RF_WH_14_SUDHA.pdf

Media

extracted text
'»/>H 12-; r
06-05-1998 1?:37

FROM

roductiveHealthMatters

TO

i3i-iiiiiia>a8O09iii

P.01

VIA
RF_WH_14_SUDHA

: r4 r
H? I-

Investigating Women’s Gynaecological
Morbidity in India:
Not Just Another KAP Survey
Michael Koenig, Shireen Jejcebhoy, Sagri Singh and S. Sridhar
This paper discussef; major lessons learnt from seven comm unity’ba sod prevalence studies of
gynaecological morbidity among poor women in India, methodological challenges associated
with this research and implications for future research: (1) dose interaction with the community is
an important determinant of a successful study; (it) appropriate medical treatment should be
provided to respondents, as it may be the only opportunity to address many women's
reproductive health problems; (Hi) it is important to complement quantitative assessments of
morbidity with in-depth qualitative research; (iv) women may significantly under-report
gynaecological problems; (v) specific questions, detailed probing and consideration ofseverity
may improve reporting validity; (vi) sample loss and self-selection due to women's unwillingness
to undergo clinical examination are serious problems: (vii) there is significant variation in clinical
definitions and criteria for gynaecological morbidity across studies. The paper concludes that
future research can be designed to achieve greater accuracy regarding prevalence and sh -iijld go
beyond a narrow medical focus to include causal factors, social and sexual behavioural issues and
consequences ofmorbidity for women's lives, and suggests allernati ves to full studies, which
voluntary groups and researchers could more easily undertake

*

ziiMHE last decade has witnessed growing

’international recognition of the magnitude
and significance of gynaecological morbi-1L dity among poor women in developing
countries.12 In docurhenting that gynaecological
problems are not corffincd to special clinic-based
populations, but are widespread within the
community at large, scientifically rigorous and
carefully conducted research has played a key
role. In India, seven community-based preva­
lence studies of gynaecological morbidity have
been completed to date.3’7 Together with several
other studies from developing countries, these
data constitute some of the most important
empirical evidence available on the magnitude of
this public health problem globally.8"11

Community-based surveys of
gynaecological morbidity in India
The seven studies reviewed here took place

4
I

between 1985 and 1994, independently of one
another, in a range of geographical and cultural
settings within India. All of them undertook
community-based surveys of gynaecological
morbidity among women, both self-reported
morbidity and clinical examinations. In five of the
studies, laboratory investigations wc*c also
carried out.12 However, as laboratory tests and
procedures varied widely across sites and were
the least comparable data collected, they are not
considered here.
Six of the studies were carried out under the
auspices of voluntary organisations with long­
standing records of addressing the health needs
of the communities studied. (Table 1) The study in
Karnataka was undertaken by a team of resear­
chers from the Indian Institute of Management.
Bangalore, who achieved very high study partici­
pation rates, although they were unaffilia ed with
an NC-O and had no record of service pr ovision
or engagement in Karnataka
1

I06-MAY-1998

%T5
22:00

44 171 242 9696

91X

P.01

ZO/t/JO

J1X.

roductlueHealthMatters

FROM

06-05-1998 17=38
pm

I'V'jt

q>-

4

131-1111111111009111

TO

P.02

Koenig, Jeji.'ebhoy, Singh and Sridhar

Table 1. Description of community-based gynaecological morbidity studies in India
Location
year of study

*60
•filiation

Maharashtra

SEARCH

(1989)

Sample

Date cofiecfion
procedure

SMh •mphnem
period

Study papal atioti

Ever merned and

100 per cent sample

Interview and

5 months,

in 2 rural villages

examination conducted

3 days a wee*

simultaneously at
village-level clinics

West Bengal

CINI

(1990-91)

unmarried women
aged 13 years and

above

Quota sample of 500

Interviewand

women in 8 rural

examination conducted

2 clinics • week.

un ma rned women

villages

simuttancouslyatthe

2-3 repeat visits

aged 13-45yvars

18 months.

Cvermamed and

central medical facility

Bombay

Streehrtakafini

(1988-89)

10 per cent random

Interview conducted

12 months, open

Ever married

sample of project

at home simultaneously

enrolment for

women aged

households in 1

with clinical examination;

clinical

15 years and above

urban slum area

examination conducted

examination over

attheNGO clinic

the ernke penod

or near home
Baroda

Baroda Citizens

50 per cent random

interview conducted

12 months; within

Ever marred

(1990-91)

Council

sample from 2 urban

at home simultaneously
with clinical examination;

2-3 days of

women aged

interview

18-45 yeers

12 months, the day
following the last

Ever married
women less than

interview

35 years of age

slum areas

examination conducted at

the NGO clime
Karnataka

No affiliation

(1992)

Subsample of a

Interview conducted

random sample of 43

monthly at home;
examination conducted

villages and 1 town

with a child

at project office clmtc

<<J morrths
Gujarat

SEWA-Rural

(1988-891

lOOper cent sample

in 10 rural villages

Interview and

2-4 days during

Ever married

examination conducted

the village health

women aged

simultaneously at village

fair

15 years and above

1

health fairs

a

Rajasthan
(19941

URMUL Trust

100 per cent sample

Interview conducted at

2 montht; 1-2dayt

Ever married

in 2 rural viflages

home, examination

in the third month

women aged

conducted at village

15 years and above

health clinics

Sampling designs and procedures varied
considerably across the seven studies. (7’ab/C 1)
Three rural studies (Maharashtra, Gujarat and
Rajasthan) enumerated and sought to obtain
Information on all eligible women in a limited
number of villages. In contrast, the West Bengal
study adopted a quota sampling approach,
whereby data collection continued until a requi­
site number of respondents from selected villages
were interviewed. Only three studies (Bombay.
Baroda and Karnataka) adopted study designs
based on representative sampling procedures.
Two studies (West Bengal and Rajasthan) found it

necessary to expand their original sampling
frames to obtain the desired sample size, due to
women's unwillingness to participate
In all the studies, the primary study population
was cvcr-married women of reproductive age,
and at least two of the studies (Maharashtra and
West Bengal) also collected data from nonmarried women. Two studies limited investiga­
tion to women of reproductive age (West Bengal
and Baroda); four included women beyond
reproductive age (Maharashtra, Gujarat, Rajas­
than and Bombay).
Given the reluctance of many women to

2

06-MAY-1998

22=01

44 171 242 9696

93X

P.02

06-05-1998 17 = 38
Xoenig

pm

FROM
rage j

roductiveHealthMatters

-e

TO

131-■■■■■■■■81009111

P.03

Reproductive Health Matters. Vol. (i. No. 11, May 1998

undergo

gynaecological

examinations,

data

Table 2. Prevalence of self-reported and

collection strategies were tailored to local condi­

clinically diagnosed gynaecological

tions in order to enhance participation. (Table 1)
In four studies (Maharashtra.

morbidity: six community-based studies.

West Bengal.

India

Gujarat and Rajasthan), structured interviews
and

clinical

examinations

were

conducted

Women reporting

together; prior to the clinical examination women
were administered a survey questionnaire. In the

Self-reported conditiont

other three studies, the survey preceded the

Menstrual proWem
Excessive discharge

clinical examination by one or more days.

lower abdominal pain’

The study enrolment periods ranged ttom (wo
or three days during health fairs in the Gujarat

study and one week in the Baroda study, to 12
months in the Bombay study and 18 months for

Lower backache*
Oyspareuma’
One or more conditions

the West Bengal study. The studies also differed

CTimctlly diagnosed condrtiow

FvrctifUffl

33-55

13-57
S-21

$•«
1-7
5SM

Vaginitis

♦-62

efforts (Gujarat and Baroda) to intensive sub­

Carvitilis
Cervical erosion

sequent efforts (Karnataka).

Pelvic inflammatory disease

8-48
2-46
t>4
<1-7

in terms of their efforts to enrol women who did
not originally participate, ranging from limited

Prolapse
One or more conditions

26-74

Main findings
There was a marked variation in patterns and
levels of both self-reported and clinically diag­

*0qc$ not include results from me Rejasrhen study,
e
community- and clinic-based samples were combine 1.

nosed gynaecological morbidity across the six

studies reporting results separately for the
community-based sample. (Table 2) The percen­
tage of women reporting menstrual disorders,
for example, ranged from 33 to 65 per cent, and
those reporting excessive discharge ranged from

13 to 57 per cent. Significant variation was also
apparent for other reported conditions, such as
lower abdominal pain (ranging from 9 to 21 per
cent) and lower backache (from 5 to 39 per cent).
Overall, the percentage of women reporting one
or more gynaecological problems ranged from
55 per cent to as high as 84 per cent in the studies

considered.
With respect to clinically diagnosed condi­

Lessons learned
We have distilled seven main lessons from these
studies to identify challenges and implications
for future research These are as follows.

Close interaction with the community is an
important determinant of the successful
completion of the studies
A common element in virtually all the studies was
the high level of interaction and rapport between

the investigators and the communities Studied. All

tions, the prevalence of vaginitis ranged from 4

but one of the studies were carried out through

per cent to 62 per cent, cervicitis from 8 per cent

voluntary organisations with a long-standing

to 48 per cent, cervical erosion from 2 per cent to
46 per cent, and pelvic inflammatory disease

record of service to the communities, either by the
organisations themselves or by researchers
working under their auspices. However, careful

from 1 per cent to as high as 24 per cent. Overall
the percentage of women with one or more
clinically diagnosed conditions ranged from 26
per cent in the Baroda study, to 40 per cent or
more in the Gujarat and West Bengal studies, to
70 per cent or more in the three other studies
where data were available n

4

and exhaustive preparatory efforts and engage­
ment with the communities were nonetheless
required in order for the researchers to gain their
co-operation and active support.
Given the asymmetric nature of gender
relations in India, convincing the men in the
community and male family members of the
rationale and need for the study assumed

3

e

*

06-MAY-1998

22=01

44 171 242 9696

93X

P.03

FROM

06-05-1998 17=39

I

Aoemg

zo/«/>c

pm

roductiveHea1thMatters

TO

131-II1IIIIIII0091U

--cp

rage <i

P.04

I

Koenig, Jcjcvbhvy, Singh and Sndhar

paramount importance in successfully enlisting
the participation of women The Maharashtra
study, for example, held extensive meetings with
village leaders, as well as men and women in the
village separately, to inform them of the rationale
for the study, male village volunteers also played
a critical role in organising the survey and
clinical field components Similarly, both the
Gujarat and Rajasthan studies undertook
extensive advance publicity activities in an effort
to inform and persuade members of the
community - both male and female - of the
merits of their study. Although not as well
documented, it Is clear that three of the other
projects (West Bengal, Baroda and Bombay) also
engaged in intensive mobilisation efforts to
promote women's participation in their studies,
using the infrastructure from their existing
community health programmes

Rajasthan study, URML’L Trust continued to
follow up women who participated in the study,
and went so far as to establish a fund to defray
the costs of any needed surgical procedures for
conditions such as prolapse. The Maharashtra
Study provided treatment free of charge for all
health problems to all participating women and
for other members of the women's families as
well The Karnataka study provided medicines
for simple problems, and referred more
complicated cases either to local hospitals or to
the examining physician based in Bangalore, 70
km away The Baroda study treated less serious
cases through their clinic and referred more
complicated cases to the municipal government
medical college.
'Chough not well documented, there seemed to
be considerable variation in the extent to which
each project assumed ongoing responsibility for
resolving the health problems identified, or in
guiding women with more complicated cases
through what constitutes an intimidating system
of higher medical care, especially for many who
are poor.

Appropriate medical treatment should be
provided to respondents as a component of
the study, because for many women this may
represent the only opportunity to address
their reproductive health problems
The provision of medical treatment as a pan of
any study of women's reproductive health
problems should be considered mandatory in
settings such as India, where few poor women
have access to high quality gynaecological
care 14 It is evident from the studies reviewed
that arranging required care may be a pre
requisite for enlisting women's co-operation and
participation in studies of this nature, as well as
an ethical imperative. It is unlikely that many
women would willingly acknowledge sensitive
problems or agree to submit to what constitutes
an invasive and embarrassing examination for
them, if effective treatment was not provided.
Studies differed in the extent to which they
included medical treatment within their study
designs. The Gujarat study, through health fairs
in each village, offered detailed health check-ups
to all married women in the villages who
attended, and less comprehensive check-ups for
other community members The Bombay, Rajas­
than, Gujarat and West Bengal studies provided
medicine to women with readily treatable
conditions, and in more complicated cases,
referred women to known medical facilities for
further investigation and treatment. In the

It is important to complement quantitative
assessments of gynaecological morbidity
with information derived from in-depth
qualitative research
in almost all the studies, considerable emphasis
was placed on understanding women's per­
ceptions of morbidity, and in establishing local
terms and expressions for describing gynae­
cological complaints that were understandable
and salient to the women themselves. For
example, in rural Maharashtra women used 12
different expressions for and distinguished five
main categories of while discharge.15 Detailed
ethnographic research in the Bombay study
provided insights into the range of terms used by
slum women for conditions such as vaginal
discharge and prolapse 16 A detailed knowledge
of the local language and terminology also
assisted investigators in understanding the
subtle and more indirect ways in which women
describe some gynaecological problems. For
example, women in the Maharashtra study
experiencing white discharge often described
this as 'weakness'. Without such knowledge,
important conditions arc likely to go undetected
by both researchers and service providers 17

4

;-'l I06-MAY-1998

22=02

44 171 242 9696

91X

P.04

06-05-1998 17.’40

FROM

roductiveHealthMatters

TO

131-IRIIIRIIBIC09111

P.05

Rcproductivt HrJllh Miners. Vol. r>. Nq. 1i, May JWa

*

In-depth research allows the investigator to
explore women's attitudes to revealing or dis­
cussing certain kinds of morbidities, and
provides an idea of morbidities that might be
under-reported from a single question in
conventional surveys. For example, the West
Bengal team highlighted women's reluctance to
classify certain conditions as problems (eg.
dysmenorrhoea or pain during intercourse), and
even to view some conditions as indicative of
good health (excessive bleeding).
Exploratory in-depth research can also offer
important insights into beliefs regarding the
causes of morbidity, the types of treatment
considered appropriate for different problems,
treatment-seeking decision-making and beha­
viour, and the consequences for women’s lives.
Finally, qualitative information offers the oppor­
tunity to corroborate findings from standard
survey techniques on the magnitude of morbidity.
For example, the focus of attention in almost all
studies on leucorrhoca - the terminology to
describe different kinds of discharge, associated
symptoms and probable causes is consistent
with the common finding that excessive dis­
charge is indeed the most widespread gynaecolo­
gical morbidity reported.

In self-reported responses on
gynaecological morbidity, under-reporting
ofproblems by women appears to be a
significant methodological problem
Two sources of data provide strong evidence of
significant levels of under reporting of gynaeco­
logical problems in response to survey questions
The first comes from the study in Karnataka,
where a subsample of 440 women from a larger
study were interviewed monthly for one year
regarding gynaecological morbidity in the
preceding month.5 The percentage of women
reporting gynaecological problems rose steadily
over the course of the study, from 8 per cent in
the first round to 33 per cent in the final round,
with the increase particularly notable for
excessive discharge (from 5 to 19 per cent).
Similar results were evident from a study in
Haryana in north India (not otherwise reviewed
here), which as part of a larger prospective study
included six bi-monthly visits to a sample of 276
women to inquire about reproductive morbidity,
among other health problems, in the preceding

15 days. A sharp increase in the percentage
of women acknowledging gynaecological
problems was reported in that study also, from
5 per cent in the first round to 38 per cent in the
last round.,n
While the possibility of over-reporting of
some gynaecological problems as these studies
progressed cannot be ruled out, a more likely
explanation, in our view, is significant under­
reporting of gynaecological problems, especially
in the early rounds. Several factors may
contribute to women's increased acknowledge­
ment of gynaecological problems - women
becoming increasingly comfortable with the
interviewers and more confident that confi­
dentiality would be safeguarded; interviewers
becoming more proficient in eliciting informa­
tion; and women increasingly coming to view
such conditions as legitimate health concerns
and problems, rather than women's 'lol in life'
which they have been conditioned to tolerate.R
The reporting of gynaecological morbidity
may also be strongly influenced by the back­
ground of the interviewer. Morbidity levels
reported to medical practitioners or to inter­
viewers shortly preceding clinical examinations
tended to be significantly higher than those
reported to interviewers yi home visits, a finding
also reported cl$cwhertro In the Karnataka
study, for example, whilfronly 15 per cent of all
women reported menstrual problems during
interviews in their homes one day preceding the
Clinical examination, as many as 65 per cent
reported a problem the following day to the
examining clinician.5 In the Maharashtra
study, while 21 per cent of women reported
menstrual disturbances to a nun-medical inter­
viewer, this percentage rose to 58 per cent with
the examining physician.4 Similar differences
were observed in the Baroda and Bombay
studies; differences were so large that both
Studies chose to rely only on responses made to
the clinician.
A contributing factor may have been the
greater propensity of the physicians thin the
investigators, who mainly had a social work
background, to probe thoroughly for possible
morbidity. The Bombay study also observed
that although non-medical investigators already
had a rapport with the community and explained
at length the need for this study, they faced
reluctance from respondents to answer their

5

06-MAY-1998

22 = 03

44 171 242 9696

89X

P.05

06-05-1998 1?:41

FROM

roductiveHealthMatters

_

---

TO

131 ■■■■■■■■■■009111

P.06

Koenig. Jejeebhoy. Singh tnd S^dhtr

questions, given their inability
medical problems.

to

resolve

Framing specific questions, detailed
probing, and consideration of severity may
significantly improve the reporting validity
of gynaecological morbidity and reduce
under- and over-reporting

Gynaecological morbidity may be underreported due to the Widespread perception
among many women that such conditions arc
normal and merit neither acknowledgement nor
complaint. Special efforts are therefore required
to elicit such information from women.
The Indian studies illustrate the importance of
question specificity and probing beyond respon­
dents' initial responses in ascertaining the extent
of gynaecological morbidity The specificity of
questions varied considerably across studies. In
the Baroda and West Bengal studies, for example,
women were asked open-ended questions on
problems they were experiencing, and their
answers were checked against a pre-recorded list
of symptoms on the questionnaire The Gujarat,
Maharashtra and Bombay studies, in contrast,
queried respondents using a checklist of indi
vidual gynaecological symptoms, and asked
women about each one. The Rajasthan study
adopted a two-stage approach, whereby women
were initially questioned about any problems that
they had at the time, and subsequently about a
scries of other specific conditions and symptoms
that had not been mentioned earlier.
Several of these studies illustrate the import
ance of detailed probing. In the Rajasthan study,
after women listed their gynaecological prob­
lems, they were prompted for symptoms that they
had not reported earlier using a check list-type
question; 'You have mentioned., symptoms. Do
you have any of the following problems?'. If
women reported additional symptoms on probing,
they were asked why they had omitted this
problem - a common answer was embarrassment
Prompting by interviewers markedly raised the
percentages of women reporting specific
conditions; from 28 per cent to 48 per cent for
menstrual problems, and from 10 per cent to 20
per cent for prolapse The most marked increase
was in the percentage of women reporting
problems during sexual intercourse, from 1 per
‘ cent to 48 per cent Significant increases were also

evident m the percentages of women reporting
vaginal discharge and infertility.21
In other studies, probing was less systematic,
yet there is evidence that where it was built in.
reported rates of morbidity were correspon­
dingly higher. For example, only the Gujarat
study probed specifically for dysuria (burning on
urination), probably accounting for the much
higher rate of this condition reported in this study.
Similarly in the West Bengal study, only 2 per cent
of respondents spontaneously mentioned pain
during intercourse as a problem in the larger
survey, while 15 per cent of participants m focus
group sessions acknowledged this problem on
probing As one woman observed; 'Does anyone
discuss this even if she suffers?22
These results suggest that while all gynae­
cological problems should be probed, those that
are linked directly with sexual behaviour are
likely to be most sensitive in terms of reporting.
A more generic problem that all such studies
face is the absence of agreed-upon definitions of
morbidity, and a resulting variation in specificity’
with which individual conditions are described.
While the .Maharashtra. Rajasthan and Bombay
studies inquired at length about vaginal
discharge (eg. colour, odour, consistency and
quantity), the Gujarat, Baroda and Karnataka
studies did not; this may have contributed to the
relatively lower reported rates in at least some of
the studies. In at least the Gujarat and Baroda
studies, women were not asked about dyspareunia (painful intercourse).
It is. also possible that some gynaecological
morbidities are over-reported. One reason for
this is that questions on severity are omitted, eg
on the level of discomfort or pain, and on the
extent to which a reported morbidity interferes
with a woman's daily routine and responsi­
bilities. For example, it is likely that in the
absence of specific questions on colour, texture
and odour when asking about discharge, some
women who have normal vaginal discharge were
recorded as having leucorrhoea.
A more general and problematic issue is that
questions haw often tended to be vague and open
to subjective interpretation, eg. 'scant/ periods or
'exccssivc/abnormal' vaginal discharge Although
some of the studies attempted to qualify
meanings, eg 'Are your periods scanty? For how
many days do you bleed?, others did not, making

■I

comparisons problematic.

6

I

06-MAY-1998

22=04

44 171 242 9696

89X

P.06

06-05-1998 1?:41

FROM

roduct1veHealthMattens

TO

131-aillllllll009111

P.07

Y
HcprcxiucT/vc Health Matters. Vol. 6. No 11, May 199fl

r-' ■

Table 3. Participation rates for the clinical
component of community-based
gynaecological morbidity studies in India

% Wom«it
participating

Stedy

Sample lor c!tn«cal
examinttjoa

Karnataka

3K «orn«n

86

Bombay

756 women

72

Baroda

548 women

65

Maharashtra

650 women and unmarried girls

59

West Benzil

500 women and unmarried girls

44

Gujarat

324 women

29

RfjwtMn

51 women

15

Significant sample loss and self-selection
due to women's unwillingness to undergo
clinical examination represents a serious
methodological problem in community­
based studies of gynaecological morbidity

a

than studies highlight the importance of male
opinion and the inability to blunt opposition from
male community members as a primary factor
behind the lower than expected enrolment rates
A frequently mentioned reason for this opposi­
tion was a lingering distrust of the government

■■

Significant sample loss for the clinical compo­
nent was a common feature in almost all the
studies considered (Tabic 3) At one extreme
were two of the rural studies in Rajasthan and
Gujarat where despite intensive efforts at
recruitment and a long-standing presence in the
communities, only 19 per cent and 29 per cent,
respectively, of sampled women were success­
fully examined. Of the other two rural studies,
the West Bengal study successfully enlisted 44
per cent of selected respondents for the clinical
phase, while the Maharashtra study successfully
enlisted 59 per cent of women of reproductive
age.23 The two urban studies achieved higher
participation rates: 65 per cent for the Baroda
study and 72 per cent for the Bombay study (the
latter aided by a 12-month open enrolment
period for the clinical examination). The
Karnataka study stood out, however, having
successfully enrolled 86 per cent of sampled
women in the clinical phase. It should be noted
that the clinical component of this study took
place after one year of regular contact with the
women, consisting of monthly interviews regar­
ding their own and their children's health status.
Non-compliance occurred for a number of
■ reasons. The West Bengal, Bombay and Rajas­

family planning programme and a fear that
gynaecological examinations would prove a
guise for conducting sterilisation procedures. In
some cases, restrictions on the availability of
trained medical personnel also contributed to
lower participation rates.
In the Gujarat study, for example, while
almost half the women who attended the health
fairs expressed a willingness to undergo clinical
examination, a much smaller percentage did so,
largely because a gynaecologist was available
only on certain days of the fair and many women
were reluctant or unable to come back By far the
most significant reason for non-compliance,
however, was that women who had no apparent
or significant symptoms of reproductive morbi­
dity were unwilling to consent to a clinical
examination; this was found in the Gujarat,
Bombay, Baroda and Rajasthan studies.
There is evidence to suggest that women with
reproductive health problems may be signifi­
cantly more likely to consent to a clinical
examination than women without apparent
symptoms. In several studies, it was possible to
assess the extent of selection bias by comparing
the reported morbidity of women who under­
went clinical examination with those who
refused.** In the Gujarat study, where sample
loss for clinical examination was pronounced,
the percentage of women reporting menstrual
problems was somewhat higher (85 versus 70
per cent) and vaginal discharge significantly
higher (56 versus 22 per cent) among examined
versus non-examined women. In the Baroda
study, the reported prevalence of white
discharge was notably higher among examined
versus non-examined women (22 versus 15 per
cent) though no differences were evident in
reported levels of menstrual problems. In the
Maharashtra study, while reported menstrual
problems were similar or even lower in the
examined group, reported vaginal discharge was
higher among women who participated in the
clinical examination compared with a random
sample who refused (14 versus 8 per cent).
Selectivity was most apparent in the

<T

7

4>
06—MAY-1998

22=04

44 171 242 9696

89X

P.07

06-05-1998

Koenig

17=42

28/4/98 5:23 pm

FROM
Page

roductiveHealthMatters

TO

131-1111111111003111

P.08

Kucnig. JcjecDhoy, Singh and Sridhar

Table 4. Clinical definitions of morbidity employed in community-based studies of

gynaecological morbidity

Mah»rjshva

Vafiftith

Cervical
•ro*ion

Cervicrti*

Inflammation of vaginal

All efojtons

waR and lab tvidence

Pelvic wiflammatory
disaaca

FW«p»e

Cervical inflammation

Palpable and tender

Based on traction

with pus discharge and

tissues around uterus,

and straining

lab evidence of infection

cervical motion
tenderness and lab
evidence of infection

West Bengal

Inflammation of vaginal

All erosions

Cervical inflammation

Tenderness or fixity

wall wnh or without

with pus discharge and

of tissues around

discharge and lab

lab evidence of infection

uterus end lab

evidence of infection

evidence

Bombay

Inflammation of vaginal

All erosions

wall with discharge and
itching/pain

Baroda

Based on straining

Cervical inflammation

Lower abdominal pun

Based on traction

with pus discharge or

and tenderness or

and straming

rases with inflamed

rhickeningoftrssues

scarred cervix

around uterus

Inflammation of vaginal

All erosions,

Cervical inflammation

Tenderness or fixity

wall

except small,

with pus discharge or

of tissues around

and straining;

asymptomatic
onaa

cases with inflamed

uterus

excludes mild

Based on traction

scarred cervix

cystocele and

Grade I descent

r

Karnataka

Inflammation ol vagina l

All erosions,

Cervical inflammation

Utenne tenderness,

wall with discW|e

except small,

with pus discharge

tenderness of
tissues around uterus

asymptomatic

and clinical cervicitis

ones
Gujarat

Rajasthan

Inflammation of vaginal

Based on straining

Alt erosions

discharge

cases with scarred cervix

uta r us

Inflammation of vaginal
wall with abnormal

Cervical inflammation

Abdominal pam and

with pus discharge

utenne tenderness

All erosions

NA

Tenderness or fixity
of tissues around

wall with abnormal

Cervical inflammation
with pus discharge or all

Based on traction
and straining

discharge
NA = could not bo oscertoinod from available data

Rajasthan study; of the 274 women originally

An additional source of variability, which may

selected for interview, the 51 women who agreed

have contributed to the overestimation of clinic­

to a physical examination were all symptomatic

ally diagnosed morbidity, is the lack of consis­

These results suggest that the overall prevalence

tency across studies in clinicians' diagnoses of

biased

gynaecological morbidity. Undoubtedly a contri­

upward as a result of sample selectivity, with the

buting factor is the absence of consensus and

bias most pronounced in studies with higher

uniformity within medical

rates of sample loss.

medical pro fession concerning diagnostic crit­

of gynaecological

morbidity

may

be

textbooks and the

eria for gynaecological morbidity'.

Signtflcani variation was evident across studies

Clear differences were evident among studies
respect to clinical definitions employed

in clinical definitions and criteria employed for

with

gynaecological morbidity. In many cases,

(fable 4) For example, while most studies used

definitions may have led to an overestimation of

inflammation of the vaginal wall plus discharge

levels of clinically diagnosed morbidity

as their basic definition of vaginitis, the Mahara-

8

06-MAY-1998

22=05

-4

44 171 242 9696

93X

P.08

06-05-1998 17’43

FROM

roductiveHealthMatters

TO

i3i-iiiiiiiiiiea9iii

P.09

-T~.
Reproductive Health Matters, VqI. fi. No. II. Muy

largely to standard demographic or socio­
economic variables.
In the studies which did investigate determi­
nants, some significant (although not entirely
consistent) associations were found between
demographic and socio-economic characteris­
tics and childbirth practices, and specific self­
reported or clinically diagnosed gynaecological
problems.25-^ Perhaps the most sinking finding

shtra and West Bengal studies were more
specific, using laboratory evidence of infection.
For cervical erosion, while five studies included
all erosions, two studies (Baroda and Karnataka)

excluded small, asymptomatic erosions.
Clinical definitions also differed significantly
with respect to cervicitis. While three studies

(Bombay, Baroda and Gujarat) included all cases
of cervical inflammation with pus discharge

z)-

to emerge was the strong association between

and/or evidence of prior or present scarred
cervix, two (Karnataka and Rajasthan) considered
only current inflammation with pus, and two
(West Bengal and Maharashtra) also required
laboratory evidence of infection. In diagnosing
pelvic inflammatory disease, different combina­
tions of a wide range of criteria were employed,
including abdominal pain or tenderness, cervical
motion tenderness, uterine tenderness, adnexal
tenderness or fixity, palpable adnexae, thickened
vaginal fomiccs, clinical cervicitis, and/or
laboratory evidence of systemic infection.
Among the six studies for which data on
prolapse are available, descent below the ischial
spine served as the basic criterion, with four
studies utilising both straining and cervical
traction in diagnosing this condition, and two
studies using straining only. In the Baroda study,
mild cases of cystocele and uterine prolapse were
not recorded.
The fact that the studies were independently
designed and conducted served to increase the
likelihood of variation in clinical definitions of
gynaecological morbidity The possibility of
significant intra-study variation also cannot be
ruled out, since several studies involved more
than one gynaecologist.

Significant lacunae persist in understanding
the factors influencing such morbidity as
well as their consequences for women's
lives

In almost all the studies reviewed, the primary
objective was to estimate the prevalence of
gynaecological morbidity within the commun­
ities studied. Reflecting the predominantly
biomedical or epidemiological orientation of the
researchers in these studies, comparatively less
emphasis was placed on understanding the
social, behavioural or biomedical antecedents of
that morbidity. The data on the possible determi­
nants of morbidity were therefore confined

use of female sterilisation or in some cases the
JUD, and reported or diagnosed gynaecological
morbidity, although the causal mechanisms, if
any, remain poorly understood.4Similar
findings have been reported from studies in
fcgypt and Bangladesh,*9 while a study in Turkey
found no such association.”
These findings aside, gaps exist in our
understanding of the determinants of gynae­
cological morbidity - behavioural, environ­
mental, iatrogenic and biomedical - in increasing
women's vulnerability. These include sexual
behaviour and practice, the role of male partners
and their sexual and reproductive health, and
men's role in assisting or impeding women's
ability to address and resolve reproductive
health problems
Existing studies also shed little light on the
consequences and implications of gynaecological
morbidity for poor women's daily lives Little is
known about how such morbidity impacts on
women's ability to fulfil (heir various roles economic, domestic, marital and sexual - or on

1

their mental health and well-being. Moreover,
gynaecological problems include a continuum of
conditions which range from inconvenience to
extreme disability or even death; in the absence of
more detailed research on severity, there has
been a tendency to view all such conditions as
equivalent. It is clear, however, that from the
perspectives of both individual women and public
health policy, these conditions have widely
differing implications in terms of severity,
priority and required intervention.

Discussion
Community-based surveys of gynaecological
morbidity
based
upon
self-reports
hold
considerable value, as long as investigators arc
fully appreciative of both their strengths and
their limitations. Data from self-reports should
9

(I)
06-MAY-1998

22:06

44 171 242 9696

89X

P.09

06-05-1998 17:43
ALTViixy

FROM
pm

t-age

roduct iveHea. IthMatters

TO

i3i-mimm009i 11

P. 10

iu

Koanig, Jqwbhoy, Singh and Sridhar

:)■

not be interpreted as providing iaccurate
estimates of the true prevalence of■ gynaecological morbidity in a community. They have
proven useful, however, in assessing women's
perceived problems and need for gynaecological
services, in understanding existing patterns of
treatment-seeking behaviour and expenditure,
and in understanding the impact on women's
lives They have been less effective in elucid
ating the precise biomedical nature of illness
or infection, as evidenced by the very low
levels of correspondence between self-reported
and
clinically
or
laboratory
diagnosed
morbidity.4'20'*7*8
These studies from India, however, also argue
for caution in presuming the simple adaptation
of conventional knowledge, attitude and practice
(KAP) survey methodologies for investigation of
self-reported gynaecological morbidity, given
the complexity and cultural sensitivity of this
issue Our review has identified a number of
concrete steps for improving the reliability and
validity of future community‘based studies based
upon self-reported morbidity, including in-depth
understanding of local terminology used by
women to describe problems; greater standard­
isation in questionnaire design,- more carefully
worded questions defining morbidity and
ascertaining its severity; inclusion of detailed
probes; and when possible, undertaking repeat
interviews rather than single-round, crosssectional surveys.29
The addition of clinical and laboratory
components provide important complements
to self-reported community morbidity surveys
In addition to considerations of cost, however,
participant compliance remains a central
concern, especially for Che clinical component.
The establishment of trust and rapport by the
investigators is an essential pre-condition. Our
review has identified a number of concrete steps
for achieving this, including intensive advance
engagement with the communities under study,
working under the auspices of known voluntary
organisations, and building in comprehensive
medical care for detected reproductive health
problems as an integral component of the Study.
A central lesson to emerge from this review is
(hat despite the best efforts of researchers,
significant sample attrition for the clinical
component of community-based studies is
frequently unavoidable in conservative settings

such as ^rrh India. Participation is likely to be
most problematic in settings where women's
ability to discuss gynaecological morbidity is
highly constrained, and where women lack
autonomy to control resources or seek treat­
ment. Future study designs must anticipate that
nun-symptomatic women are more likely to
refuse to participate and build in appropriate
mechanisms to learn more about morbidity
patterns among those who do refuse, to be able
to gauge more accurately the effects of selection
bias on results.
together with the groundbreaking study by
Bang et al., the studies reviewed here represent
some of the earliest efforts to document and
understand the prevalence of gynaecological
morbidity in developing countries. The wide
variations found in both self-reported and
clinically diagnosed levels and patterns of
gynaecological morbidity may be a reflection of
the extreme geographical and cultural hetero­
geneity of India and innate differences across the
populations .studied A$ with much pioneering
work, however, many of the studies considered
are Characterised by serious limitations with
respect to both study design and methodology,
rendering many of the findings non-comparable
and in many cases, inconclusive.
For all these reasons, there is scope for further
research on gynaecological morbidity in India in
order to boiler understand morbidity levels,
patterns and variability. In general, the larger
and more heterogeneous the setting, the greater
the rationale for multiple studies; the case for
multiple studies in more homogeneous sellings
is likely to be less persuasive In addition lo
providing more precise estimates of prevalence,
the next generation of research studies must
advance beyond an exclusive focus on disease
prevalence. Subsequent research must also seek
to illuminate the socio-economic, behavioural
and biomedical causes and consequences of
reproductive morbidity for the quality of
women's lives, and the factors that impede
women from prioritising these problems and
seeking appropriate treatment. Such studies will
require more broad-based, multi-disciplinary
approaches and expertise, and an array of skills
which relatively few projects thus rar have been
able to assemble.
There are a rapidly growing number of
researchers and community organisations in

io

06-MAY-1998

22 £06

44 171 242 9696

94X

p. 10

08-05-1998 09:34

roductiveHealthMatters

ppori

I

TO

131-IIIIIIIIII009111

p.01

——

'
Reproductive Health Mailers,. Vol 6. No H. Mav

India (and elsewhere) who are interested in
policy and programme shift towards a repro­
undertaking community-based studies on the
ductive health approach,34 and there are plans to
prevalence of gynaecological morbidity. Given
introduce interventions for prevention, recogni­
the cost,30 the time-intensiveness and the array
tion and rrcalinent at all levels of the health
of technical skills required to conduct such
system.”'5
rigorous research, full-scale community-based
More precise and detailed information on
prevalence studies of gynaecological morbidity
cause-specific gynaecological morbidity would
are likely to prove beyond the capacity of most
clearly be helpful when difficult choices arise
community-based organisations?1 Several alter­
with regard to affordability, cost effectiveness
natives exist, however.
and feasibility of specific services, especially in a
One is the organisation of interventions such
resource-poor context such as India's public
as women's health fairs. Women who participate
sector health care system 36 Further research can
in these special interventions, while clearly
provide insights into the constraints and barriers
unrepresentative of the general population, do
which women presently face in accessing care
provide an indication of the need and demand
for these problems, and how current health
for reproductive health care within the broader
delivery must be reformulated
community. A second approach remains surveys
At the same time, the absence of such data
of gynaecological morbidity based solely upon
should not obscure the fact that sufficient
women's self-reports, which provide at least a
information exists to undertake many of the
partial picture of morbidity and its sequelae within
required programme changes - most notably, a
the community. A third, under-utilised approach
shift away from a narrow family planning target
is through studies of clinic-based populations (eg.
focus, an emphasis upon quality of care and
women attending family planning or antenatal
client-orientated services, the reorientation and
clinics), for whom participation may be less
sensitisation of health staff to recognise and
problematic and who may derive direct benefits
treat reproductive tract infections in a non­
in the form of detection and treatment.12 Finally,
stigmatising manner, and the design of inter­
the introduction of new, Jess intrusive field
ventions to enable women to understand,
diagnostic techniques (eg. self-administered
prevent, and address gynaecological problems.
vaginal swabs, urine-based diagnosis) holds
considerable promise for complementing and
Acknowledgements
extending the utility of existing research and
This is a modified and abridged revised version of
increasing participation levels '3i
a paper entitled 'Undertaking Community-based
There will undoubtedly remain a need to
Research on the Prevalence of Women's
assess systematically the impact of intervention
Gynaecological Morbidity: Lessons from India’,
programmes on the prevalence of gynaeco­
to be published by the International Union for
logical morbidity. Given the cost, complexity and
the Scientific Study of Population (1USSP) in
organisational expertise required to carry out
a forthcoming volume entitled The Assessment of
such studies successfully, these impact assess­
Reproductive Health: Innovative Approaches. Al
ments should, in our view, be limited to a small
Mundiqo and O Campbell (eds). It is published
number of carefully designed and selected
here with kind permission of rhe IUSSP Wc are
experimental intervention studies
also indebted to current and former.members of
While the Indian studies raise as many
the Seven research projects for taking (he time and
questions as they answer with respect to
effort to provide additional details and infor­
prevailing levels and patterns of gynaecological
mation about each study, and to John Cleland and
morbidity, they leave lirtle doubt that such
Jane Hughes for helpful comments.
morbidity constitutes a major public health
problem in India. Together, these studies have
Correspondence
been an important catalyst In building a con­
Michael Koenig, Ford Foundation. 55 Lcdi Estate.
sensus to address Illis neglected dimension of
New Delhi 110 003, India. E-mail:
reproductive health. The management of repro­
m koenigt&fordfound org
ductive tract infections occupies a prominent
place within the Government of India's recent

1

11

08-MAY-1998

13=57

44 171 242 9696

90X

p.01

08-05-1998 09=35

roductlueHealthMatters

FROM

TO

i3i-aiiiaiiiii009iii

P.02

Koenig, Jejvcbhoy, Singh and Sridhur

References and Notes
1. Dixon-Mueller R. Wasserhclt .1,
1991. The Culture of Silence:

2

K |eds). Population Council, New
York

Reproductive Traci Infections

9. Wasserheit JN. I larris JR.
Chakraborty J et al V)89.

Health Coalition, New York.

a family planning population in

Germain A. Holmes KK. Piot P et

possible that the higher

rural Bangladesh. Studies m

prevalence estimates in the later

bamily Planning. 20(2)-69-80.
10. Brabin L. Kemp J, Obunqe OK et

underestimates of true levels of

Infections: CloOal Impact and

Reproductive tract infections in

Priorities for Women's

al. 1995 Reproductive tract

Reproductive Health Plenum

infections and abortion among

Press. New York.

adolesc ent girls in rural Nigeria

3. The studies are by SEARCH in
rural Maharashtra: Child-in-

Need Institute (CINI) in rural
West Bengal; Streehitakarini in
Bombay; Bar<>da Citizens
Council in Baroda. Gujarat;
Indian Institute of Management.

Bangalore, in rural and urban

Lancet 8945 (4 February);
300-04

18 Council for Social Development,
unpublished results.

19. Even with these factors, it is

rounds represent significant
gynaecological morbidity, given
the absence or low levels of
reported conditions, such as
prolapse or dysuria
20. In a study in Istanbul, for

11 Bulut A. Filippi V. Marshall Tet

example-, while 81 per 1 ent of the

al 1997 Contraceptive choice

women reported experiencing
one or more symptoms

and reproductive morbidity in
Istanbul. Studies in Family
Planning. 28(l):35-43.

12. These were undertaken in

associated with reproductive

morbidity when interviewed by
a physician, the corresponding

Karnataka.- SEWA-Rund in rural

Maharashtra, Bombay, West

figure when interviewed al

Gujarat; and URMULTrust in

Bengal, Karnataka and

home by an interviewer was 65

Rajasthan

percent. See Bulut a. Yolsay N.
Filippi V et al. 1995 In search of

rural Rajasthan. Two other
studies in rural Gqjarat were

13 Although this information was

cither discontinued or under

not reported separately in the
Maharashti*a study, it is evident

way at the time of writing. See
notes 23 and 31

4. Bang RA. Bang AT. Baitule M ct
al 1989. High prevalence of

gynaecological diseases in rural
Indian women. Lancet. 8629(14
January):85-88.
5. Bhatia JC, Cleland J, Bhagavan L
etal. 1997. Gynaecological

morbidity in south India. Studies

that the overall level of clinically
diagnosed morbidity in this

study also exceeded 70 per cent
of all women
14. The Maharashtrasludy, for
example, reported thatonly 8
per cent Of women respondents
had ever undergone a

truth: comparing alternative
reproductive tract infection.
Reproductive Health Matters b
lNov).-31-3y

21. Since this Study population was

comprised largely of selfselected clinic attendees who
would be expected to volunteer
information more readily, an

gynaecological examination. See
note 4

Oomman NM, 1996. Poverty and

15. Rang R. Bang a, 1994. Women's

responses might be expected

Rajasthani Women's

perceptions of white vaginal
discharge: ethnographic data

Ethnomedical Models with

from rural Maharashtra

among the- general population of
women.
22. Bhattacharya S. 1994

Biomedical Models of

Listening to Women Talk about
Their Health. Giltelsohn J et al.

Reproductive Behavior. PhD
thesis. Johns Hopkins
University.
7. Latha K. Kanani SJ, Maitra N «t

al 1997. Prevalence ot clinically
detectable gynaecological

morbidity in India: results of
four community-based studies.
Journal ot Family Welfare.
43(4):8 16.

Younis N, Khattab H, Zurayk H
el al. 1993. A community study of
gynaecological and related
morbidities in rural Egypt.

(eds). Har-Anand Publications,
New Delhi
16. Streehitakarini. 1995
Gynaecological diseases ai*d
perceptions about them in a
Bombay slum area.
(Unpublished report.)
17. Bang R. Bang A, 1996 A

1

sources of informanon on

in Family Planning. 28(2) 95-103

Pathology. Comparing Rural

8

24(3):175-86

among Women in the Third
World Intt-rnational Women's

al. 1992. Reproductive Tncr

f>

Sludie*; in Family Planning

even larger gap between

unprompted and prompted

Gynaecological morbidities
among women in West Bengal
Child-in-Nerd Institute

(Unpublished report.)
23. Preliminary x-csuits from a

separate study in Giyarat
indicate that only 46 per cent of
sampled women consented to
undergo a clinical examination.
See Visaria L 1997.

community study of

Gynaecological morbidity in

gynaecological disease m Indian
villages: iome experiences and

rural Gujarat: some preliminary
findings. Gujarat Institute of

reflections. Lrarrung about
Seauahty Zcidenstein S Moore

(Unpublished paper.)

Development Research.

12

08-MAY-1998

13=58

44 171 242 9696

90X

P.02

08-05-1998 09:36

roductiveHealthMatters

FROM

TO

131-llliaillVI009111

P.03

T

i

Tieptoductiy.e Htuilh Matter*. Vo/, o. /Vn. 11. May 1998

3-

24. This was only possible in studies
which fielded their surveys of
sell-reported morbidity among
all sampled women prior to and
separately from the clinical
examination, or who returned to
interview a representative
sample of women who refused
the clinical component
25. See Bhatia JC, Cleland J, 1995.
Self-reported symptoms of
gynaecological morbidity and
their treatment in south India.
Studies in Family Planning.
26(41:203-16. The Karnataka
analysis was based on data from
a larger cross-sectional survey.
26. Parikh 1, Taskar V. Dharap N et
al. 1996. Gynaecological
morbidity among women in a
Bombay slum. Streehitakarini
(Unpublished paper.)
27. Zurayk H, Khattab H. Younis N
et al. 1995. Comparing women's
reports with medical diagnoses
of reproductive morbidity
conditions in rural bqypt
Studies in Family Plunninq.
26(1):14-21.
2B. Filippi V, Marshall T, Bulut A et
al. 1997, Asking questions about
women's reproductive health,
validity and reliability of survey

findings from Istanbul Tropica!
Medicine and International
Health. 2(l);47-56.
29 Graham W, Ronsmans CCA.
Filippi VGA ct al. 1995. Asking
Questions about Women s
Reproductive Ilea hh in
Community -bused Surveys:
Guidelines on Scope and
Content. London School of
I lygkne and Tropical Medicine.
Maternal and Child
Epidemiokxp' Unit Publication
No 6, April.
30. lhe field data collection costs,
exclusive of data entry and
management, from an ongoing
clinic-based study of
gynaecological morbidity in
Maharashtra ar? estimated at
roughly US$50 per respondent
Arundhali Char, personal
communication. 1998. The pvrcase costs for community-based
studies of gynaecological
morbidity are likely to be
substantially higher.
31 For a discussion of this issue, see
Khanna R, 1997. Dilemmas and
conflicts in clinical research on
women's reproductive health.
Reproductive Health Matters
9(MayM68-73.

32. See, for example, HopcraH M,
Verhagen AR. Ngigi $ ex a!
1973. Genital infections «n
developing countriesexperienc*? in a ranr.il> planning
clinic. Bulletin of the World
Health OrqanizancM} 4«:531-86
33. For a discussion of dlagm^tic
techniques, see Morse SA.
Moreland A A, Holmes KK, 1996.
Arfas of Sexually Transmitted
Di\e>i*e$ and AIDS MosbyWoife, I ondon.
34. Sec Mcasham AR Heaver RA.
1996 India's Family Welfare
Program: Moving to «
Reproductive and Child Health
Appffyttch. World Bank,
Washington DC
35. Government of Indii. n d.
Reproductive and Child Health
Programme. Ministry <rf Health
and Family Welfare. New Delhi.
36 It has been estimated dial India
presently spends roughly
US$i) w per capita on maternal
and child health and family
planning services, against the
World Bank's recommended
level of US$5.40 per capita for
lovv - income develop ng
countries. See 1341 above.
p. 51.

13

I
,1

08-MAY-1998

13:58

44 171 242 9696

90X

P.03

GJH 17-;

Gynaecological Morbidity among Women in a Bombay Slum

Dr. Indumati Parikh
Dr. Vijaylaxmi Taskar

Dr. Neela Dharap
Dr. Veena Mulgaonkar

Streeihitakarini

Lokamanya Nagar Compound
Kakasaheb Gadgil Marg

Dadar, Mumbai 400 025

*

Gynaecological Morbidity among Women In A Bombay Slum

Dr. Indmnati Parikh
Dr. Vijayalaxmi Taskar

Dr. Nee la Dharap
Dr. Veena Mulgaonkar

Objective

There is a growing recognition that gynaecological morbidity is an important health problem among poor

women in India. Yet, information on the levels and patterns of gynaecological problems experienced by
women in India is sparse. There are few community-based studies; since a large proportion of women suffer
morbidity silently, and are reluctant to seek care or to visit clinics and hospitals, it is difficult to assess the
true magnitude of the problem or the patterns of morbidity from which women suffer. Yet the small amount

of data available suggests startlingly high levels of morbidity, for which treatment is rarely sought. One of
the very few community-based studies conducted in a rural area of Maharashtra s Gadchiroli district reports

that of 650 women aged 13 and above. 55% reported gynaecological complaints, but as many as 92% were

reported on clinical examination to have one or more gynaecological or sexually transmitted diseases. Yet
only 8% had sought treatment (Bang et al., 1990). Other studies from India and other developing

countries—Karnataka. Bangladesh, Egypt, and Nigeria—corroborate significant though lower levels of
reproductive morbidity among the general population (Younis et. al., 1993 for Egypt; Brabin et al., 1995 for
Nigeria; Bhatia et al., 1995 for Karnataka) or among specific sub-populations such as contraceptive users

(Wasserheit et al., 1989).
This picture of high levels of morbidity, combined with a reluctance to seek treatment is
corroborated b\ the working experiences of many non-governmental organizations addressing health needs
in both, rural and urban India. Yet more rigorous information about levels and patterns of morbidity, their

perceptions and correlates, is virtually non-existent. Community based studies of the prevalence and nature
of gynaecological morbidity in different settings that can provide information to health planners and policy

makers regarding appropriate strategies to improve women's reproductive health are much needed.

i

This study is intended to fill that gap in knowledge of gynaecological morbidity in a slum area of
Bombav served b\ Streehitakarini, a health-based voluntary organization in Bombay( 1). Streehitakarini is a
prominent non-goxernmental organization with a long history in this area, starting in 1964 with a medical

clinic set up in its headquarters in the area. The clinic has been catering to the health needs of poor women
and children and has succeeded in involving local women as grass-roots workers, in becoming a nucleus foi

activities focusing on women's welfare as well as establishing itself as a major health provider in the area.
Streehitakarini's intensive health education and child survival programmes led to a concern among women

first about their children's health, but increasingly about their own health needs as well. 1 he study hence

grew out of Streehitakarini's need to understand the full range of health needs of the women served. I he
objective of this study is thus to determine the levels, patterns and correlates of gynaecological morbidity in
an urban slum, focusing on both women's perceptions and assessment of their gynaecological health as well

the conclusions of medical assessments and laboratory tests.

Data and Methods
Data for this study w ere drawn from a survey of ever married women residing in one slum area of Bombay.

The surx ey comprised of several elements: (a) a socio-demographic survey of respondents including their

reported symptoms and morbidity and reproductive histories; (b) a clinical examination and (c) laboratory
tests. Other qualitative data were also obtained through group discussions with health workers, informal
inten iews with health practitioners and 100 community women on their perceptions of disease patterns

among women in the community(2).

The study was designed to draw a random sample of ten percent or (1500 of the estimated total of

15000) ever married women residing in this slum. No attempt was made to keep the sample size constant by

replacing respondents whose dwelling units had been demolished or who had out-migrated by the date of
the survey, resulting in a sample loss of 446 women. Of the remaining 1054, 298 refused a gynaecological

examination and hence vital information on symptoms and clinical findings is not available for them.

Hence, the effective sample comprises a total of 756 women, representing an overall sample loss of 50%;

and a refusal rate of 28%.

2

Field work for the survey was undertaken in 1989. Interviews were conducted by two extension

workers who were trained investigators(3). Interviews were conducted largely in the Marathi language, at
the homes of the respondents.
Respondents were then requested to attend the Streehitakarini clinic for subsequent medical

examination. While the examination was generally conducted at the clinic premises, if women were
reluctant to attend the clinic, examinations were occasionally conducted at convenient locations close to the

home of the respondent in well-equipped medical vans. Two experienced female gynaecologists conducted
all the examinations. Examinations comprised of (a) noting medical history; (b) general medical

examination; (c) speculum examination (including swabs taken from the cervix and posterior fornix) and a
bimanual examination; blood and urine samples were taken from all those who agreed. Samples were sent
to prominent Bombay hospitals for testing(4). Those observed to be in need of treatment were immediately

provided medical attention or referred to appropriate hospitals for further investigation and treatment.

Gynaecologists report that once they overcame their shyness regarding this subject, respondents were quite
willing to discuss their gynaecological histories and problems.

Given the fact that the sample has been drawn from a single, homogeneous slum area, there is
relatively narrow variation in such characteristics as income, education, sanitation or hygiene levels.

Socioeconomic and Demographic Profile

Respondents reside in a typical slum of Bombay. Homes either consist of huts (zopadis) constructed largely
of wooden planks, cane, bamboo and occasionally brick and tiles or of single-room flats in concrete

buildings called (chawls), each of which contains about 20 flats. The area is congested and has few

amenities. A profile of the community suggests that two in three families reside in homes measuring 100
square feet or less; these homes usually accommodate entire families (Table 1). The average household size

is 5.8 and as many as two in five households consist of more than 5 and less than 10 members. Water
facilities are erratic and insufficient for over 90% of the slum, and clothes and utensils are washed, beside

the open drains. Toilet facilities are poor (not shown): an average of one for 15-20 chawl residents and for
250 zopadi residents. Narrow pathways crisscross the slums and are lined with open drains and open
garbage. There is no open space and the air is highly polluted due to the smoke emitted by nearby factories.

3

The large majority of respondents (97%) are Hindu. 1 he mother tongue of 77% respondents is
Marathi; other mother tongues included Telegu (9%), Hindi (8%) and Gujarati (5%). Respondents are
uenerally poor and poorly educated. As many as 61% live below the poverty line; one in three (31%) is
illiterate and another 30% have had no more than a lower primary education. Few women work: 14 per cent
are employed for wages.

Among demographic characteristics, the average respondent was aged 30, had 2.6 surviving

children and was married at 17.1 years of age. I he mean age at first pregnancy was 19.5 years. As far as
contraceptive patterns are concerned, as in the general population, terminal and female methods were most
likely to be used: 43% were sterilized at the time of interview, 6% and 3% were using IUDs and oral

contraceptives respectively and 2% were using male methods (vasectomy or condoms).
Hygiene and sanitation conditions, as assessed by the investigator, are generally poor: sanitation

facilities are reported as extremely poor among half of all respondents and two-thirds are reported to have
poor personal hygiene.

Given the relatively high level of sample loss, it is reassuring to observe, that women who refused
to be examined have virtually identical socioeconomic and demographic characteristics as those who were

examined (Table 1). Living conditions are virtually identical, as are age and parity distributions; women

who agreed to be examined are slightly better educated, somewhat more likely to use non-terminal methoc|s

and less likely to use no contraception than those who refused examination. These kinds of results suggest
that the 756 women who constitute the sample of this study were randomly drawn, and the exclusion of the
others is not likely to affect results significantly.

Gynaecological Histories and Perceptions of Morbidity
Clinical histories included detailed information on menstrual and obstetric histories, as well as an enquiry
about the presence of any gynecological complaints described in commonly used and understood local
terms. Among gynaecological disorders, seven major conditions were probed, as follows:

1.

menstrual problems: dysmenorrhoea (painful menstruation) was specifically probed; all other

menstrual problems were discerned from menstrual histories: polymenorrhoea (frequent menstruation with
cycle length shorter than three weeks); menorrhagia (duration of bleeding more than five days or excessive

in amount as assessed by the clinician); oligomenorrhoea (duration of bleeding less than three days or cycle
length more than five weeks); and metrorrhagia (irregular or intermenstrual bleeding);
4

excessive vaginal discharge, as expressed by the woman in the vernacular;
2

lower abdominal pain; or pain in hypogastrium or iliac regions;

4.

low backache: or pain in the lumbo-sacral region;

5.

"something coming out" from vagina (as sign of genital prolapse).

6.

pain or burning sensation while passing urine (as sign of dysuria);

7.

infertility or difficulty in becoming pregnant.

Table 2 shows that almost three quarters (73%) of all women reported one or more of the above

gynaecological conditions. Almost one-third (30%) reported white discharge (pandharepani)(5), a typical
symptom of lower reproductive tract infection. Two in five (39%) reported lower back pain (kambar dukhi)
and one in five (21%) reported abdominal pain (otipot dukhcine). Almost two fifths (39%) report symptoms

suggestive to the physician of one or more menstrual disorders, ranging from 23% who report symptoms
suggestive of dysmenorrhoea (palitpot dukhate) to 13% who report oligomenorrhoea (scanty periods,
angavar kamijane) and 5%-7% reporting irregular periods (aniyamitpuli) or profuse bleeding (palit just

jane) (that is, polymenorrhoea, menorrhagia or metrorrhagia). Fewer women (3%-6%) report such
conditions as 'something coming out’ (anga baharyete) or infertility (moot na hone) or dysuria (laghavila
aag or garam laghavi).
Low backache and lower abdominal pain are frequently, but not necessarily, symptoms relating to

such gynaecological morbidities as pelvic inflammatory disease, cervicitis, vaginitis, cystocoele, rectocoele,

polyp and fibroids. Hence, they are included among reported gynaecological complaints (although some
studies have excluded both) (see, for example. Bang et al., 1989). As many as 39% and 21% of all

respondents report low backache and lower abdominal pain respectively.
Qualitative data from group discussions held with one hundred women also suggest the prevalence
of these conditions(6). In a free listing of illnesses which women commonly suffer from, gynaecological

conditions featured prominently. Among the 15 leading conditions listed, eight reflect gynaecological

conditions: white discharge (61), low back pain (54), low abdominal pain (44), severe menstrual bleeding

(29), painful menstruation (21), burning micturition (11), uterine cancer (25), lump in the breast (22).

Others included pain in the legs (36), general weakness (29), headache (22), pain in the joints (18) and
tuberculosis (18), giddiness (17) and abdominal lump (11), some of which may also be manifestations of

gynaecological conditions.

5

In short, the results of this section suggest high levels of reported gynaecological morbidity

including leucorrhoea, dysmenorrhoea and other menstrual complaints and low back pain.

Clinical Examination

The gynaecological examination included a per speculum examination followed by a bimanual examination
and the following five conditions were generally diagnosed:

I.

vaginitis: inflammation of the vagina (redness and/or ulcer), with or without discharge;

2

cervicitis: all diagnoses of acute cervicitis, endocervicitis and chronic cervicitis; with and without
erosion;

3.

pelvic inflammatory disease (PID): tender or palpable or thickened fornices; fixity of uterus and/or
tubo-ovarian tender masses or tender movements of uterus;

4.

prolapse, including cystocoele, rectocoele or uterine

5.

fibroid and polyp.

Table 3 points once again to high levels of gynaecological morbidity. On clinical examination, as
from gynaecological histories, the results suggest that 73% of all women suffer one or more gynaecological
morbidities. Leading conditions include cervicitis (39%), with and without erosion, prolapse (19%),

including cystocoele, rectocoele or uterine prolapse, and pelvic inflammatory disease (16%). Polyps and
fibroids were rarely observed; uterine cancer was not observed at all (Table 3).
In summary, a large proportion of women—almost three in four—were diagnosed as having one or

more gynaecological morbidities. Leading causes of morbidity include cervicitis, prolapse and pelvic
inflammatory' disease.

Laboratory Test Results:
Laboratory' tests of cervical and vaginal smears, and blood and urine samples were conducted on a random

sub-sample of 569 of the 756 women for whom histories and clinical data are also available. In addition,
tests for chlamydial vaginitis (chlamydia trachomatis) were independently conducted on a random sub­

sample of 399 of the 756 women for whom histories and clinical data, as well as other laboratory tests, are

6

also available. Although a variety of laboratory tests were conducted, Table 4 presents only such indicators

of gynaecological morbidity as sexually transmitted diseases (STDs), that is, trichomoniasis (trichomonas
vaginalis), gonorrhoea (gonococci), syphilis (positive VDR.L) and chlamydial vaginitis (chlamydia
trachomatis); and endogenous infections such as candidiasis (Candida albicans or yeast); clue cells; or

gardnerella vaginalis(7) (Table 4).
Table 4 presents rates of STDs and endogenous infections for both the complete sample for whom^

laboratory testing had been conducted and the randomly selected sub-sample of 399 for whom all laboratory
testing, including tests for chlamydia, were conducted. Results suggest that the two samples were indeed

randomly selected: patterns of STDs (excluding chlamydia) and endogenous infections are similar among

the larger group of 569 women for whom laboratory testing was conducted and the sub-sample of 399 for

whom chlamydia testing was also undertaken. In addition, it is clear from Table 1 that the socioeconomic
and demographic conditions of the 569 and 399 women for whom various laboratory tests were conducted

are virtually identical to those of women for whom this was not done.
A look at the rates in Table 4 for the 399 women for whom complete data are available suggests

high rates of laboratory-detected morbidity. As many as 49% of the 399 women for whom all laboratory
tests were conducted had one or more lab detected illness. Among specific diseases, the most commonly
observed is chlamydia, observed in as many as 14% of all women. Among other sexually transmitted
diseases, trichomoniasis was observed among 10% of all women (8); but other STDs, such as gonorrhoea

and syphilis (VDRL) were rarely observed—among 1.5% and 0.3% of all women, respectively. In total, the
prevalence of sexually transmitted diseases assessed from these tests is extremely high among this slum

population: 23%.
In addition to sexually transmitted diseases, high rates of endogenous infections are also commonly

observed. As many as 31% of the 399 women report one or more endogenous infections: candidiasis

(Candida albicans) is observed in 17%, and bacterial vaginosis (as measured by the presence of gardnerella
vaginalis and clue cells) is observed in 15% of the women.

Infections such as syphilis, gonorrhoea and chlamydia are difficult to diagnose in women in the
absence of microbiological or serological examination. Such infections as Candida, gardnerella vaginitis and

trichomonas vaginitis are also difficult to diagnose as a large number of women are asymptomatic. It is
hence difficult to diagnose and treat these infections, without incorporating periodic gynaecological check­
ups in general health services.
7

Such high rates of sexually transmitted diseases and endogenous infections among women in this

apparently low-risk urban slum community reinforce concerns about their partners' sexual behaviour and

male responsibility in general. Evidence has shown that men can transfer organisms from one woman to
another without developing any signs of the disease themselves. Yet, the experience is that husbands of

women suffering from sexually transmitted infections resist treatment, resulting in the risk of re-infection
among women.

Socioeconomic and Demographic Correlates of Gynaecological Morbidity

Table 5 presents selected socioeconomic and demographic correlates of the more widely prevalent reported,
clinically diagnosed, and laboratory diagnosed gynaecological conditions. By and large, associations
between socioeconomic indicators and morbidity, however measured, tend to be weak, reflecting perhaps
the overall homogeneity in environmental and socioeconomic conditions in this slum population. Among

social and economic correlates of morbidity, of some interest is the finding that respondents with income
levels above the poverty line are somewhat more likely to report a menstrual condition or any
gynaecological condition, and more likely to have a laboratory diagnosed sexually transmitted disease.

Working women, similarly, are somewhat more likely than non-working women to report a menstrual
complaint. Neither women's education levels nor their sanitary conditions, in contrast, are related to

morbidity, however measured.

More variation is seen in morbidity levels by age and parity. Older and high parity women are
systematically more likely to report low back or lower abdominal pain and menstrual problems (women
over 40 are less likely to report the latter). Not reported here because numbers are small, they are also more
likely to report 'something coming out', and dysuria. Clinically diagnosed and laboratory-detected morbidity

tends to be much more sensitive to demographic indicators, albeit in different ways. For example, parity is
positively associated, and marital age inversely associated with morbidity, and especially with reproductive

tract infections (cervicitis, vaginitis or pelvic inflammatory disease) and prolapse. Age is clearly and
positively related w ith prolapse. In contrast, the correlates of laboratory-detected morbidity suggest that

older women are somewhat less likely than younger women to experience either a sexually transmitted

disease, or any endogenous infection (particularly candidiasis). Also evident is a consistent inverse
relationship between infection and parity: childless women and particularly those with 1-2 children
8

experience higher rates of sexually transmitted diseases than higher parity women; childless women
experience considerably higher rates of endogenous infections, particularly candidiasis (Candida albicans),

than other women.
Current contraceptive status appears to have some bearing on morbidity. For example, women who

have had tubectomies or use IUDs report considerably higher rates of reported morbidity and particularly
menstrual complaints and low backache or lower abdominal pain, than other women. When clinically
diagnosed morbidity is considered, sterilized women are moderately more likely (78%), and oral

contraceptive and male method users somewhat less likely (66%) to experience at least one diagnosed

morbidity. In particular, prolapse is more evident among sterilized women (30%) than any other group,
suggestive perhaps of their relatively advanced age and greater likelihood of repeated pregnancy. Also,

reproductive tract infections (RTIs) are moderately more evident among sterilized women and those using

IUDs (55%-56%) than any other group. Finally, the relationship between contraceptive use and laboratorydetected morbidity suggests a different picture: here, oral pill and male method users are considerably less

likely than other groups to have experienced an STD; endogenous infections are particularly low among
these and sterilized women and correspondingly high among IUD users and non-users.

In order to examine the relative importance of the independent variables reported in Table 5 on
morbidity. Table 6 presents the calculated odds ratios from a series of logistic regressions in which each
measure of reported (Panel A), clinically-diagnosed (Panel B) and laboratory-diagnosed morbidity (Panel
C) is a function of such other indicators as age, parity, income level, education and current contraceptive

status. These models thus give us an idea of the strength of the relationship between each independent

variable and morbidity, net of the effects of other independent variables.
The results reported in Panel A of Table 6 generally suggest that complaints are significantly more

likely to be reported by sterilized women and IUD users than women using no method (the reference
category ). In particular they are significantly more likely to report menstrual disturbances; sterilized women

are additionally more likely to report low backache or abdominal pain. The only other relationships of
significance are the higher morbidity, notably menstrual problems, reported by women with incomes above
the poverty line compared to poorer women (the reference category); and the lower rates of low backache

and lower abdominal pain reported by educated women compared to the uneducated (the reference
category ). The results reported in Panel B of Table 6, relating socioeconomic and demographic factors to

clinically diagnosed morbidity, suggest a somewhat different pattern. Socioeconomic determinants drop out
9

as significant predictors, and age and parity become more important correlates of clinically diagnosed
morbidity. Among specific conditions, reproductive tract infections are unrelated to any independent

variable. Prolapse. ho\ve\er. is significantly and positively affected by age and parity; in addition, sterilized
women are significantly more likely than other women to have experienced prolapse, even after age and
parity are controlled. Any clinically diagnosed morbidity is significantly affected only by age and parity;

however, women with incomes above the poverty line, educated women and sterilized women report
moderately higher levels of clinically diagnosed morbidity compared to other women.
Panel C of Table 6 relates socioeconomic, demographic and contraception variables to laboratory
<r'

detected morbidity. Confirming the pattern observed in Table 5, we find that older women are significantly

less likely than younger women to experience an STD or any laboratory detected morbidity; and now,
sterilized women are significantly less likely than non-users and IUD users to experience any endogenous

infection. Other noteworthy findings include the absence of any relationship between socioeconomic factors
and morbidity ; and the suggestion that sterilized women and women using oral contraceptives and male

methods, by and large, are moderately /e.s.v likely to experience an STD than non-users.
This discussion of the socioeconomic and demographic correlates of morbidity suggests wide

differences in the net determinants of reported, clinically diagnosed and laboratory detected morbidity.

Reported morbidity is significantly influenced by socioeconomic factors and contraceptive status. In
contrast, significant determinants of clinically diagnosed morbidity include age and parity; now, while
contraceptive status is unrelated to RTls. sterilized women experience significantly higher levels of prolapse
than other women. Finally, there are few significant determinants of laboratory detected morbidity: age

confers a net protection, and contraceptive-users appear, in general, to be mildly but not significantly less
likely to experience STDs than non-users; the only significant influence of contraception now is the finding

that sterilized women are significantly less at risk of an endogenous infection than non-users and IUD users.

Health Seeking Behaviour
There is considerable ethnographic evidence of women's reluctance to seek health care, especially

gynaecological health care in India. In this study, information on women's health seeking behaviour comes
from group discussions with providers, that is, health workers, local medical practitioners and faith-healers
and pharmacists; as w ell as with a group of local women.

10

There was a consensus among workers that slum women were generally reluctant to seek care for

their gynaecological problems for several reasons. Above all, workers perceived a lack of understanding

about what constituted a gynaecological problem; many women perceived problems as a normal aspect of

womanhood. Second, even if motivated to seek help, prohibitive costs of treatment act as a second barrier.
In addition, women tend to be inhibited from discussing gynaecological problems with male physicians.

Interviews were also held with 22 local health practitioners practising various alternative medical
systems, of w hom only four were women. Corroborating the views of workers, not a single male
practitioner had e\er conducted a gynaecological examination, but were, nevertheless, approached by

women in order to discuss their gynaecological problems. Diagnosis was made and treatment conducted on
the basis of reported symptoms. Only in a few persistent cases were women then referred to public hospitals
or gynaecological clinics. Female practitioners corroborated a high incidence of leucorrhoea and infertility
among their slum women patients. Not unusually, the large majority of practitioners did not consider

gynaecological illnesses serious unless cancer was suspected.
Interview s w ith three male and two female faith-healers revealed extensive use of herbal medicines.

Among them, one male faith-healer treated gynaecological (and other) problems by his mystical healing
power and incantations (meditation and reciting mantras). By and large, the most common gynaecological

condition for which women visit faith healers is infertility.

Interviews w ith a few local pharmacists, and /^///-shopkeepers revealed that local women often

sought from them medication for menstrual pain or vaginal tablets for white discharge. Through this
approach, women were able not only to save the time and money necessary for visiting a doctor or clinic,
but also were spared the awkwardness of undergoing a medical examination.

Finally, discussions with women residing in the community suggest that health seeking for

gynaecological conditions is minimal and that home remedies are widely known, but less frequently
utilized. A variety of remedies were described for irregular periods—ranging from a paste of cress seeds,

water or milk to detes, jaggery, mutton, pigeon and those things that produce 'heaf in the body. In contrast,

remedies for excessive bleeding were quite the opposite: 'cooling' foods (boiled coriander seeds, banana,

yogurt and fresh fruits).
Women associated leucorrhoea with husband's promiscuity (navaryacha bahar khyalipana).
alcoholic husbands, as well as too many deliveries, lifting heavy objects, over-work, general weakness—as

well as having sexual relations during menstruation or within 10/12 days of delivery. Leucorrhoea is

11

perceived as a 'hot' condition and remedies include 'cooling' foods (rice, green vegetables). Other remedies
included hibiscus {jaswandiche) flowers and roots, various spices (cumin, coriander, basil (tu/si) and roots

(banyan, neem).

Discussion
The main limitation of this study is the high rate of sample loss. As indicated earlier, an initial sample loss
of 30% occurred as a result of closed dwellings and out-migration. The remaining 1054 ever married

women w ere successfully interviewed. However, 298 of them refused a clinical examination, resulting in a

further loss of 20% of the original sample and 28% of the interviewed sample. Of those who refused to

undergo clinical examination, two in five women refused for fear of the examination, blood tests and other
procedures or simply because they were too shy; and one in five refused because their husbands objected.

The extent to which sample loss may have biased overall prevalence rates and patterns is unclear. As many
as 40% of those who refused to participate in the study did so for fear of the examination or procedures or
because they were too shy; another 22% refused because their husbands objected; and one quarter refused
because they perceived themselves as healthy. This last finding suggests an element of self-selection that is

ine\ itable in a study of this nature, and could result in influencing community morbidity levels.
«■

Nevertheless, a comparison of socio-demographic characteristics of the 756 women for whom

clinical data are available and those for whom it is not suggests little variation, as seen in Table 1. Age and
paritx profiles are similar; however, women who refused examination tended to be less educated, although

no w orse off economically. Finally their contraceptive profiles are somewhat different. Prevalence rates are
slightly higher among women who underwent clinical examination (54%) than those who did not (48%);

among contraceptive-users, the women who accepted the clinical examination are somewhat more likely to
have used non-terminal methods and correspondingly less likely to be sterilized than women who refused.

Of the 756 women who obtained clinical examinations, laboratory tests were conducted for only

569 (75%) and comprehensive laboratory tests, including chlamydia, were conducted for only 399 (53%).

Again, a look al the socio-demographic and morbidity profiles, presented in I able 1, suggest there are no

major differences between those women for whom laboratory findings are available and those for whom it

is not. suggesting that women for whom all reported, clinjcal and all laboratory detected morbidity are
available are in fact randomly selected from the slum population under study.

12

Apart from sample loss, another limitation of this study is the considerable homogeneity of the
sample in terms of such socioeconomic characteristics as income, education, and such environmental

factors as sanitation and hygiene conditions. Moreover, there is not much variation in terms of contraceptive
method profile.

This stud) has highlighted several noteworthy findings on the state of reproductive health among
slum women in Bombay. The single most important conclusion of this study is the high prevalence of
gynaecological morbidity in this urban slum community. Over 70% of all respondents reported
gynaecological complaints; over 70% had clinical evidence of either vaginitis, cervicitis, prolapse or PID;

and almost half (49%) are observed to have either an STD or an endogenous infection as assessed by

laboratory tests.

Equally important and disturbing is the finding that almost a quarter of all women (23%), thus far
assumed to be at low risk, suffer from one or more sexually transmitted diseases. Evidence of such STD

infections as chlamydia and trichomoniasis was found in 15% and 10% of all cases, respectively; however,

syphilis and gonorrhoea were rarely observed.
A third important set of conclusions relate to the socioeconomic and demographic correlates of
morbidity. Results have suggested wide differences in the net determinants of reported, clinically diagnosed

and laboratory-detected morbidity. Reported morbidity is significantly influenced by socioeconomic factors
and contraceptive status, with sterilized women and those using IUDs reporting significantly higher levels
of morbidity than other women. In contrast, significant determinants of clinically diagnosed morbidity

include age and parity; now, while contraceptive status is unrelated to RTls, sterilized women experience

significantly higher levels of prolapse than other women. Finally, there are few significant determinants of
laboratory detected morbidity. Older women appear to have a reduced risk of STDs, and generally, of any
laboratory-detected morbidity. Also, contraceptive-users appear, in general, to be mildly but insignificantly
less likely to experience STDs than are non-contraceptive-users ; the only significant influence of

contraception is the finding that sterilized women are significantly less at risk of an endogenous infection

than non-users and IUD users. Further investigations into these relationships and their aetiology is

necessary, using larger samples and more in-depth probing.

13

Finally, results suggest that few women would resort to clinics or doctors for gynaecological
problems. Gynaecological conditions are rarely taken seriously by either women themselves or local care­
providers and are hence perceived as worthy of medical attention only if extreme. Instead, there is a wide

reliance on home remedies—herbs, pastes and roots—for the most common gynaecological conditions
including leucorrhoea and menstrual disorders.
These results stress the fact that gynaecological morbidity is unacceptably high and constitutes a
major public health problem, one that has remained largely unaddressed within current programmes.
Results underscore the need to broaden the scope of family welfare to incorporate among its reproductive

health services, the screening and treatment of STDs and other gynaecological infections. High rates of
sexually transmitted diseases and other infections highlight the urgent need to focus on responsible sexual

behaviour among men. In short, results present a forceful plea for greater attention to. and investment in, the

reproductive health needs of poor Indian women.

14

5

Notes:

1.

This study, conducted by Streeliitakarini was funded by the Ford Foundation, and was carried out in

a slum area of Bombay's Municipal Ward "G". By the mid-1980s, Streehitakarini was serving a population
of over 1,00,000 including over 15000 ever-married women.

2.

The survey comprised a socio-demographic questionnaire, a questionnaire filled by the medical

practitioner on both symptoms and results of clinical examination, and a laboratory findings report.
Questionnaires had been pretested extensively and were largely precoded, with fixed response categories.
The socio-demographic questionnaire contained questions on the age, marital status, birth histories, and

contraceptive use: household characteristics, including available amenities; and individual characteristics
such as economic activity status, education and husband's characteristics. The medical practitioner's
questionnaire contained questions on the reported menstrual history, gynaecological symptoms reported by

the respondent as well as details of the pelvic examination. The laboratory form contained information on
any abnormalities detected in lab tests.

3.

The two trained investigators (one college graduate and one under-graduate) were Streehitakarini

extension workers, using pre-tested questionnaires. Although investigators already had some rapport with
the community and although they explained at great length the need for this study, they faced some

skepticism from respondents, who considered gynaecological morbidity a natural part of womanhood.
4.

Hinduja and Tata Memorial Hospitals.

5.

Commonly used Marathi terms for leucorrhoea includepcmdhcirepani or sqfedpani literally, white

liquid. Other terms included pandhari dhupuni, kapade kharab hotat or kapdyala dag (soiling clothes), dhat
padate, angawar jute or angavar safedjate.

6.

Discussions with workers also confirm the predominance of leucorrhoea in this community, along

with low back and abdominal pain, cancer of the uterus, menstrual problems, and infertility.

7.

While testing was also conducted for iatrogenic infections observed on smears (staphylococcus

aureus, klebsiella pneumoneae, E. coli, pseudomonas aeruginosa, staphylococci and streptococci), some
samples are suspected to have been contaminated and hence these data are net presented. Tests for systemic

infections (tuberculosis etc.) were not conducted.

15

8.

Trichomonas vaginalis is not always venereal in origin; transfer of the organism can occur from one

individual to another by indirect contact. It has clearly been shown that a man can transfer the organisms
from one woman to another without himself developing any signs of the disease. This does not mean that

trichomonas vaginalis is always venereal in origin. Transfer of the organism from one individual to another
by indirect contact certainly happens (Jeff Coates. "Infections as they affect individual organs" in Jeff

Coates. Principles of Gynaecology, chapter 20, p. 3 18; fifth edition).

References

Bang. R.A.. A.T. Bang. M. Baitule, Y. Choudhary, S. Sarinikaddam, and O. Tale. 1989. High Prevalence of
Gynaecological Diseases in Rural Indian women. Lancet. 1: 85-87.

Bhatia J.C., J. Cleland. L. Bhagavan, et al. 1995. Prevalence of Gynecological Morbidity among Women in

South India. Studies in Family Planning. 26(4):203-216.
Brabin, L.. J. Kemp. O.K. Obunge, et al. 1995. Reproductive Tract Infections and Abortion among
Adolescent Girls in Rural Nigeria. Lancet. 345: 300-304.

Coates. Jeff. 19?? Principles of Gynaecology, chapter 20, p. 318; fifth edition).
Wasserheit. J.N., J.R. Harris, J. Chakraborty, B.A. Kay, K. Mason. 1989. Reproductive Tract Infections in a

Family Planning Population in Rural Bangladesh. Studies in Family.Planning. 20: 69-80.
Younis, N., H. Khattab, H. Zurayk. et al. 1993. A Community Study of Gynaecological and Related
Morbidities in Rural Egypt. Studies in Family Planning,. 24:175-86.

16

Table 1
Socio-demographic profile of respondents

Women who
were examined

Women
who refused
examination

Women who were examined and
for whom laboratory tests were
conducted (1)
all lab tests inclany lab tests
uding chlamydia

Women
examined
for
Chlamydia

404

TOTAL

756

298

569

399

Age
15-24
25-29
30-39
40-59

26.3

23.9
27.8
35.9

22.3

30.1
36.6

30.0

12.7

25.8
27.9
34.6
11.7

22.3

27.8

12.5

11.0

% iiterat*

69.0

58.4

69.9

68.4

68.3

Mun year* of education

4.5

3.5

4.6

4.5

4.5

% women working

13.6

14.0

10.7

10.3

10.1

5.

Mean age at marriage

17.1

16.3

0.2

17.3

17.3

6.

Mean ago at first pregnancy

19.5

19.0

19.6

19.7

19.7

Mean no. of surviving children

2.6

2.6

2.7

2.7

2.7

43.0
5.6
3.4
2.4

44.6

46.9
5.4

47.9

% IUD
% oral pills
% male methods

47.6
5.5

% non users

9.

Religion: % Hindu

10.

Mother tongue
Marathi

1.

3.

8.

33.2

47.8

current contraceptive status
% tubectomy

3.2

5.3
3.8

45.6

1.3
52.3

2.3
42.2

2.0
41.1

41.2

96.7

98.0

97.0

97.2

97.3

77.0
4.6

65.1

77.9
4.4

74.9
5.5
9.3

8.6
2.1

17.4

7.9
2.1

8.8

3.4

1.5

75.2
5.4
9.2
8.7
1.5

100 sq. ft. or less

65.4

69.5

65.2

101-200 sq.ft.
200 + sq.ft.

31.2
3.3

28.5
2.0

31.3
3.5

64.4
32.8
2.8

64.6
32.7
2.7

Below poverty line < Rs. 201

60.6

56.4

61.5

61.7

61.9

Rs. 201 to 450
Rs. 451 +

34.4

35.9

33.7

33.8

33.4

5.0

7.7

4.5

4.7

13.

water supply: % inadequate

91.8

96.7

91.9

93.2

93.1

14.

Sanitation: % poor

50.9

57.7

52.9

55.9

55.9

69.1

Gujarati
Hindi
Telegu
Others

11.

12.

15.
|l6.

1.3
0.3

5.7
8.4

Living space

3.7
2.0

Per Capita Income (per month)

Hygiene: % poor

Mean household size

67.9
5.8

73.8

67.8

69.2

5.3

5.9

5.8

(1 > 569 women were tested for STDs and endogenous infections; only 399 of these were additionally

5.8

tested for chlamydia.

30 +

45.9

46.3

48.3

Hindu

97.0
1.2
1.4
0.3

96.7

98.0

Nav Budha
Muslim

1.5

Christian

0.3

1.3
0.3
0.3

Sufficient
insufficient

8.2

3.4

91.8

96.6

1.6

8.0
92.0

47.6

47.8

Table 2
Reported gynaecological complaints

TOTAL No. EXAMINED WOMEN

% women
reporting
the
following
symptoms
756

1.

any menstrual complaint
dysmenorrhoea
polymenorrhoea
menorrhagia
oligomenorrhoea
metrorrhagia

39.3
23.3
4.9
7.0
13.2
7.3

2.

Excessive white discharge

30.2

3.

Low abdominal pain

21.4

4.

Lower back pain

38.9

5.

Something coming out

2.8

6.

Pain, burning during urination

5.7

7.

Infertility

4.9

8.

Any morbidity

73.0

Table 3
Clinically diagnosed gynaecological morbidity

TOTAL No. EXAMINED WOMEN

% women
diagnosed
to have:
756

1.

Vaginitis

14.9

2.

Erosion

20.8

3.

Cervicitis
Cervicitis (but no erosion)
Cervicitis with erosion

39.4
21.3
18.1

4.

Pelvic inflammatory disease

16.1

5.

Prolapse (any)
Cystocoele (only)
Rectocoele (only)
Cysto- and Rectocoele
Uterine prolapse

18.9
5.7
3.7
7.8
1.7

6.

Polyp

1.3

7.

Fibroid

0.5

0.0

9>.
Any morbidity

72.8

Table 4
Gynaecological morbidity as observed in lab findings

TOTAL No. EXAMINED WOMEN

% women
with lab
tests
excluding
chlamydia
diagnosed
as having:
569

% women
with all
lab tests
including
chlamydia
diagnosed
as having:
399

1.

Any abnormal finding (excluding chlamydia)
Any abnormal finding (including chlamydia)

36.4
na

40.6
48.9

2.

STDs (excluding chlamydia)
STDs (including chlamydia)
trichomonas vaginalis
gonorrhoea
chlamydia
Syphilis/VDRL +

8.8
na
7.7
1.1
na
0.2

11.3

Endogenous infection
Bacterial vaginosis (gardnerella or clue cells)
Gardnerella vaginalis
Clue cells
Candidiasis albicans

28.8

30.8

13.4

14.5
13.5
8.3
17.3

3.

NB:

12.7
8.3
16.5

Iatrogenic infections have been excluded from this analysis
because of inconsistent findings and possible contamination of
some samples; systemic infections were not assessed.

22.8
9.8
1.5
14.5
0.3

Table 5
Socio-demographic correlates of reported, clinically diagnosed and laboratory-detected morbidity: percentage distributions

LABORATORY-DIAGNOSED

CLINICALLY DIAGNOSED

REPORTED MORBIDITY (a)

MORBIDITY (a)

1.

2.

5.

back or
abdom­

one or
more

reported
morb­
idity^

rual
comp­
laint

inal
pain

clinically
diagnosed
morbidity

Any
RTI
(vaginitis
cervicitis
PID)

(cysto-,
rectocoele,
uterine

MORBIDITY (b)
Any
STD

one or
more

laboratory
detected
infection

Any
endo­
genous
infection

39.3

47.4

72.8

52.1

18.9

48.9

22.8

30.3

Per capita income
below poverty line (upto Rs. 200)
above poverty line (Rs. 200 + )

70.1
77.5

35.8
44.6

47.6

71.8
74.2

53.1
50.7

19.4

47.2
51.6

19.5

32.5

18.1

28.1

28.1

53.9
43.5

71.4

53.4
54.4

20.1
22.6

46.0
53.0

20.6

29.4

76.5

32.2

45.2

70.9

49.3

15.1

48.1

26.1
22.2

47.0

Education

73.9

38.0

Middle/ +

70.0
74.7

36.5
42.5

Work status
Not working
working

72.0
80.0

37.8
49.0

47.7
45.0

72.7
73.0

51.4

18.6

21.0

48.8
49.1

22.2
26.3

31.3

57.0

Sanitary condition
Poor
Fair

73.5
72.5

39.5
39.1

47.5
47.2

74.8
70.6

52.7
51.5

18.4
19.4

47.1
51.1

23.8
21.6

29.6
32.4

Age
15-24

7.0.4

35.2

39.2

71.4
74.3
74.9
66.7

53.8
49.0

54.2
42.5

24.2

78.1
66.7

47.6
52.6
66.7

6.0
17.2
24.7

27.0

39.1
45.8
31.3

51.3
52.4

56.2

72.4

34.4

40.9

19.9
20.5

34.8
35.8
26.7

71.1
75.1
73.9

36.9
39.9
. 41.9

50.5
44.6
46.1

77.7
73.7
66.0

55.2
54.5
46.5

23.9
16.6

44.8
49.1
53.4

24.0
24.1
20.3

26.0
30.4
36.8

86.9
67.9
77.8
71.5

44.3
33.1
43.8
43.0

36.1
43.2
50.7

45.9
72.6
75.4

53.9

78.8

36.1
52.7
53.2
54.9

3.3
1 1.8
23.2
30.6

69.0
48.6
47.9
43.9

24.1
27.5
19.3
18.7

48.3
29.6
31.1
28.0

76.3

47.7

IUD
other methods (c)

83.3
61.4

53.2
50.0

78.2
71.4

55.7
54.8

29.5

52.4

43.5
57.1

65.9
68.7

47.7
49.0

34.8

70.1

38.6
42.6

42.9
26.1

none

38.6
29.9

21.5
28.6
8.7

56.1

25.6

37.8

Age at marriage
14-16
17-18
19 +
Surviving children

0
3

8.

prolapse

73.0

25-29
30-39
40-59

6.

any
menst­

TOTAL

None
Primary

3.

one or
more

Current contraception
tubectomy

Notes
refers to 756 women for whom both socio-demographic and clinical data are available
a
refers to 569 women for whom laboratory tests were also conducted
b
includes oral contraceptives, vasectomy, condoms.
c

14.6

7.2
18.2

10.4

31.0

28.1

22.7

24.1

Table 6
Socioeconomic, demographic and contraceptive use correlates of gynaecological morbidity: logistic regressions

Any
reported
morbidity

1.

2.

3.

REPORTED MORBIDITY
Any
Low backmenstrual
ache/lower
problem
abdominal
pain

CLINICALLY DIAGNOSED MORBIDITY
Any
Any RTI
Genital
diagnosed
(vaginitis,
prolapse
morbidity
cervicitis or
PID)

LAB DIAGNOSED MORBIDITY (1)
Any
Any
Any
laboratory
STDs
endogenous
diagnosed
diseases
morbidity

SOCIOECONOMIC FACTORS

1.5857

1.0861

1.3165

0.9704

1.2882

1.1202

1.0396

0.7152

0.7730
0.9905

0.8689
1.1700

0.6795
0.7645

1.2958
0.9960

1.0326
0.8667

1.4549
1.0571

1.1731
0.9240

1.1805
0.9547

1.0405
0.9895

Age

0.9970

0.9905

0.9997

0.9616 **’

0.9839

1.0475

0.9719

0.9714

0.9757

Children ever born

0.9225

0.9946

1.0776

1.3704

’*•

1.1009

1.2513

1.0160

0.9919

1.0722

1.3767 *
1.4067
0.8716

1.2208
0.8648
0.8068

1.1919
1.1429
0.9607

2.2338
0.6959
1.6471

0.6926
1.0287
0.4809

0.6984
0.9745
0.4628

0.4901
1.0602
0.5192

Income above poverty line

1.4993

Primary education
Middle or higher education

**

DEMOGRAPHIC FACTORS

CONTRACEPTIVE STATUS

Tubectomy
IUD
Other methods (2)

1.6827
2.4403
0.7912

•*
**

2.4854
2.7864
1.5957

’•*
•••

NB: Reference categories: for education, uneducated women; for income: low income (below poverty line);
for contraception: non-users.
* significant at the .10 level.
** significant at the .05 level.
*** significant at the .01 level.
(1) N = 399 women for whom all laboratory tests including chlamydia had been conducted
(2) OC and male methods

Table 6 A
Socioeconomic, demographic and contraceptive use correlates of gynaecological morbidity: logistic regressions

MORBIDITY

1.

2.

3.

C. LABORATORY

B. CLINICALLY

A. SELF-REPORTED

DIAGNOSED

DIAGNOSED

MORBIDITY

MORBIDITYd)________
Any

Any

Low back-

Any RTI

Genital

Any

menstrual

ache/lower

(vaginitis,

prolapse

STDs

problem

abdominal

cervicitis or

pain

PID)

endogenous

diseases

SOCIOECONOMIC FACTORS
Income above poverty line

1.59

1.09

0.97

1.29

1.04

0.72

Primary education

0.87

1.45

1.17

0.68
0.76

1.03

Middle or higher education

0.87

1.06

1.18
0.95

0.99

Age

0.99

1.00

0.98

1.05

0.97 *

0.98

Children ever born

0.99

1.08

1.10

1.25

0.99

1.07

1.38

1.19

1.41

1.14

2.23
0.70

Other methods (2)

1.60

0.87

0.96

1.65

0.70
0.97
0.46

0.49

IUD

2.49
2.79

1.04

DEMOGRAPHIC FACTORS

CONTRACEPTIVE STATUS

Tubectomy

NB: Reference categories: for education, uneducated women; for income: low income (below poverty line);
for contraception: non-users.

* significant at the .10 level.

** significant at the .05 level.
*** significant at the .01 level.

(1) N = 399 women for whom all laboratory tests including chlamydia had been conducted
(2) OC and male methods

1.06

0.52

X- *

Table 6 b
Socioeconomic, demographic and contraceptive use correlates of gynaecological morbidity: logistic regressions

A. SELF REPORTED
MORBIDITY
Any
menstrual
problem

Low back­
ache/lower
abdominal
pain

B. CLINICALLY
DIAGNOSED
______ MORBIDITY
Any RTI
Genital
(vaginitis,
prolapse
cervicitis or
PIP)

C. LABORATORY
DIAGNOSED
MORBIDITY! 1 >_______
Any
Any
STDs
endogenous
diseases

new chlam
Leucorrhoea

Any
reported
morbidity

Any
diagnosed
morbidity

Any
laboratory
diagnosed
morbidity

SOCIOECONOMIC FACTORS

Income above poverty line

1 59

1.09

0.97

1.29

1.04

0.72

1.513

Primary education
Middle or higher education

0.87
1 17

0.68
0.76

1.03
0.87

1.45
1.06

1.18
0.95

1.04

0.99

Age

0.99

1.00

0.98

1.05

0.97

Children ever born

0.99

1.08

1.10

1.25

0.99

1.38

1.19
1.14

0.87

0.96

2.23
0.70
1 65

•*

1.2599

14993

1.3165

1.1202

1.3143
1.1692

0.9182
0.8859

0.7730
0.9905

1 2958
0.9960

1.1731
0 9240

0.98

0.965

0.9855

0.9970

0.9616

0.9719

1.07

1 0829

1.135

0.9225

1.3704

1.0160

0.70
0.97
0.46

0.49
1.06
0.52

0.9859
1 1518
0 2559

1.1212
2.8097
0.5881

1.6827
2.4403
0.7912

1.2208
0.8648
0.8068

0.6926
1 0237
0.4809

BBBBBBB

BBBBBBB

330 2059

920.131

881 7304

550 0500

•*

DEMOGRAPHIC FACTORS

CONTRACEPTIVE STATUS

Tubectomy
IUD
Other methods (2)

2 49
2 79
1.60

•••
•••

NB: Reference categories: for education, uneducated women; for income: low income (below poverty line!;
for contraception: non-users.
* significant at the 10 level.
* * significant at the 05 level.
* • • significant at the 01 level.
(1) N 399 women for whom all laboratory tests including chlamydia had been conducted
(2) OC and male methods

LOGLIKELIHOOD

atimtn

ng» a nn a

n u a a g a it

Table 6 G
Socioeconomic, demographic and contraceptive use correlates of gynaecological morbidity: logistic regressions

Any

REPORTED MORBIDITY
Any menst
leuc

back/
abd

CLINICALLY DIAGNOSED MORBIDITY
Any
RTI
Prolapse

Any

LAB DIAGNOSED MORBIDITY
Any
Any
STD
endog

new chlam

Age
Children ever born

0.9967
0.9219

0.99
0.9884

0.9852
1.1262

0.9996
1.0784

0.9614
1.35 ***

0.9839
1.0947

1.0472
1 2331

0.9719 *
1.016

0.981
0.9383

0 9769
1.0854

0.9639
1.127

Socioeconomic status
Inc >pov line
Primary education
Middle or higher education

1.1115 *
0.7825
0.9942

1.1006 *
0.8837
1.1836

1.0325
0 9279
0.8972

1.0245
0.6809 **
0.7644

1.0227
1.3127
1.0113

0.9722
1.0345
0.8721

1.015
1.4839
1 0848

1.1202
1.1731
0.924

1.183 **
1.2268
0 9751

0 9777
1.0245
0 9544

1.2591
1.3165
1.1402

Contraceptive status
T ubectomy
IUD
Other methods @

1.6605
2.3951
0.7865

2.4534 ***
2.6994 ***
1.5789

1.119
2.7569
0.5844

1.3732 *
1.4022
0.8708

1.2215
0.8455
0.8025

1 1968
1.1399
0.9591

2 2307
0.6823
1.6245

0.6926
1.0287
0.4809

1.0505
1.3013
0 3594

0 4985
1.0954
0 5464

0.9676
1.2515
0.2613

-LOGLIKELIHOOD

-877.77

-881.73

-1042.6

732 0487

550.05

422.5267

491.45

330.2059

1010.052

-920.131

1041.859

NB: Reference categories: for education, uneducated women; for income: low income;
for contraception non-users.
* significant at the 10 level.
** significant at the .05 level
*** significant at the 01 level
@ OC and male methods

TabMC 4
Sodoecorranac. demoprspdc md contraceptive use correlates o< gynaecological morbidity: logistic regressions

I Any
reports I

morbid :y

REPORTED MORBIDITY
low backAny
ache,Iowar
menstrual
abdomino
problem

Any

LAB DIAGNOSED MORBIDITY (II
Any
Any
endoy*noui
STDs
di>*as*i

CLINICALLY DIAGNOSED MORBIDITY
Genital
Any RTI

(vaginitis,
cervicitis or

di*gnos*d
mwbtdily

Any
l»boFatofy
diagno>*d

prolapsa

new chlani
Uhicorrhoea

morbidity

PIO I
SOCIOECONOMIC FACTORS

Incom* abov* povwlv bn*

Primary education
Midde or hrgtver educatran

1 4993

1 585 7

0 9704

1.3! 65

I 0861

1.2882

I 1202

1.2599

1.0396

0.7152

1.3143
1.1692

0.9182
0.8859

1.0326
0 866 7

1.4549
1.0571

0 9240

1.1805
0 954 7

1.0405
0.9895

0 98'19

1 04/5

0 9/19

0.9714

0.9/57

0 965

0.9855

1.3 704

1.1009

1 251'1

1 0160

0.9919

1.0722

1 0829

1.135

1.2208
0.8648
0,8068

1 1919
1 1429
0.9607

2.2338
0 6959
1.64 71

0.6926

0.6984

1.0287

0.9745
0.4628

0.4901
1.0602
0.5192

0.9859
1.1518
0.2559

C 7710
0.99 )5

0 8639
1 170G

0.6795
0 7645

1.2958
0.9960

C.99'0

0 9905

0.9997

0 9616

0 92/5

0.9946

1.0726

1.6827 ••
2 44 )3 ’•
0 79'2____

2.4854
2 7864
1 5967

1 3767
1 4067
0 8716

1.1/31

DEMOGRAPHIC FACTORS

CMdran *<r*r tram

3.

•••

••

CONTRACEPTIVE STATUS

Tubaci omy
IUD
Other mattrads 121



0 4809

1.1212
2 8097 •••
0.5881

M: R«lw*nc* cat*g<>r>*a: lor *ducatran. un«ducat*d woman; for mcoma: low incoma Ibalow poverty linal;

for contracaptran: rran-uiar>.
• significant at th* .10 lavai.
•• sagnrficant at th* 05 lavai.
••• significant ar th* .01 lavai.
(1) N-399 woman tu whom aS laboratory lasts mduding chlamydia had b**n conducted

(2) OC and male methods

-LOGLIXEtlHOOD

nifiti

imm

lllllll

lllllll

330.2059

-920.131

UJH \2_; 3

Prevalence of clinically detectable gynaecological
morbidity in India: Results of four community
based studies
Dr. K. Latha, Dr. S.J. Kanani, Dr. N. Maitra, Dr. R.V. Bhatt (BCC);
Dr. S.K. Senapati, Dr. S. Bhattacharya (CINI); Dr. S. Sridhar, Dr. G.B. Giri,
Dr P.P. Shah, Dr. S.P. Shah, Dr. L.A. Desai (SEWA-Rural); Dr. I. Parikh,
Dr. V. Taskar, Dr. N. Dharap and Dr. V. Mulgaonkar (Streehitakarini)

Introduction
In recent years, there has been increased
recognition of the scope and significance of
gynaecological problems experienced by
poor women in developing countries. The
first and perhaps the most compelling evi­
dence on the importance of gynaecological
morbidity came from a community-based
study undertaken in rural Maharashtra, In­
dia in the mid-1980s.1 Subsequent empirical
studies from Bangladesh, Egypt, Nigeria and
Karnataka have all documented significant
though lower levels of reproductive morbidity among the general population2*3'4 or
among specific sub-populations such as con­
traceptive users.5
In the Indian study1 conducted in a rural
area in Maharashtra state, 92 per cent of the

650 women clinically examined had evidence of one or more gynaecological dis­
eases, with an average of 3.5 conditions per
woman. The findings of this study were
striking, but raised questions about their
broader generalisability, given the small and
possibly atypical nature of the population
studied, and the size and geographical and
cultural diversity of India as a whole. Find­
ings from the four community-based stud­
ies reported in this paper, conducted in geo­
graphically and culturally distinct areas of
India, provide important additional evi­
dence on the prevalence of gynaecological
morbidity among poor women. Two stud­
ies were conducted in urban slum areas in
Bombay, and in Baroda in the state of
Gujarat. A third was conducted in rural West

Dr. K. Latha, Dr. S.J. Kanani, Dr. N. Maitra and Dr. R.V. Bhatt are with the Baroda Citizens Council (BCC),
Above Health Museum, Sayaji Baug, Baroda 390018, Gujarat, India; Dr. S.K. Senapati and Dr. S. Bhattacharya
are with the Child-in-Need Institute (CINI), Post Box No. 16742, Calcutta 700027, India; Dr. S. Sridhar,
Dr. C.B. Giri, Dr. P.P. Shah, Dr. S.P. Shah and Dr. L.A. Desai are with the Society for Education, Welfare and
Action (SEWA)- Rural, Jhagadia 393110, District Bharuch, Gujarat, India; and Dr. Indumati Parikh, Dr. V.
Taskar, Dr. N. Dharap and Dr. V. Mulgaonkar are with Streehitakarini, Lokmanya Nagar Compound, Kakasaheb
Gadgil Marg, Dadar, Bombay 400025, India.

8

The Journal of Family Welfare

Gynaecologic,]I morbidity

Bengal and the fourth among a rural popu­
lation in southern Gujarat. The studies,
which were conducted between 1988-91,
grew out of the need of four non-govern­
mental organisations providing health ser­
vices to understand the health needs of the
women in their project sites.
All four studies were population-based
and collected comparable data on aspects of
gynaecological morbidity. The studies were,
however, undertaken separately and inde­
pendently, and varied somewhat in terms of
study design, data collection procedures,
and the range of other information obtained.
In this paper, we present findings on aspects
of gynaecological morbidity common to all
four studies.
Subjects and Methods

The major objective in all studies was to
estimate the prevalence of gynaecological
morbidity among women in poor commu­
nities as assessed by the women themselves
as well as by a gynaecologist on clinical ex­
amination. As such, all studies contain clini­
cal histories provided by the respondent as
well as a clinical assessment of all respon­
dents who agreed to undergo a pelvic ex­
amination. For the sake of convenience, the
four studies are subsequently referred to by
their locations; rural West Bengal, rural
Gujarat, Baroda and Bombay. Table 1
summarises some of the important features
of each study and points out differences in
design.
Study Areas and Samples

Given the reluctance of many women to
undergo pelvic examinations, data collection
strategies in each study were tailored to lo­
cal conditions in order to enhance partici­
pation. In three of the studies, interviews
were carried out at the homes of the respon­
dents by trained interviewers. All inter­
viewed women were encouraged to visit the
organisation's health facility shortly there-

Vol. 43, No. 4, December 1997

after for a clinical examination. In the rural
Gujarat study, in contrast, both interviews
and clinical examinations were carried out
as part of health fairs organised in each vil­
lage.
Eligibility criteria for inclusion’varied
slightly among studies, as seen in Table 1. In
the West Bengal study, both ever married
and vimarried women were included in the
study; the other three studies included only
ever married women. Age criteria also var­
ied, with the lower age cut-off ranging from
13 years in West Bengal to 18 years in Baroda,
and the upper age limit extending beyond
the reproductive ages in both the rural
Gujarat and the Bombay slums studies. To
achieve comparability, the study popula­
tions in the present paper are restricted to
ever married women aged 15 to 45 years.
Significant proportions of interviewed
women refused to undergo a subsequent
clinical examination, ranging from 29 per
cent and 35 per cent in the two urban stud­
ies in Bombay and Baroda respectively, to
55 per cent and 65 per cent respectively in
the rural West Bengal and the rural Gujarat
studies.
Investigation

Common to each study was a socio-demographic profile and a morbidity history
provided by the respondent and a clinical
examination
conducted
by
the
gynaecologist. Clinical histories included
detailed information on menstrual and ob­
stetric histories, and an enquiry about any
gynaecological complaints, described in
commonly used and understood local terms.
Most such complaints as vaginal discharge,
urinary complaints, backache and lower
abdominal pain were specifically probed in
all studies except in tl 1 Baroda study, where
only volunteered information was recorded.
The clinical examinations, which in all stud­
ies were conducted by teams of women
gynaecologists, included a per speculum

9

o

TABLE 1
Settings and study populations
Study site
and period

Study population characteristics

Age (years)

Marital status

Sample size
(all women)

Residence

Survey

Clinical
examin­
ation

Survey

Clinical
examin­
ation

% of surveyed
women success­
fully examined

Urban
slum

840

548

840

548

65.2

Ever married
+ Single

Rural

1130

500

875

395

45.1

Ever married

Rural

1103*

324

835

293

35.1

Ever married

Urban
slum

1054**

756

1001

715

71.6

J

i

BCC

18-45

(1990-91)

CINI
(1990-91)

13-45

SEWA-Rural

15 +

(1988-89)
Streeh itakar i ni
(1989-93)
Purveyed230

15-50

W°men

Sample for current study
(Ever married women aged 15-45 yrs)

Ever married

10 S,Udy Vi"a8eS 1103 Ch°Se ,0 Par,iciPate in vi"a8e health

conducted by the organisation, and were

tally selected from ari earlier household listing, household visits subsequently failed to locate the other 446 women,
earlier household listing, household visits subsequently failed to locate the other 446 women,
largely aue to outmigration or demolition of original dwelling unit.
hi'geiy^^to^ut^

£
|
Oj

3

2_

r

II

Gynaecological morbidity

examination followed by a bimanual exami­
nation. Although laboratory tests were also
conducted in three of the studies, they var­
ied substantially in terms of the range of tests
conducted and the completeness of cover­
age, and have hence been excluded from the
present paper.

metrorrhagia: irregular or intermenstrual bleeding.
2. Excessive vaginal discharge, as ex­
pressed by the woman in the vernacu­
lar;
3. Low backache or lumbosacral pain;
4. Lower abdominal pain; i.e. pain in hypogastrium or either iliac region;
5. Dysuria: Pain or burning sensation
while passing urine; and
6. "Something coming out” from vagina (as
a possible sign of genital prolapse).

Gynaecological
conditions
were
standardised across all four studies as fol­
lows

A.
1.

Clinical history
Menstrual problems:
dysmenorrhoea (painful menses) was
specifically probed; all other menstrual
problems were discerned from a de­
tailed menstrual history as follows:

B.
1.

2.

Clinical examination
Vaginitis: inflammation of the vagina,
with or without visible discharge;
Cervicitis: all diagnoses of acute cervi­
citis, endocervicitis and chronic cervici­
tis;
Pelvic inflammatory disease (PID): ten­
der or palpable or thickened fornices;

polymenorrhoea frequent menses with
cycle length shorter than three weeks;

3.

menorrhagia duration of bleeding more
y s or excessive in amount
as assessed by the clinician;

The identification of other conditions fol­
lowed standard clinical definitions.

oligomcnorrhoea duration of bleeding
less than three days or cycle length more
than five weeks, or scanty in amount as
assessed by the clinician; and

Results
The socioeconomic and demographic profile of the women in each sample is
summarised in Table 2.

TABLE 2

Sociodemographic profile of respondents

RURAL

West Bengal
(N-395)

URBAN
Gujarat
(N - 293)

Baroda
(N - 548)

Bombay
(N - 715)

Religion
% Hindu
% Muslim

84
16

100

62
38

97
3

Education status
% literate

52

50

59

71

Employment status
% working for wages

6

59

17

13

Mean age (years)

28

29

30

29

Mean parity

3.2

2.7

2.6

2.6

Vol. 43, No. 4, December 1997

11

Gynaecological morbidity

Demographic differences between the
four studies were relatively moderate. For
example, the mean age of the sample varied
from 28 years to 31 years and mean parity
from 2.6 to 3.2. In contrast/socioeconomic
differences were wider. Significant propor­
tions of women in all four studies were un­
educated, ranging from 29 per cent in the
Bombay study to 50 per cent in the rural
Gujarat study. Relatively few women work
for wages (between six per cent and 17 per
cent) in three of the four studies; but in ru­
ral Gujarat, as many as 60 per cent work for
wages, largely as agricultural labourers. Fi­
nally, with the exception of the Baroda study,
populations were overwhelmingly Hindu;
about half of the sample in the rural Gujarat
study was tribal.

Table 3 presents gynaecological morbid­
ity as reported by the women prior to the
clinical examination.

The results show that a large majority of
the respondents in each site — ranging from
65 per cent to 84 per cent - reported one or
more gynaecological morbidities. Among
women reporting gynaecological morbidity,
the mean number of reported conditions
ranged from 2.0 to 2.6 in the four studies.
Substantial variation was evident, however,
in the specific patterns of reported morbid­
ity. The leading causes of morbidity in each
study were menstrual problems (33 to 59 per
cent of respondents), excessive discharge (22
to 57 per cent) followed by low backache (5
to 39 per cent).

Among menstrual disorders, dysmenor­
rhoea and oligomenorrhoea were the most
commonly reported problems. While com­
mon in all four studies, excessive discharge
was a particularly prominent condition in
the two rural studies in West Bengal and
Gujarat (57 per cent and 50 per cent respec-

TABLE 3
Gynaecological morbidity by clinical history
(% women reporting morbidity)

Gynaecological
condition

URBAN

RURAL
West Bengal
(N-395)

Gujarat
(N — 293)

Baroda
(N-548)

Bombay
(N - 715)

A. Menstrual problems
Dysmenorrhoea
Polymenorrhoea
Menorrhagia
Oiligomenorrhoea
Metrorrhagia

32.7
11.4
4.8
3.3
18.0
8.6

58.9
47.4
7.2
14.7
23.9
8.2

58.0
35.0
2.9
12.8
28.3
NR

40.7
24.3
4.9
7.3
13.4
7.4

B. Excessive discharge

50.1

57.0

22.4

30.8

C. Childlessness

3.3

2.7

1.8

5.0

D. Something coming out per vaginum

6.1

2.4

1.1

2.7

E. Lower abdominal pain

17.5

NR

9.3

21.5

F. Low backache

5.3

29.7

24.1

39.3

G. Dysuria

2.3

25.9

2.7

5.6

Women reporting any morbidity

65.3

84.3

64.6

74.1

Mean number of morbidities among women
reporting any morbidity

2.00

2.61

1.96

2.19

NR: Not recorded.

12

The* journal of Family Welfare

Gynaecological morbidity

tively). In three of the four studies low back­
ache was common (24 to 39 per cent); only
in rural West Bengal was it rarely reported.
Lower abdominal pain was reported by nine
per cent to 22 per cent of women in the three
studies which collected information on this
condition. Childlessness, dysuria and symp­
toms indicative of genital prolapse ("some­
thing coming out") were generally reported
by fewer women in each site, although some
variation was still evident.
Table 4 presents data on gynaecological
morbidity obtained through clinical exami­
nation.

Most notable is the considerable variation
in levels of gynaecological morbidity across
the four sites — ranging from 26 per cent in

the Baroda study to 43 per cent in the rural
West Bengal and Gujarat studies, to as high
as 74 per cent in the Bombay study. Morbid­
ity rates were considerably lower when mea­
sured by examination than by history (Table
3) in three of the four studies. In the Baroda
study for example, where 65 per cent of all
respondents reported one or more
gynaecological morbidities, only 26 per cent
were observed to have a gynaecological con­
dition on examination. In the rural West Ben­
gal study, these proportions fell from 65 per
cent to 43 per cent and in rural Gujarat from
83 per cent to 43 per cent. Only in Bombay
did the rates coincide. Here, 74.1 per cent of
all women reported one or more
gynaecological conditions and 73.6 per cent
of all women, although not necessarily the

TABLE 4
Gynaecological morbidity by clinical history
(% women having morbidity)

Gynaecological
condition

URBAN

RURAL

West Bengal
(N-395)

Gujarat
(N-293)

Baroda
N-548)

Bombay
(N - 715)

A.

Vaginitis

3.8

10.2

1 1.3

15.4

B.

Cervical erosion/ectopy alone

2.4

19.8

5.5

21.5

C.

Cervicitis
Cervicitis alone
Cervicitis with erosion

14.4
4.8
9.6

7.8
5.5
2.3

13.5
4.9
8.6

39.6
21.1
18.5

D.

Pelvic inflammatory disease

1.0

8.2

8.4

16.5

E.

Genital Prolapse
Anterior (cystocoele) only
Posterior (rectocoele) only
Anterior and posterior
Uterine

1 7.3
5.7
1.5
3.0
7.1

NR
NR
NR
NR
NR

4.6
1.6
0
2.4
0.5

18.2
5,5
3.5
7.7
1.5

F.

Other
Polyp
Fibroid

0.2
0.0

0.3
1.0

0.2
0.0

1.3
0.6

Women with any morbidity
on clinical examination

42.8

42.7

26.1

73.6

Mean number of morbidities among
women having any morbidity

1.22

1.17

1.99

1.79

NR: Not recorded.

Vol. 43, No. 4, December 1997

13

Gynaecological morbidity

same women, had morbidity on clinical ex­
amination.

Considerable heterogeneity also exists
among studies in the prevalence and rela­
tive importance of specific gynaecological
morbidities. Marked variation between stud­
ies was evident for both cervicitis (ranging
from eight per . ent in the rural Gujarat study
to 40 per cent in the Bombay study) and for
cervical erosion (from two per cent in the
rural West Bengal study to over 20 per cent
in the rural Gujarat and the Bombay stud­
ies). Vaginitis was also an important source
of morbidity (10-15 per cent of women) in
all but the rural West Bengal study. Rates of
pelvic inflammatory disease ranged between
one per cent and 17 per cent in all studies.
Rates of genital prolapse were similar (1718 per cent) in two of the three studies where
this information was recorded. The low rates
in the Baroda study may be indicative of the
more conservative diagnostic procedures
adopted there.
Discussion

This paper has presented data on the
prevalence of gynaecological morbidity at
four very different sites in India, based upon
both women's self-reported histories as well
as clinical examinations. Three major find­
ings emerge. First, levels of gynaecological
morbidity are unacceptably high at all sites.
Over two in three women report one or more
conditions in all studies. No fewer than one
in four women, and as many as three in four
women are observed on clinical examination
to have one or more gynaecological morbidi­
ties. Such conditions as vaginitis, cervicitis
and pelvic inflammatory disease affect more
than ten per cent of women in most studies.
These results must be regarded as minimum
estimates, given that the results of labora­
tory tests, which could detect additional in­
fections of the reproductive tract, have not
been considered.

14

Second, considerable inter-study varia­
tion is evident in the levels and patterns of
morbidity. For example, between 65 per cent
and 84 per cent of all respondents report
gynaecological problems, and between 26
per cent and 74 per cent are assessed, on
clinical examination, to have one or more
gynaecological conditions. The Bombay
study stands out in terms of its high level of
morbidity in general (74 per cent), and of
cervicitis in particular (40 per cent), a find­
ing which warrants further investigation.
Third, despite this variation, a common
constellation of disorders emerged in all
studies. From women's self-reports, men­
strual problems and excessive discharge
were the most commonly cited morbidities.
Among morbidities revealed by clinical ex­
amination, cervicitis, vaginitis and pelvic
inflammatory diseases — all infections of the
reproductive tract — are among the leading
morbidities in all studies. Such other condi­
tions as genital prolapse and cervical erosion
were also prominent morbidities in certain
sites. For example, genital prolapse was
found in 17-18 per cent of women in the
Bombay and rural West Bengal studies; cer­
vical erosion was found in 40 per cent of
women in the Bombay study and 22 per cent
of women in the rural Gujarat study.

A number of factors may account for the
observed variations between studies in lev­
els and patterns of gynaecological morbid­
ity. For one, studies were independently de­
signed and implemented, with attendant
differences in sampling procedures and
sample loss. While all studies were commu­
nity based, they differed in their success in
recruiting women to participate in the clini­
cal phase of their studies. Since large pro­
portions refused the clinical examination,
sample loss, ranged from 29 per cent and 35
per cent in the urban studies in Bombay and
Baroda respectively, to 55 per cent and 65
per cent in the rural studies in West Bengal

The Journal of Family Welfare

Gynaecological morbidity

and Gujarat respectively. Such high rates of
sample loss are not surprising given the gen­
eral reluctance of Indian women to undergo
an intrusive procedure such as a pelvic ex­
amination.
The issue of sample loss assumes increas­
ing importance given the likelihood of se­
lection bias among women who agreed to
undergo clinical examinations. Women with
serious reproductive health problems may
be significantly more likely to consent to
clinical examination compared to women
without pronounced symptoms. In two of
the four studies it was possible to assess the
extent to selection bias by comparing re­
ported morbidity of women who underwent
clinical examination with those who refused.
In the Baroda study, where sample loss was
moderate, while similar proportions of ex­
amined and non-€xamined women reported
menstrual problems (58 per cent versus 56
per cent), a somewhat higher proportion of
examined women reported excessive dis­
charge, relative to women who refused ex­
amination (22 per cent versus 15 per cent).
In the rural Gujarat study, where the sample
loss was the highest, selection basis was
much more pronounced, with higher pro­
portions of examined women reporting
menstrual problems (59 per cent versus 43
per cent among women who refused) and
markedly higher proportions reporting ex­
cessive discharge (57 per cent versus 29 per
cent). These results suggest that the overall
prevalence of gynaecological morbidity may
be biased upward, as a result of sample se­
lectivity, with the effect most marked in stud­
ies with higher rates of sample loss.

Differences in data collection procedures
may also have accounted in part for the ob­
served differences between studies. For ex­
ample, the extent of probing varied from
study to study and across conditions. Com­
pared to other studies, in the rural Gujarat
study, questions on dysuria were specifically
probed, and may have accounted for the
Vol. 43, No. 4, December 1997

relatively high rates of this condition re­
ported. Similarly, as indicated earlier, while
three of the studies probed for other condi­
tions, the Baroda study did not, possibly
accounting for the substantially lower rates
of such morbidities as excessive discharge
reported here.
Another difference across studies was the
lack of uniformity in the identification of
certain morbidities by examination. For ex­
ample, in the Barod study, cervical erosion
and genital prolapse were not classified as
morbidity if they were assessed by the
gynaecologist to be mild. In contrast, in the
Bombay study, milder cases of prolapse and
erosion were classified as morbidity. This
may account for the relatively higher rates
of these conditions in the Bombay study and
the contrastingly lower incidence of these
conditions in the Baroda study.

A final and perhaps most plausible expla­
nation for the observed differences is that
there exist genuine differences between the
study sites in the prevalence and patterns of
gynaecological morbidity. The four sites rep- •
resent markedly different socioeconomic
and cultural settings with possible attendant
differences in living conditions, sexual and
reproductive health behaviour, and access
to health care, all of which might affect pat­
terns of gynaecological morbidity. Further
research to better understand the underly­
ing environmental and behavioural factors
which predispose women to the risk of
gynaecological problems, or influence the
women's ability to resolve them, is clearly
warranted. It remains likely that no single
pattern of reproductive morbidity could be
considered as representative in a country as
large and heterogenous as India.
Despite these variations, it is clear from
these studies that gynaecological morbidity
constitutes a major public health problem,
one which remains largely unaddressed
within the current health system. Taken col15

Gynaecological morbidity

lectively, these results present a forceful ar­
gument for greater attention to, and invest­
ment in, the reproductive health needs of
poor Indian women.
ACKNOWLEDGEMENTS

These four studies have been funded by
grants from the Ford Foundation, New
Delhi; the support is gratefully acknowl­
edged. A large number of colleagues at each
study site have provided invaluable assis­
tance at various stages — field work, data

collection, analysis and report preparation of these studies: it is only the conventions of
publications that prevent all from being
named here. We are also very grateful to Dr
Shireen Jejeebhoy and Dr Michael Koenig for
their comments and suggestions which went
a long way in bringing this paper to its
present form. Most of all, we hope that this
paper may contribute to improvements in
the lives of women of the kind who went
out of their way to make these studies pos­
sible; the studies were carried out for them.

REFERENCES

Bang RA, Bang AT, Baitule M, Choudhary Y, Sarmukadan S and Tale O: 'High prevalence of gynaecological
diseases in rural Indian women'. Lancet, 1:85-87 (1989).
2. Younis N, Khattab H, Zurayk H, El. Mouethy M, Amin MF and Farag AM: 'A community study of gynaecological
and related morbidities in rural Egypt', Studies in Family Planning, 24 : 174-86 (1993).
3. Brabin L, Kemp J, Obunge OK. Ikimalo J, Dollimore N, Odu NN, Hart CA and Briggs ND: 'Reproductive tract
infections and abortion among adolescent girls in rural Nigeria', Lancet, 345 : 300-304 (1995).
4. Bhatia JC, Cleland J, Bhagavan L, et al.: (Unpublished) 'Prevalence of gynaecological morbidity among women
in south India.'
5. Wasserheit JN, Harris JR, Chakraborty J, Kay BA and Mason KJ: 'Reproductive tract infections in a family
planning population in rural Bangladesh', Studies in Family Planning, 20 : 69-80 (1989).
1.

16

The Journal of F.imily W'elfare

ujh ix; s'

f

Soc. Sei. Med. Vol. 43. No. 10. pp. 1507-1516. 1996
Copyright © 1996 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
0277-9536/96 $15.00 + 0.00
80277-9536(96)00105-0

Pergamon

OBSTETRIC MORBIDITY IN SOUTH INDIA: RESULTS-FROM
A COMMUNITY SURVEY
JAGDISH C. BHATIA1 and JOHN CLELAND2
'Indian Institute of Management. Bangalore, India and -Centre for Population Studies, London
School of Hygiene and Tropical Medicine, London, U.K.
Abstract—A sample of 3600 mothers with at least#one pre-school age child were interviewed in detail about
obstetric problems associated with their last confinement. About 10% reported one or more of the classic
symptoms of pre-eclampsia; and 8% reported symptoms of potentially life-threatening conditions during
delivery, most notably prolonged labour of over 18 hr. Disorders during the post-partum period were
more common; 10% reported excessive bleeding, loss of consciousness or convulsions (all indicative of
potentially serious conditions) and an additional 17% reported symptoms of infections. The level of health
care received by women is described. Copyright © 1996 Elsevier Science Ltd
Key wo>ds—obstetric, maternal, morbidity, self-reports, health services

INTRODUCTION

The policies and plans of many nations have long
accorded priority to maternal and child health
(MCH) programmes. Though better maternal health
has been an implicit goal, this concern has rarely been
translated into effective services. Priority has always
been given to child survival and mother’s health has
remained relatively neglected. This neglect is all the
more regrettable in view of the well documented
linkages between mother's and child’s health and the
difficulty of ensuring the health and survival of a child
without adequately safeguarding the health of the
mother. Moreover maternal health is important in its
own right, irrespective of its implications for child
welfare.
Maternal health received greater attention after the
safe motherhood initiative was launched at an
international conference held in Nairobi in 1987 [1].
During the conference and thereafter, emphasis has
been mostly on maternal mortality. Mortality
statistics are often, and justifiably, used as a weapon
of international advocacy. Just as the case for
prioritising child health programmes was buttressed
by publication of infant mortality rates, maternal
mortality estimates are now' used to highlight the
plight of pregnant women in less developed countries.
It is commonly claimed that about half a million
women die each year because of the complications of
pregnancy and childbirth and that all but 6000 of
tnese deaths t ike place in developing countries [2]. Il
has also been reported that maternal mortality ratios
are 15-20 times higher in developing countries and
that the life time risk of maternal death in developing
countries is 1 in 51 as compared to 1 in 1687 in
developed countries [3, 4].

■»^WH0 estimates also suggest that 88-98% of
pregnancy-related deaths are avoidable. A study in
India reported that 78% of maternal deaths occurring
in the study population could have been prevented by
specific timely actions [5]. In the developed countries
also, serious pregnancy-related morbidities arise, but
deaths are averted through early medical interven­
tions. In a study conducted in the U.K. obstetric
morbidity was reported in approx, one-fourth of the
cases and about 1 % of these were life-threatening or
“near miss” episodes [6].
Maternal mortality is just the tip of the iceberg of
the obstetric health problems of women. Many
women do not die of causes related to pregnancy but
suffer severe morbidities. A small prospective study
conducted in a village in India reported that there are
16.5 pregnancy-related morbidities for every ma­
ternal death [7]. A study from Nigeria found that, for
every woman who dies of maternal causes, 15 suffer
from permanent disabilities [8]. Another analysis
indicates that in developing countries for each
maternal death, a further 10-15 women suffer serious
impairments [9]. Based on some of these estimates, it
has been calculated that there are 8.25 million
mprbidities each year worldwide [10]. Others have
calculated that there are over 100 acute morbid
episodes for every maternal death, giving a global
total of 62 million morbidities annually [11]. These
estimates, though crude and unreliable, nevertheless
point to the magnitude of the problem of maternal
morbidity.
As the realisation grew about the magnitude of
maternal morbidity and preventability of maternal
deaths, attention shifted from studying levels of
maternal mortality to investigating the levels and
determinants of pregnancy-related morbidity. There

1507

<voX

9

A-Moaai.-*- -

1508

Jagdish C. Bhatia and John Cleland

is no clear definition of maternal morbidity. knowledge and to improve our understanding of the
However, it is usually considered to encompass any aetiology of maternal morbidity, and thus start to
pregnancy-related problems during the ante-natal, form a basis* for effective management of healtft
delivery and post-natal period. This approach services to pregnant women.
enforces consistency with the definition of maternal
mortality, which according to International Classifi­
cation of Diseases is “the death of woman while
STUDY DESIGN AND METHODS
pregnant or within 42 days of termination of
pregnancy, irrespective of the duration and the site of
The study was conducted in the Indian State of
the pregnancy, from any cause related to or Karnataka which is located in the southern part of
aggravated by the pregnancy or its management but the country. According to the 1991 census the state
not from accidental or incidental causes” [12]. The has a population of 44.8 million which accounts for
restriction to 6 week after delivery implied in the 5.3% of the total Indian population. The male,
definition of maternal morbidity has been criticized female and combined literacy rates in the state are 67,
on the grounds that several pregnancy-relatedI 44 and 56 as against the All India averages of 63, 39
problems continue much beyond the 6 week post­ and 52, respectively. A little more than two-fifths
partum period [13]. The absence of an agreed (44.0%) of married couples are protected against
definition of maternal morbidity creates conceptual pregnancy through contraception, mainly through
and measurement problems that may contribute to female sterilisation. In terms of health and family
the lack of information in this important area of planning programmes, Karnataka maintains a
woman’s health [14]. There is considerable overlap progressive profile compared to many other Indian
between women’s health, reproductive health and states. Health and family planning services in the
maternal health and these terms are often used rural areas are made available through a large
interchangeably. This paper is limited to maternal or network of primary health centres (PHCs) and
obstetric morbidity, namely problems occurring subcentres. In urban areas, maternity centres and
during pregnancy, delivery and puerperium.
hospitals run by the government provide free curative
There is a dearth of community-level information and preventive health services. In addition, private
on reproductive morbidity in developing countries. health provision in the state is becoming increasingly
Although a few studies on gynaecological morbidity common. Although many indicators of development
have been reported [15-17], community-based data place Karnataka at a much higher level than many
on obstetric morbidity are very rare. A search of other Indian states, there are considerable regional
literature on the subject reveals that in addition to a variations within the state. The districts located in the
small scale study in two villages of Rajasthan (India), northern part of the state remain particularly
referred to above, information about levels and types backward in terms of social and economic develop­
of maternal mortality at the community level in the ment.
developing countries is available from three prospec­
The present study is part of a major research effort,
tive studies conducted in Indonesia, Kenya and funded by the Ford Foundation, to investigate the
Bangladesh [18-20] and one retrospective study in pathways through which mother’s education influ­
Egypt [17]. The rates of maternal morbidity in all ences child survival. The main study has several
these studies are found to be very high, except in components: anthropological studies; investigation of
Kenya where less than one-fifth of the women primary schools in three states of India; a
reported complaints during the last childbirth. The cross-sectional survey; and a prospective study. The
Kenyan study, however, was designed to collect anthropological studies aimed to identify possible
information about the diseases of children and linkages and to develop instruments and methods for
information about maternal morbidity was obtained subsequent quantitative work. During these in-depth
only incidentally. In addition, a few hospital based investigations it was found that the mother’s health
studies have attempted to estimate the occurrence of is intricately related with that of the child. It was
certain specific conditions during pregnancy, delivery decided therefore to collect detailed information in
and post-partum period but there are very few the cross-sectional and prospective studies on
estimates of the percentage of all women who different aspects of mothers’ health.
experience specified problems during the entire
a
The analysis contained in V111O
this paper
is based on a
46 week period spanning pregnancy, labour, delivery cross-sectional survey conducted
.
------- 1 a sub-district of
and puerperium. Furthermore none of the studies Karnataka state, situated about 70 km from the
have attempted to relate problems during different capital „
Bangalore. This sub-district was chosen
stages of the child bearing process. In view of scarcity because
it was typical of rural Karnataka and within
..It
of data on the subject, the WHO Technical Group on reasonable distance of the capital city. According to
Reproductive Morbidity recommended that the the 1991 census, the sub-district has a total
highest research priority should be given
to studying
.
population of about 117,000. It has one town with
levels of obstetric morbidity (21). The present study 47,000 inhabitants. The study population comprised
was conducted with a view to fill this gap ini mothers who were less than 35 year old and had at

1509

Obstetric morbidity in south India
least one child under 5 years of age. The upper age
' limit was imposed because it was anticipated that
the number of older women with pre-school children
would be too few to sustain separate analysis.
Because the major purpose of the survey was
explanatory rather than descriptive, no attempt was
made to draw a strictly representative sample of the
sub-district. Rather, the sampling strategy was based
on logistical considerations; all eligible women living
in the town and the 48 villages having a population
of at least 500 persons were included in the sample.
A complete listing was made of all eligible women
currently residing in these locations and attempts
were made to interview all of them. To maximize the
response rate, at least three call backs were made,
including visits early in the morning and late at
night to contact respondents who worked away from
home. The achieved sample size was 3600 (2400 in
rural areas and 1200 from the town), with a contact
and response rate of over 95%. The main reason for
non-response was that women were repeatedly not
at home; only five women refused to be interviewed.
Interviewing was performed by female interviewers
who had degrees in social sciences or related
subjects. They were familiar with the local culture
and had prior experience of conducting similar
surveys.
As mentioned earlier, the main purpose of the
survey was to identify pathways by which the
mother's education influences the health and survival
of child. Accordingly, most of the interview was
devoted to this theme. Questions on respondents'
reproductive morbidity formed one of 14 sections in
the questionnaire. In order to improve the correspon­
dence between bio-medical conditions and self­
reported symptoms, the first step in framing
questions was to ask an experienced female
obstetrician/gynaecologist to prepare a comprehen­
sive list of reproductive morbidities (both obstetric
and gynaecological) along with complete details of
symptoms for each condition. Piloting and pre-test­
ing ensured that these symptoms were described in
everyday terminology that women could understand.
This paper considers only obstetric problems. The
questions on symptoms of obstetric morbidity were
administered in sequence, starting with symptoms
during pregnancy and then proceeding to those
during delivery and ending w-ith symptoms in the
post-partum period. All questions related to the most
recent live birth. As mentioned earlier, all respon­
dents had a surviving child under 5 years of age. Thus
the recall period ranged between about 2 and
60 months. This period is rather long for eliciting
information on minor complications. However,
exploratory analysis showed that the number and
range of reported morbidities were not related to the
time elapsed since the birth.
The morbid conditions reported by women have
been grouped according to their potential severity
and life-threatening nature. The number of groups

and conditions included under each is indicated in
Table 1. Obstetric morbidity is first described by
frequency distributions and cross tabulated by type of
consultation/treatment sought for each specific
condition. Subsequent analyses are based on the
seven groups or categories. All respondents were
classified by whether or not they reported one or
more symptoms in each of the seven categories.
Bivariate analysis has been carried out to compare
the prevalence of symptom groups in various
socio-economic and demographic classes. Analyses of
variance were performed and Pearson’s chi-square
values and the probabilities associated with them
were calculated. The variables included in the
bivariate analysis are: urban-rural residence; religion/caste; woman’s education; household economic
status; age at pregnancy; and pregnancy order. With
the exception of economic status, the definition of
these variables is straightforward. Economic status
was based on the imputed financial value of consumer
durables, such as radio, fan, bicycle, furniture and so
on. After calculating the total monetary value of
possessions, households were divided into three
categories of approximately equal size. In order to
estimate the net effect of each of these variables on
specific grouped types of morbidity a logistic
regression analysis was then performed. In addition
to the socio-economic and demographic factors used
in the bivariate analysis, two additional variables
were added. In the analysis of natal morbidity, the
occurrence of ante-natal problems was included as a
predictor. Similarly post-partum morbidity was
analyzed by the occurrence of natal problems. These
additions permit exploration of the links between
morbidity at different stages of the childbearing
process. All variables were categorical in nature or
grouped, and for each variable one category was
selected as the reference category. Regression analysis
estimates the coefficient for each of the remaining
categories of the variable. Results are presented in
terms of odds ratios, which express the magnitude of
the effect of each category on the outcome, relative
to the reference category.

FUNDINGS

Approximately two-fifths of all respondents re­
ported at least one morbid condition associated with
their last pregnancy. A further breakdown shows that
about one-fifth (18%) reported at least one problem
during the ante-natal period; about 8% experienced
a problem during delivery and another 23% indicated
problems during the post-partum period. A total of
2305 abnormalities were reported, equivalent to 0.64
episodes on an average for the entire sample and
about 1.6 episodes each for those reporting at least
one morbid condition. The types of problem reported
during pregnancy, delivery and post-partum are
analyzed below.

*■

1510

Jagdish C. Bhatia and John Cleland

Ante-natal problems

matic heart disease and varicose veins form the third
group.
A total of 485 adverse conditions during pregnancy
were reported. Altogether 10.2% of women reported
one or more of the potentially life-thieatening
conditions in the Group I category. The most
common symptoms reported were swelling of hands
and face (4.3%), hypertension (3.8%) and fever for
3 or more days (2.9%)
Nearly 10% of the sample experienced severe
vomiting during pregnancy. Some have argued that
nausea during pregnancy is a sign of good health and
certainly should not be classified as a morbid
condition [22], However, if the condition is perceived
to be so serious that it requires medical attention or
if it prevents the performance of normal activity, it is
certainly a departure from normality and, from the
sufferer’s perspective, is rightly regarded as an illness.
Furthermore, vomiting is usually accompanied by
diminished food intake, while may deplete the body’s
carbohydrate stores, giving rise to ketosis and other
toxic symptoms. It may be noted, at this juncture,

Problems reported during pregnancy have been
grouped into three categories. Group 1 includes
potentially severe or life-threatening conditions such
as .swelling of hands and face, fits and convulsions
and hypertension which are all classical signs of
prc-eclampsia. Uterine bleeding is also included in
Group 1. It may result from premature separation of
the placenta, trauma to the abdomen and genital
region or various pathological lesions of the cervix.
The final two symptoms in Group 1 are fever for
three or more days and fever with rigor. The former
typically results from a viral infection and the latter
is indicative of malaria. Among possible conse­
quences are foetal loss or malformation and
premature labour.
Group 2 comprises severe vomiting which may
cause extreme discomfort but is not usually medically
dangerous. Preexisting conditions which have the
potential to complicate the process of delivery, such
as urinary problems, tuberculosis, jaundice, rheu-

table I. Obstetric problems among women, according to' type of consultation/treatment

Type of me rbidilv
During p«cgnsncy
Group I
Swelling of hands
and face
Fits and convulsions
Hypertension
Bleeding
Fever for 3 davs
or more
Fever with rigor
Group 2
Severe vomiting
Group 3
Urinary problems
Tuberculosis
Jaundice
Rheumatic he«.n
disease
Varicose veins
During delivery
Group I
Labour > 18 hr
Fits or convulsions
Excessive bleeding
Loss of consciousness
Ruptured uterus
Tom vagina or cervix

No. & %
reporting
problem

Place of consultation/treatment (%)
No. & % seeking
consultation treatment

PHCPHU

Sub-centre

Government
hospital

Private
hospital

Other

Total

156 (4.3)

129 (82.7)

0.0

2.3

41.1

53.5

3.1

100

12 (0.3)
135 (3.8)
34 (0.9)
106 (2.9)

11 (91.7)
129 (95.5)
27 (79.4)
96 (91 4)

0.0
1.6
11.1
2.1

0.0
0.8
7.4
2.1

45.5
27.9
4^1.4
31.3

54.5
69.0
37.0
61.5

0.0
0.8
0.0
3.1

100
100
100
100

42 1.2)

37 (88.1)

0.0

2.7

43.2

54.1

0.0

100

347 (9.7)

287 (82.7)

1.4

4.2

32.8

60.6

0.7

100

34 (0 9)
2 (0.1)
6 (0.2)
16 (0.4)

30 (88.2)
2 (100.0)
5 (88.3)
12 (80.0)

3.3
0.0
0.0
8.3

0.0
0.0
0.0
0.0

40.0
1000
20.0
50.0

56.7
0.0
80.0
41.7

0.0
0.0
0.0
0.0

100
100
100
100

46 (1.3)

42 (93.3)

0.0

2.4

35.7

59.5

2.4

100

204 (5.7)
4 (0.1)
45(1.3)
33 (0.9)
4(0.1)
26 (0.7)

147 (72.4)
3 (75.0)
28 (62.2)
31 (96.9)
2 (50.0)
26 (100.0)

2.0
00.0
3.6
0.0
0.0
0.0

6.8
0.0
7.1
6.5
0.0
0.0

36.7
33.3
46.4
25.8
50.0
50.0

44.2
66.7
46.4
67.7
50.0
50.0

10.2
0.0
3.6
0.0
0.0
0.0

100
100
100
100
100
100

204 (57.0)
43 (87.8)

2.5
0.0

7.8
7.0

34.3

32.6

52.5
58.1

2.9
2.3

100
100

4 (100.0)

0.0

0.0

0.0

100.0

0.0

100

111 (92.5)
34 69.4)
293 (70.3)
213 (76.6)

0.0
0.0
2.4
1.4

5.4
8.8
12.3
7.5

29.7
47.1
38.9
37.1

59.5
41.2
36.2
46.5

5.4
2.9
9.9
7.0

100
100
100
100

21 (80.8)
73 (85.9)
22 (28.9;

0.0
1.4
4.8

4.8
4.1
0.0

33.3
37.0
45.4

57.1
50.7
45.4

4.8
5.5
4.8

100
100
100

During post-rut«l period
Group I
Excessive bleeding
359 (10.0)
Shock (loss of
49 (1.4)
consciousness!
Fits or convulsions
4 (0.1)
Group 2
High fever
120 3.3)
Foul discharge
50 (1.4)
Lower abdominal pain 417 (11.6)
Pain in pelvic region
278 (7.7)
Group 3
Breast abscess
26 (0.7)
Painful urination
86 (2.4)
Depression
81 (2.3)

Obstetric morbidity in south India
that 83% of those experiencing severe vomiting
during pregnancy sought consultation or treatment.
Problems in the third group, comprising an­
tecedent risk factors, are reported by very few
women. Only 94 respondents, 2.6% of the total
sample, reported one or more of these conditions.
Problems during labour and delivery

Conditions reported during labour and delivery are
all potentially serious or life-threatening problems. In
total, 7.7% of the sample reported a symptom.
Prolonged labour was the dominant problem,
reported by 5.7% of the women. This condition may
be caused by malpresentation of the foetus or by
defects of womb muscles. It may lead to severe
maternal and foetal distress, with possible additional
complications of haemorrhage and uterine rupture.
Excessive bleeding and loss of consciousness were
each reported by about 1% of the sample. The
aetiology of these conditions is complex and varied
but both are potentially life-threatening.
Problems during post-partum

Problems of various types during the 6 week
post-partum were more common than problems
reported during pregnancy or delivery. Nearly
one-quarter (23%) of all respondents reported at least
one adverse health condition. These conditions have
been grouped into three categories. The first,
comprising excessive bleeding, loss of consciousness
and convulsions, arc all indicative of potentially
life-threatening conditions. A total of 390, or 10.8%
of respondents, reported at least one such condition,
excessive bleeding (10%) being by far the most
common symptom. Group 2 symptoms are all
indicative of infection, which if left untreated could
lead to chronic gynaecological-morbidity. A total of
596 women, equivalent to 16.6% of the sample,
reported at least one such symptom (fever, discharge,
lower abdominal or pelvic pain). Lower abdominal
pain (11.6%) was the most common symptom among
this group. Group 3 comprises three miscellaneous
conditions: breast abscess, painful urination and
depression. Only 4.8% of the total sample reported
one or more of these symptoms.

Socio-economic and demographic predictors of
morbidity

The relationships between socio-economic and
demographic factors and reported obstetric morbid­
ity are first described in Table 2 in the form of
bivariate analyses and are later assessed by logistic
regression (Table 3). In general terms, it might be
expected that rural, less educated, lower caste women
from households of low economic status might
experience more problems during their last preg­
nancy, because of limited access to health care and
low health status prior to conception. However, this
pattern does not emerge. Indeed, for many con­
ditions, it is the relatively privileged sectors of the

1511

population who report more problems. Thus high
caste Hindus (and non-Hindus, mainly Moslems) are
more likely to report ante-natal problems than
middle or lower caste Hindus. Similarly, there tends
to be a positive relationship between education and
likelihood of reporting ante-natal problems, and
greater reported morbidity in urban than in rural
areas.
One possible explanation for these unexpected
variations is that some types of women are more
likely to undergo routine ante-natal checks and that
some symptoms or conditions emerge during these
consultations. In this population, use of ante-natal
services is high. About 90% of women reported at
least one consultation with a health-care provider
during pregnancy. Receipt of ante-natal care is as
high in rural as in urban areas and thus diagnoses by
practitioners cannot account for the residential
variation in reported morbidity. However, an
analysis of health care shows that educated women
and non-Hindu women are significantly more likely
to seek ante-natal care than uneducated women and
hindus, repectively. Can these contrasts in health­
seeking behaviour explain the reported morbidity
differentials? There is certainly a link between
reported problems during pregnancy and ante-natal
care. Among women with problems, 96% received
ante-natal attention, compared to 88% for women
with no problems; this pattern is to be expected
because women who feel unwell during pregnancy
will be particularly likely to seek medical advice.
However it is impossible to ascertain from the
information collected in the survey how many
problems or symptoms were first detected during
consultations. Thus the possibility remains that the
higher levels of ante-natal morbidity reported by
educated women and by non-Hindus may reflect their
greater use of health services. However two factors
suggest that this is unlikely to be the major
explanation. First, this line of argument does not
explain the urban-rural difference in reported
morbidity. And second, it is implausible to attribute
the greater propensity of educated women to report
severe nausea to clinical diagnoses. On balance
the most likely explanation is that urban and
educated women, and non-Hindus, are more aware
of symptoms of ill-health and thus more prone to
report them, than rural, less educated or Hindu
women.
For post-natal problems, however, socio-economic
differences tend to be smaller and in the opposite
direction. For instance, women from poorer house­
holds report a higher incidence of symptoms of
infection in the post-partum than those from more
affluent households, and there is a parallel difference
between low and high caste Hindus not only for
symptoms of infection but also for the potentially
life-threatening conditions of Group 1. Finally the
bivariate analysis reveals no statistically significant
socio-economic differences in the incidence of

«

Tabic 2. Ma.er.,,1 n.orbid.ty

Characteristics of women

Residence
Urban
Rural
Caste/religion
High caste
Middle caste
Lower caste
Non-Hindus
Education
None
I 5 yr
6 + yr
Economic
status
Low
Middle
High
Other characteristics
Age at pregnancy
< 18 yr
18-24 yr

25+ yr
Pregnancy under
1
2-4

No. of
respondents
N = 3595

1197
2398
182
2128
951
333

1888
504
1203
1194
1188
1208

236
2096
1263
674

2518
403
•/’<0.05; •*/’<O.OI; ***/»< 0.001.

by wonlen by

Percentage of mothers reporting morbidity
Ante-natal
Group I N = 366

11.8*
9.4
12.6***

9.8
8.1

17.1

7.71
11.9
13.3
10.9
9.5
10.2

8.5
9.9
11.0
12.3*
9.3
11.9

Group 2 N = 347

12.2***

8.4
12.6

9.7
8.3
11.7
7.7***
9.1
12.9

10.1
10.1

8.8

8.9
9.5
10.0
12.9***
9.3
6.5

_____ _______
Group 3 N = 94

Group I N = 274

3.3
2.3

7.4

5.5***
2.4
1.6

7.8
4.9
7.4

Natal
Group I A' = 390

9.8
11.4

6.6*
10.0

Pos(-natal
Group 2 N = 596
Group 3 N = 173

17.4
16.2
14.3*

5.8*
4.3
2.2*
4.4

5.4

7.8
9.9

12.8
13.2

15.4
17.7
22.2

2.2
2.4
3.3

7.5
9.7
7.0

10.9
III
10.7

16.5*
20.0
15.3

4.1*
6.3

11.8

I9.5***

6.2***
4.8

2.8
3.3
1.8

2.5*
2.3

4.5

2.5*
2.3

4.5

9.0
7.1
6.9

6.4
7.3
8.4

10.3

10.5

5.3
7.5

f£.
Sn
D3

16.2
14.4

5.3

3.5

E’
M

O
S'

O
3
Q.

9.3**

9.8
12.9

10.1 •
6.9

12.0

7.7

11.2

10.5

17.8
15.5
18.1

3.8
4.4
5.6

14.1*
16.7
19.9

4.3
6.9

I

1513

Obstetric morbidity in south India
potentially life-threatening natal conditions, such as
' prolonged labour or excessive bleeding.
Most of these relationships remain substantially
unchanged when they are re-assessed in the
multivariate analysis. In particular the greater
tendency of non-Hindus and better educated women
to report ante-natal problems is confirmed, and these
are among the strongest net effects that emerge from
Table 3. The general impression, however, is that
obstetric problems are experienced by women of all
backgrounds and that powerful socio-economic or
cultural risk factors do not exist.
Demographic factors are represented by two
variables: age at pregnancy and pregnancy order.
Neither factor appears to be a consistent predictor
of obstetric problems in this study population.
However, primigravidae are more likely than other
women to report ante-natal problems and exper­
ience difficulties in labour. The logistic regression
confirms these effects, but there is no parallel
difference with regard to post-natal problems.
Differentials by current age are mostly small and
non-significant.
The logistic regression also permits an assessment of the links between occurrence of ante-natal,
natal and post-natal problems. As expected, ante­
natal problems are strongly related to diffi­
culties experienced during delivery. Similarly,
difficulty during labour is a strong predictor of the
reported occurrence of problems in the post-natal
period.

Treatment of obstetric morbidity
All women who reported obstetric morbidity were
asked whether they had sought treatment or
consultation for each specific symptom and, if so,
what was that main source of care. The results are
shown in Table 1. The percentages who reported that
they had sought treatment are remarkably high,
indicating that these conditions were not regarded by
respondents as trivial or unimportant. For ante-natal
problems, between about 80 and 100% reported
treatment for each specific condition. As already
mentioned, a further reason for these high treatment
levels is that some problems are no doubt diagnosed
and treated during routine ante-natal checks. The
prevalence of health seeking for post-natal problems
tends to be lower. For instance, only about 70%
sought advice or treatment for symptoms of infection
(discharge and pain) and the percentage is even lower
(57%) in the case of excessive post-partum bleeding.
Loss of consciousness and high fever, on the other
hand, are more likely to evoke a health seeking
response. Finally, data on treatment of problems
during labour and delivery are difficult to interpret
because the distinction between emergency and
routine admissions was not ascertained in the study.
Altogether 38% of women delivered in a hospital.
The high levels of hospital treatment of prolonged
labour, excessive bleeding and loss of consciousness
suggest that emergency admissions for these compli­
cations of delivery were common.

Table 3. Logistic regression of types of maternal morbidity by selected characteristics of women
Type of morbidity odds ratios

Characteristics of women
Socio-economic
Residence
Urban
Caste/religion
High caste
Middle caste
Lower caste
Education
1-5 yr
6 + yr
Economic status
Middle
High

Other characteristics
Age at pregnancy
< 18 yr
25+ yr
Pregnancy order
1
5+
Ante-natal problems
Group I—Yes
Group 2—Yes
Group 3—Yes
Natal problems
Group I—Yes
Constant
• P < 0.05; ** P < 0.01; ’

Ante-natal

Post-natal

Natal

Reference
category

Group I

Group 2

Group
Group 33

Rural

1.03

1.29*

1.02

Non-Hindus

0.63
0.57***
O.53***

0.93
0.89
0.86

None

1.54***
1.79***

Low

Group 1

Group 2

Group 3

0.79 ,

0.76*

1.01

0.98

1.04
0.52*
0.32***

0.45*
0.86
0.91

0.43 ♦*
0.70
1.00

0.68
0.71*
0.82

0.28*
0.70
0.97

1.14
1.57'

0.91
1.29

1.30
0.82

1.06
1.04

1.27
0.99

1.57*
1.30

0.73*
0.82

0.86
0.78

1.02
0.58

0.80
0.75

0.86
0.88

0.82
0.75*

0.67*
0.49***

18-24 yr

0.89
1.15

0.94
1.26

0.87
1.48

0.72
1.34*

0.88
1.48***

1.27
1.11

0.90
1.22

2-4

1.31*
1.25

1.42**
0.63*

1.13
1.63

1.65***
0.84

1.11
0.86

0.71••
1.15

0.92
1.69*

4.90***
- 2.17

3.81***
- 1.45

2.22***
- 3.07

3.78***
2.69***
1.63

No

No
P < 0.001.

Gr.oup 1

- 1.94

- 2.40

- 3.20

- 2.69

1514

Jagdish C. Bhatia and John Cleland

Two sources of treatment are dominant: govern­
available at the doorstep [25,26]. In a study
ment hospitals and private practitioners or clinics. conducted in southern India, only one:third of the
- ,For the majority of conditions,%the private sector is
maternal deaths identified by community enquiries
preferred to the public sector. The most striking
were found in the records maintained by the rural
finding is clear evidence of the underutilization of
health facilities [5]. Clinical examinations and
primary health centres and sub-centres. Very small laboratory tests gre too expensive to be carried out on
minorities of women, typically well under 10%, a large scale and they do not provide information on
report such facilities as their main source of advice or socio-economic and behavioural aspects. As morbid
treatment. The complex interplay of factors that conditions vary with the gestation period, point
determine health service utilization and choice of
prevalence surveys have to use very large samples to
public vs private sector facilities is addressed
capture all the conditions at various durations of
elsewhere [23].
pregnancy and puerperium. Prospective studies
represent in principle an ideal approach but are very
expensive because a large population may have to be
DISCUSSION
screened to enrol an adequate number of pregnant
There is a scarcity of data on levels of obstetric
women for a statistical analysis. Furthermore, to
morbidity in the developing countries and infor­ maintain the interest of respondents and ensure their
mation on non-medical determinants of such
cooperation for the entire duration of the study is in
morbidity is still rarer. A review of more than 60 itself a daunting task.
studies on the determinants of maternal ill-health by
Cross-sectional retrospective studies thus remain
Campbell and Graham shows that studies have
the most feasible option for the study of maternal
concentrated mostly on pathogenic and biomedical morbidity in developing countries at present. This
risk factors and information on the socio-economic
approach, as demonstrated by the present investi­
and behavioural determinants of maternal morbidity
gation and other studies, can provide a wealth of
is virtually non-existent [24]. This is primarily because
information useful for gauging the extent and broad
maternal health has always been considered a domain
nature of maternal morbidity, and for planning,
of clinical specialists and not an epidemiological or
organization and evaluation of maternal health
social issue. This narrow orientation has forced
programmes [27]. Such studies, however, should be
health managers to look for medical solutions to a
designed by multi-disciplinary teams, involving
problem th^t has important social dimensions. The
medical and social scientists, demographers, bio-statpresent study makes an
j- 0 ana,yse th<; jsticians and management specialists and should take
socio-economic and demographic determinants of into account local terminology and cultural context.
maternal morbidity through univariate and multi­ Medical professionals may criticize this approach on
variate statistical techniques. In addition, the effect of
the grounds that there is no precise correspondence
these factors on health care seeking behaviour of
between self-reported symptoms and clinical diag­
women has also been analysed. This is rarely found
noses. Moreover, there are several asymptomatic
in the available literature on the subject.
conditions which women cannot recognize and
On the basis of a review of available literature,
report. Conversely, some conditions may be imagi­
Koblinsky et al. estimated that about 40% of women
nary or undetectable by medical scientists. But it is
may suffer an acute problem in the process of child the recognition of a problem, its perceived causes and
bearing [11], The results of the present study are
feared outcome rather than a sophisticated medical
consistent with that estimate: about 40% of the
diagnosis which generally determine how a disease
respondents reported at least one morbid episode
will be managed by the patient. Thus demand for and
during ante-natal, natal or post-natal period of their
utilization of services will depend on subjective
most recent delivery. Many of these conditions are perceptions.
potentially life-threatening. Others may act synergis­
This study suggests that approx. 5% of women
tically and might have long term adverse effects on
experienced symptoms of possible pre-eclampsia in
the health of the woman.
connection with their most recent confinement.
There are several possible sources of information
About 10% reported ante-, intra- or post-partum
on maternal morbidity: health facility records; haemorrhage, and nearly 6% had a prolonged labour
clinical examinations and laboratory tests of preg­
of 18 hours or more. These potentially serious
nant women; point prevalence surveys; prospective conditions, though confined to a minority of women,
studies of a cohort of pregnant women; and
nevertheless are sufficiently common to indicate
retrospective cross-sectional surveys. The deficiencies health problems of considerable magnitude. The
of health facility records in developing countries are
other key result, concerning morbidity, is the high
well known. Furthermore, they are unrepresentative
prevalence of symptoms of post-partum infection;
because a majority of pregnant women in many
nearly 17% of the total sample reported at least one
developing countries do not use these facilities. It has such symptom.
been observed in several studies in India that
One limitation of this study is that information was
ante-natal services are not utilised even when they are collected only for the most recent live birth. The

«■

1515

Obstetric morbidity in south India
incidence of induced abortion in India is thought to
" be high and a large number are performed by
unqualified medical practitioners [28, 29]. Such
abortions no doubt contribute to appreciable

morbidity, and thus the failure of the study to cover
this dimension of obstetric morbidity, though
justifiable because of the well-known reluctance of
women to report abortions, results in an underesti­
mate of the magnitude of the problem.
The levels of consultation or treatment for
obstetric problems, particularly during the ante-natal
period, were found to be very high. The study
included-only villages with a population of 600 and
above. It has been shown that the government
medical care facilities and the private doctors tend to
be located in bigger villages, while smaller villages
have no health care facilities, whatsoever. Further­
more, smaller villages lack communication facilities
and are not easily accessible to health facilities [30].
The inclusion of women from such villages in our
study would have resulted in decreased levels of
consultation and treatment. While the level of
consultation during pregnancy was relatively high, a
large proportion of women with potentially life­
threatening conditions such as prolonged labour, fits
and convulsions, excessive bleeding and ruptured
uterus during labour and delivery did not seek any
professional assistance. Similarly several conditions
during the post-natal period like excessive bleeding,
foul discharge, lower abdominal pain and pain in the
pelvic region remained neglected. These conditions
may not endanger the life of the woman, but if
untreated, often result in severe gynaecological or
reproductive health problems. There is thus a need to
strengthen the follow up services to reach recently
delivered women and formulate effective strategies to
provide emergency obstetric care to women in
difficult and complicated labour.
This study also reveals that the majority of the
consultations took place with private practitioners.
The numl)er of unqualified private medical prac­
titioners practising western medicine is rapidly
increasing in rural areas and small urban towns.
People willingly pay for their services rather than
using free services available at government facilities
[31]. The use of the government public health system
was found to be only marginal. This under utilization
clearly indicates that perceived quality of services at
government health facilities, particularly at the
;1 in the rural areas, was
sub-centre and PHC level
need to improve quality of
relatively poor. There is a
services available at government health facilities, in
order to ensure their optimum utilization. Assessment
of services offered by private practitioners is a further
research priority.

able assistance provided by Mr N.S. Sanath Kumar and Mr
S. Ramaswamy in data processing. The interviewers who
vvuvvvvw
e_______
collected ________
data under „
rigorous
field conditions deserve our
appreciation and commendation. We are also grateful to Dr
Oona Campbell for her expert advice on analysis,

REFERENCES

1. Mahler H. (1987) The Safe Motherhood Initiative: a call
to action. Lancet I, 268-270.
2. World Health Organization (1986) Maternal mortality
rates, a tabulation of available information. Division of
Family Health, Geneva, WHO.
3. Rochat R. W. (1981) Maternal mortality in the United
States of America. World Health Statistics Quarterly 34,
2-13
4. Tinker A.
A. and Koblinsky M. A. (1993) Making
Motherhood Safer. World Bank Discussion Paper 202,
Washington: World Bank.
5. Bhatia J. C. (1993) Levels and causes of maternal
mortality in southern India. Studies in Family Planning
24, 310-318.
. ,
6. Stones W. et al. (1991) Life threatening ‘near miss
episodes. Health Trends 23, 13-15.
7. Datta K. K. et al. (1980) Morbidity pattern amongst
rural pregnant women in Al war, Rajasthan a cohort
study. Health and Population Perspectives and Issues 3,
282-292.
8. Harrison K. (1985) Child-bearing, health and social
priorities: a survey of 22774 consecutive births in Zaria,
Northern Nigeria. British Journal of Obstetrics and
Gynaecology 92, 5.
9. Measham A. R. and Rochat R. W. (1987) Slowing the
work: better health for women through family planning.
A background paper prepared on International Confer­
ence on Better Health for Women and Children Through
Family Planning, Nairobi, Kenya, 6-9 October.
10. Walsh J. A. et al. (1989) Maternal and perinatal health
problems. In: Jamison D.T. and Mosley W.H. (eds)
Evolving Health Sector Priorities in Developing
Countries. The World Bank, Washington, DC.
11. Koblinsky M. A., Campbell O. M. R and Harlow D.
(1993) Mother and more: a broader perspective on
women’s health. In: Koblinsky M., Timyan J. and Gay
J. (eds) The Health of Women: A Global Perspective.
Westview Press, Oxford.
12. World Health Organisation (1977) International
classification of diseases. Manual of the International
Statistical Classification of Diseases, Injuries, and
Causes of Death, Ninth revision. World Health
Organisation, Geneva.
13. Campbell O. M. R. and Graham W. J. (1991)
Measuring maternal mortality and morbidity—levels
and trends. Maternal and Child Epidemiology Unit,
London School of Hygiene and Tropical Medicine,
London.
14. Graham W. J. and Campbell O. M. R. (1992) Maternal
health and measurement trap. Social Science &
Medicine 358, 967-977.
15. Bang R. et al. (1989) High prevalence of gynecological
diseases
in rural Indian women.
I, 85-88. tract
wTs^rheit
°(1989)Lancet
Reproductive
-. ... - — x- - z ‘
t

infections in a family planning population in rural
Bangladesh. Studies in Family Planning 20, 2, 69-80.
17. Younis N. et al. (1993) A community study of
gynecological and related morbidities in rural Egypt.
Studies in Family Planning 24, 3, 175-186.
Acknowledgements—Financial support for the study was" 18. HullW. J. (1983) The Ngaglik study: an inquiry into
birth interval dynamics and maternal and child health
provided by the Ford Foundation which is gratefully
in rural Java. World Health Statistics Quarterly 36,
acknowledged. We would also like to thank Dr N.S.N. Rao
100-118.
in the organization and supervision of field work, and the,.

1516

Jagdish C. Bhatia and John Cleland

19. Voorhkoeve A. M„ Muller A. S. and W'Oigo H. (1984)
Child Epidemiology Unit, London School of Hygiene
The outcome of pregnancy. In: Van Gineken J.K. and
and Tropical Medicine.
Muller A.S. (eds) Maternal and Child Health in Rural 25. Kanitikar T. and Sinha R. K. (1989) Antenatal care
Kenya. An Epidemiological Study. Croom Helm,
services in five states of India. In: Singh S. N. et al. (eds)
London.
Population Transition in India, Vol. 2. B.R. Publishing
20. Goodburn E. A. el al. (1994) An investigation into the
Delhi.
nature and determinants of maternal morbidity related
26. Ramachandran L. (1989) The effect of antenatal and
to delivery and the puerperium in rural Bangladesh.
natal services on pregnancy outcome and health of the
London School of Hygiene and Tropical Medicine
mother and child. Journal of Family Welfare 35, 5
London and BRAC. Dhaka Bangladesh.
34-46.
21. World Health Organisation (1990) Measuring repro
repro-­ 27. Zurayk H. (1993) Concepts and measures of repro­
ductive morbidity. Document No. WHO/mch/90.4,
ductive morbidity. Health Transition Review 3, 1,
Gepeva, Safe Motherhood Programme.
17—38.
22. Weigel M. M. and Weigel R. M. (1988) The association
28. Bhatia J. C. and Ramaiah T. J. (1971) Incidence of
of reproductive history, demographic factors and
induced abortions in a community development area
alcohol and tobacco consumption with the risk of
Social Action 21, 3, 224-232.
developing nausea and vomiting in early pregnancy.
29. Bhatia J. C. (1973) Abortionists and abortion seekers.
American Journal of Epidemiology 127, 562-570
Indian Journal of Social Work 34, 3, 275-285
23. Bhatia J. and Cleland J. (1995) Determinants of
30. Chuttani C. S., Bhatia J. C., Dharam V. and
maternal care m a region of South India. Health
Timmappaya (1973) A survey of indigenous medical
Transition Review 5, 4, 127-142.
practitioners in rural areas of five different states in
24. Campbell O. M. R. and Graham W. J. (1991)
India. Indian Journal of Medical Research 61. 6
Measuring the determinants of maternal morbidity and
962-967.
mortality: defining and selecting outcomes and deter­
31. Bhatia J. C. (1975) Traditional healers and modern
minants, and demonstrating associations. Maternal and
medicine. Social Science & Medicine 9, 15-21.

'



it.; n

Levels and Determinants of Gynecological
Morbidity in a District of South India
Jagdish C. Bhatia, John Cleland, Leela Bhagavan, and N.S.N. Rao

This article presents the results of an assessment of gynecological morbidity among 385 women
with young children residing in a district of Karnataka State, South India. All three main modes of
assessment (clinical examination, laboratory tests, and self-reports) reveal a high burden of
reproductive tract infections. The two most common conditions, identified by laboratory tests, were
bacterial vaginosis and mucopurulent cervicitis. Approximately one-fourth of the women had clinical
evidence of pelvic inflammatory disease, cervical ectopy, and fistula. The contribution of sexually
transmitted diseases to overall gynecological morbidity appears io be relatively modest; 10 percent
were so diagnosed. Associated conditions of anemia and chronic energy deficiency were common.
Severe anemia was found in 17 percent of cases and severe chronic energy deficiency in 12 percent.
These results indicate that radical improvements in women's health in India will require far more
than the diagnosis and treatment of reproductive tract infections. (Studies in Family Planning
1997; 28,2: 95-103)

An ambitious program of action to make reproductive
health services universally available was drawn up at
the International Conference on Population and Devel­
opment held at Cairo, Egypt, in September 1994. Repro­
ductive health has been defined by the World Health Or­
ganization as a ''state of complete physical, mental and
social well being and not merely the absence of disease
or infirmity, in all matters relating to the reproductive
system and to its functions and processes" (UN, 1994).
Such a broad definition does not lend itself readily
to the identification of priorities, but many experts
would agree that the three main dimensions of repro­
ductive health among adult women are: avoidance of
unwanted pregnancies, safe motherhood, and protec­
tion against infections and dysfunctions of the repro­
ductive tract, including sexually transmitted diseases.

Jagdish C. Bhatia is Professor of Health Management, Indian
Institute of Management, Bannerghatta Road, Bangalore,
560 076 India. John Cleland is Professor of Medical
Demography, London School of Hygiene and Tropical
Medicine, London, U.K. Leela Bhagavan is Obstetrician and
Gynecologist, Bangalore Hospital. N.S.N Rao is Consultant
in Biostatistics, Ford Foundation Project, Indian Institute of
Management, Bangalore.

The first dimension has been thoroughly researched, par­
ticularly in developing countries, and a growing body
of evidence exists on obstetric morbidity and mortality.
However, the third dimension has been badly neglected.
Gynecological morbidity can be defined as structural
and functional disorders of the. genital tract not related
to pregnancy, delivery, or the puerperium. It includes
menstrual disorders, reproductive tract infections, cer­
vical cell changes, genital prolapse, and such other con­
ditions as syphilis, urinary tract infections, hypertension,
anemia, chronic energy deficiency (CED), and obesity.
Most knowledge in this area is based on hospital
statistics; information available at the community level
in developing countries is rare. A few community-based
studies have focused on specific problems, such as cer­
vical cancer, vaginal discharge, and reproductive tract
infections among selected groups of women (Wahi et
al.,1972; Bali and Bhujwela, 1969; Wasserheit et al., 1989;
Brabin et al., 1995), but only two such studies of over­
all gynecological morbidity have been reported: one
from India (Bang et al., 1989) and the other from Egypt
(Younis et al., 1993). The study in India covered two
predominantly tribal villages, and only 59 percent of
the sampled women participated in the clinical exami­
nation and laboratory investigations. Furthermore, little
information is available about the determinants of gy­
necological morbidity.

Volume 28

Number 2

June 1997

95

More information is urgently needed to assess the
nature and burden of gynecological morbidity, if the
rhetoric of the Cairo Conference is to be translated into
realistic and cost-effective action. The present study ad­
dresses this need. Its objectives were first to determine
the level of gynecological morbidities and other related
conditions among women in a district of South India
using clinical, laboratory, and other methods, and sec­
ond, to analyze socioeconomic, demographic, and cog­
nitive correlates of gynecological morbidities and re­
lated conditions.

Materials and Methods
The study was conducted in the Indian state of Karna­
taka in the southern part of the country as part of a
wider inquiry, funded by the Ford Foundation, into
pathways through which the behavior and characteris­
tics of mothers influence the health and survival of their
children. A sample of 3,600 women (2,400 living in ru­
ral areas and 1,200 in town) was selected for the origi­
nal study. These women were asked about their symp­
toms of obstetric and gynecological morbidity, as well
as about other matters (Bhatia and Cleland, 1995a,
1995b, and 1996). Because the objective of the main
study was to examine the effect of mothers' behavior
on children's health and survival, all women with chil­
dren between the ages of six to 12 months (the most vul­
nerable age group) were included in the prospective
study. In the original sample, 440 women met this cri­
terion and were included in the study. All were mar­
ried and were younger than 35. They were interviewed
at monthly intervals for one year, at which point they
were invited to undergo a medical examination. A gap
of approximately 15 months occurred between the se­
lection of the sample for prospective study and the clini­
cal examination. A total of 23 women were lost to fol­
low-up prior to the clinical phase. Of the 417 women
remaining, 11 refused to participate or were temporarily
absent (this figure includes five women who appeared
at the examination center but refused vaginal exami­
nation), 17 could not be examined because they were
menstruating and did not appear for their rescheduled
appointments, and four women's specimens were re­
jected by the laboratory'. Thus full gynecological exami­
nations and laboratory investigations were completed
for 385 women, representing 86 percent of the original
sample and 92 percent of those who completed the 12month prospective phase of the study.
A day before the medical examination was sched­
uled, women were interviewed by female social work­
ers who gathered information about the presence of

96 Studies in Family Planning

symptoms of gynecological problems. This information
was not made available to the gynecologist.
Clinics were held in the project office in'town and
at schools in all the villages included in the study. An
experienced female gynecologist visited all these loca­
tions and conducted the examinations over a threemonth period in late 1994. She was assisted by a quali­
fied female technician. The gynecologist took menstrual
histories and anthropometric measurements. A general
medical examination that included blood-pressure mea­
surement and auscultation, and examination of breasts,
adiposity, and external genitalia was performed. Later,
a speculum examination was conducted during which
specimens were taken from the cervix and the poste­
rior fornix. After removal of the speculum, a bimanual
examination was performed to determine if tenderness
was present. A 5-milliliter sample of venous blood was
collected for hemoglobin and Venereal Disease Reference
Laboratory (VDRL) assessments. A midstream sample
of urine was also collected for laboratory assessment
of urinary tract infections. In addition, four slides were
made from swab specimens of which two were fixed
immediately with propyl alcohol for a pap smear and
two were air dried. These specimens were stored in an
icebox and transported daily to one of the wellequipped laboratories in Karnataka, at Bangalore hos­
pital in the state capital. Because the laboratory where
the samples were sent was 70 to 100 kilometers (43 to
62 miles) distant from the villages where the clinics were
held, and because the time lag between the collection
of samples and their deposit at the laboratory ranged
from three to 12 hours, preparation of wet mounts for
the detection of trichomonal vaginosis and Candida
albicans was difficult.
The gynecological morbidities and related conditions
discussed in this report were measured, therefore, by sev­
eral different data-collection procedures: by means of soc­
ial workers' questions; by the gynecologist's careful ques­
tioning, anthropometric measurements, and clinical
examination; and by laboratory investigations. The
precise diagnostic criteria are given in Table 1.
The clinical criteria were in accordance with those
given in standard textbooks of gynecology (Dewhurst,
1981; Jeffcoate, 1982; Howkins and Bourne, 1985). Fur­
ther details of the tests used in the study may be found
elsewhere (Koss, 1968; Ishizuka et al., 1972; Naib, 1970).

Findings
Characteristics of Respondents

One-fourth of the sampled women were residing in
town and the remaining three-fourths in rural areas. A

I

Table 1

Criteria for diagnosing gynecological morbidity and related conditions, Karnataka, India, 1994
Condition
Clinical
Vaginitis
Cervicitis
Cervical ectc-py
Pelvic inflammatory disease (PID)
Genital prolapse

Laboratory
T nchomonal vaginalis

Gonorrhea
Chlamydia trachomatis
S/philis
Bacterial vaginosis

Candida albicans
Mucopurulent cervicitis

Cervical cell changes
Urinary tract infection

Severe anemia
Mild anemia
Chronic energy deficiency (CED)
CED grade I
CED grade II
CED grade III
Overweight
Obese
Hypertension

Diagnostic criteria

Inflammation (redness) of vaginal canal, with or without discharge
Presence of mucopurulent discharge in the cervix
Redness of cervix on speculum examination
Uterine or adnexal tenderness together with clinical cervicitis. Cervical motion not
tender in all cases. Acute cases had fever; chronic cases had history of fever
Descent of anterior or posterior vaginal wall together with uterine descent
Papanicolaou stain at 10 X 40 = 400 magnification
Gram's stain at 10 X 100 = 1,000 magnification
Leishman's stain at 10 X 100 = 1,000 magnification
VDRL/KAHN (antigen, toxin) Rapid Plasma Reagin test
Presence of clue cells on Papanicolaou stain at 10 X 40 = 400 magnification
Presence of yeast cells on Papanicolaou stain at 10 X 40 = 400 magnification
Examination of slides on Gram’s and Papanicolaou stain. The cases where
numerous polymorphs of both viable and degenerated eutrophils with a background
of fibrous material were seen were diagnosed as having the condition
Presence of nuclear atypia
Microscopic examination of centrifuged specimen showing >8 pus cells/High Power
Field
Hemoglobin < 10mg/dl
Hemoglobin < 12mg/dl

Body mass index (BMI) = Weight in kg/height in meters squared
BMI>17<18.5
BMI >16 <17
BMI <16
BMI 25-30
BMI > 30
Diastolic blood pressure 2i 90mm hg

little more than one-tenth (11 percent) were Muslim, 1
percent were Christiar, and the remaining 88 percent
were Hindu. Approximately one-half (56 percent) had
never gone to school, about one-third had more than
six years of schooling, and the remaining 11 percent had
finished schooling at the primary level (five years). The
age distribution of the sample was as follows: 12 per­
cent were younger than 20, 44 percent were between
20 and 24 years of age, and the remaining 44 percent
were older than 25. Regarding parity, 18 percent had
had only one pregnancy, 58 percent had had between
two and three pregnancies, and the remaining 24 per­
cent had had four or more pregnancies. Exactly onehalf were not using contraceptives at the time of their
interviews, about two-fifths (38 percent) were sterilized,
and the remaining 12 percent were using an intrauter­
ine device (IUD). A little more than one-tenth (13 per­
cent) were pregnant and 8 percent were lactating at the
time of the study.
Complaints Reported During Interviews

A total of 152 women reported 226 gynecological com­
plaints to the social worker who interviewed them.
Thus, the average number of complaints is 1.5 among
those reporting any complaint. Fewer than one-fourth

(23 percent) of the women complained of excessive
weakness. The details of complaints reported are shown
in Table 2. The major gynecological complaints men­
tioned were vaginal discharge with bad odor or itch­
ing or irritation (22 percent), lower abdominal pain or
vaginal discharge with fever (16 percent), and menstrual
bleeding disorders or painful menstruation or spotting
(15 percent).
Although the questions used to elicit symptoms
were identical to those administered earlier to the larger

Table 2 Percentage of 385 women who reported
gynecological and related symptoms to the social worker, by
type of complaint, Karnataka, India, 1994
Type of complaint

Percent

Feeling of excessive weakness, tiredness, or
breathlessness while performing normal household duties
White discharge from vagina with
bad odor or itching/irritation
Lower abdominal pain or vaginal discharge with fever
Menstrual problems (heavy/light/ irregular
bleeding or painful menstruation or spotting
between periods)
Abnormally frequent urination or burning sensation
while passing urine, during the last three months
Pain or bleeding while passing stools
Some mass/swelling coming out of vagina
or leakage of urine when coughing or sneezing
Constant leakage of feces from vagina

Volume 28

22.9
21.9
16.1

15.4

2.3
2.0
0.5
0.5

Number 2 June 1997 97

sample of 3,600 women (Bhatia and Cleland, 1995a),
some marked differences as well as some similarities
are found in the reported prevalence of symptom cat­
egories. The prevalence of symptoms of anemia and
lower reproductive tract infections was broadly simi­
lar in the two inquiries, but menstrual problems and symp­
toms of possible pelvic inflammatory disease (lower ab­
dominal pain or vaginal discharge with fever) were
more commonly reported in the second, smaller inquiry.
One possible reason for these discrepancies concerns dif­
ferences in sample composition. Selection of respondents
for the smaller 12-month longitudinal study that culmi­
nated in clinical examinations was restricted to those
having infants between six and 12 months of age, where­
as the larger parent sample comprised respondents
whose youngest child was younger than five. A sec­
ond reason concerns the context of the two sets of in­
terviews. In the second inquiry, respondents were aware
of the invitation to undergo a medical examination on
the following day. This factor may have influenced their
propensity to report symptoms of ill health.

The Medical Examination

As Table 4 indicates, the gynecological examinations
revealed that 13 percent and 24 percent of the women
had clinical signs of vaginitis and cervicitis, respec­
tively. Cervical ectopy was diagnosed in 10 percent, pel­
vic inflammatory disease in 11 percent, and genital pro­
lapse in 3 percent of the women. Altogether, two out of
five women had at least one clinically diagnosed repro­
ductive tract infection.
The Laboratory Investigations

The results of laboratory tests showed that approxi­
mately one-tenth of the women suffered from sexually
transmitted diseases: trichomonal vaginalis (8 percent),
gonorrhea (1 percent), chlamydia trachomatis (1 per­
cent) and syphilis (2 percent) (see Table 5). More than
one-half of the women were found to have endogenous
infections: bacterial vaginosis (18 percent), Candida al­
bicans (5 percent), mucopurulent cervicitis (37 percent),
and cervical cell changes (1 percent). In addition, 7 f
cent of the women had urinary tract infections.

Menstrual History
As shown in Table 3, a careful and detailed menstrual
history of the women taken by the gynecologist yielded
a far higher estimate of menstrual problems than was
obtained by the interview with the social worker. About
two-thirds of the women reported one or more men­
strual problems. The problems reported included ir­
regular periods (9 percent), long or short menstrual
cycles (9 percent), long or short periods (20 percent),
profuse discharge (15 percent), and passing of clots (8
percent). More than one-half of the women reported
painful menstruation (dysmenorrhea).
Table 3 Percentage of 385 women who reported menstrual
problems to the gynecologist, by type of problem, Karnataka,
India, 1994
______

Percent

Confidence Interval
(95% confidence level)

Irregular periods
Long-cycle (>5 weeks)
Short-cycle (<3 weeks)
Long-lasting period (>5 days)
Short-lasting period (<3 days)
Profuse discharge
Scanty discharge
Passes dots
Dysmenorrhea: backache
Dysmenorrhea: abdominal pain
Premenstrual dysmenorrhea
Menstrual dysmenorrhea
Postmenstrual dysmenorrhea
Any dysmenorrhea
Any menstrual problem

9.4
7.5
1.1
16.0
3.7
14.7
6.7
8.3
36.6
15.5
5.9
5.9
0.8
54.5
62.3

[6.5, 12.31
[4.9, 10.1)
[0.6,2.11
[12.3,19.7]
[1.8.5.6]
[11.2, 18.2]
[4.2, 9.2]
[5.5,11.1]
[31.8,41.4]
[11.9, 19.1]
[3.5, 8.3]
[3.5, 8.3)
[0.0,1.7]
[49.5, 59.5]
[57.5, 67.1)

Menstrual problem

98

Studies in Family Planning

Nutritional Status
As shown in Table 6, the results of the hemoglobin tests
indicated that approximately two-thirds (67 percent) of
the women were mildly anemic (hemoglobin <12gm/dl)
and 17 percent were severely anemic (hemoglobin
<10gm/dl). A body mass index (BMI) was generated
from height and weight measurements of the women.
Chronic energy deficiency (CED) was defined using the
criteria developed by the United Nations Food and Agri­
culture Organization (Shetty and James, 1994). CED
grade I (mild), II (moderate), and III (severe) w$s found
in 29 percent, 16 percent, and 12 percent of the cases, re­
spectively. Only 4 percent of the women were found to
be overweight or obese. In addition, blood-pressure nr
surements indicated that 2 percent were hypertensive.

Table 4 Percentage of 385 women with clinically diagnosed
gynecological morbidity, by type of morbidity, Karnataka,
India, 1994
Type of morbidity____________

Percent

Confidence interval
(95% confidence level)

Vaginitis
Cervicitis
Cervical ectopy
Pelvic inflammatory disease
Genital prolapse
Fistula
Dyspareunia
Any clinically diagnosed morbidity

13.4
23.9
10.0
10.7
3.4
0.3
1.5
39.8

[9.9,16.8)
[19.6, 28.2]
[7.0,13.0)
[7.6, 13.8]
[1.6, 5.2]
[0.0, 0.8]
[0.3, 2.7]
[34.9, 44.7]

Table 5 Percentage of 385 women diagnosed with
gynecological morbidity identified by laboratory tests, by type
of morbidity, Karnataka, India, 1994
Type of morbidity
Any sexually transmitted disease
Trichomonal vaginalis

Gonorrhea
Chlamydia trachomatis
Syphilis
Any endogenous infection
Bacterial vaginosis
Candida albicans
Mucopurulent cervicitis
Cervical cell changes
Urinary tract infection

Percent

10.3
7.5
0.8
0.5
1.5
53.9

18.2
5.2
36.6
1.0
6.5

Confidence Interval
(95% confidence level)

[7.3,13.3]
[4.9, 10.1]
[0.0,1.7]
[0.0,1.4]
[0.3,2.7]
[48.9, 58.9]
[14.3, 22.1]
[3.0, 7.4]
[31.8,41.4]

[0.0, 2.0]
[4.0, 8.9]

Predictors of Gynecological Morbidity and
Nutritional Status
For the purpose of identifying the predictors or corre­
lates, gynecological conditions and nutritional status
have been grouped into eight broad categories: any
menstrual problem (as elicited by the gynecologist); any
dysmenorrhea; clinically diagnosed vaginitis; any clini­
cally diagnosed reproductive tract infection; any labo­
ratory-detected sexually transmitted disease; labora­
tory-detected vaginitis (bacterial vaginosis or Candida
albicans); severe anemia; and any chronic energy defi­
ciency. In order to assess the relationships of selected
socioeconomic, demographic, and cognitive variables to
each of these broad categories, a logistic regression was
performed. The dependent variables were dichotomized
as the presence or absence of each of these conditions.
The independent or predictor variables selected for
the analysis fall into four main groups, similar to those
used in the analysis of self-reported morbidity (Bhatia
and Cleland, 1995a). The first group, representing the
social and economic backgrounds of the women, in­
cludes rural-urban residence, religion, economic status,
and years of schooling. The second group of factors.

Table 6 Percentage of 385 women observed to have
conditions related to nutritional problems and hypertension,
by type of condition, Karnataka, India, 1994
Condition

Anemia or chronic energy deficiency
Mild anemia: hemoglobin <12 gm/dl
Severe anemia, hemoglobin clOgm/dl
CEDgradel (BMI > 17 < 18.5)
CED grade II (BMI > 16 < 17)
CEO grade III (BMI < 16)
Overweight (BMI 25-30)
Obese (BMI > 30)
Hypertension

Percent

88.1
67.4

16.8
28.7

15.8
12.2

2.4
1.3
2.3

Confidence Interval
(95% confidence level)
[84.9,91.3]
[62.7, 72.1]
[13.1,20.5]
[24.2, 33.2]
[12.1, 19.4]
[8.9,15.5]
[0.9, 3.9]
[0.1,2.4]
[0.8,3.8]

which includes age of the woman, number of pregnan­
cies, and history of abortions, is demographic in nature.
The third group comprises current contraceptive sta­
tus, which is of interest because other studies have sug­
gested a link between certain contraceptive procedures
and infection of the reproductive tract. Some proce­
dures, if not properly performed, could cause infections,
and also some reproductive tract infections are con­
traindications for the use of certain contraceptives. The
last group of independent variables consists of cogni­
tive and behavioral factors. For each respondent, scores
on the following dimensions were derived: household
and environmental sanitation, personal hygiene, and ex­
posure to health education. (For the definition of these
variables, together with that of economic status, see
Bhatia and Cleland, 1995a.)
The analysis was carried out in two stages. First, in
a bivariate logistic regression, unadjusted odds ratios
and associated p-values were calculated for each of the
above gynecological and related conditions and indi­
vidual background variables. In the second stage, five
independent variables (rural-urban residence, years of
schooling, economic status, number of pregnancies, and
current use of contraceptives) were selected, and a mul­
tiple logistic regression was performed to estimate the
net relationship of each of these factors to the preva­
lence of gynecological conditions and nutritional sta­
tus. All variables were grouped or were categorical in
nature; for each variable, one category was selected as
the reference category. 'The results of the bivariate analy­
sis are shown in Table 7.
Women residing in town, those with six or more
years of schooling, and those having four or more preg­
nancies were found to be significantly more likely to
report menstrual problems than were women in the cor­
responding reference categories. With regard to painful
menstruation, the probability of reporting the problem
was found to be higher among urban women and among
women younger than 20. The prevalence of dysmenorrhea
was significantly higher among women who had been
sterilized than among those not using contraceptives.
The prevalence of clinically diagnosed vaginitis and
reproductive tract infections showed little significant
variation. Indeed, the only statistically significant as­
sociation was between clinically diagnosed RTIs and
number of pregnancies. The odds of having this condi­
tion among women with only one pregnancy were about
50 percent less than among those with two to three preg­
nancies.
Regarding laboratory-detected infections, sexually
transmitted diseases were found to be significantly low­
er among Muslim women than among Hindus (Fisher's
Exact Test p<0.01). In fact, not a single case was detected

Volume 28

Number 2 June 1997 99

Table 7 Bivariate logistic regression of gynecological and related problems, by selected background characteristics of 385
women, according to type of problem, Karnataka, India, 1994_________________________
Characteristic
.....................

Residence
Urban
Rural (r)
Religion
Hindu
Muslim (r)
Education (years)

0(0
1-5
6+
Economic status
Low(r)
Fair
Age
<20
<20-24 (r)
25+
Pregnancies
1
2- 3 (r)
4+
Abortions
No(0
Yes
Current use of contraceptives
IUD
Tubectomy
Not using (r)
Environmental sanitation
Poor(r)
Fair
Personal hygiene
Poor (r)
Fair
Exposure to health education
Yes (r)
No___________________

(N)

Any
Clinically
menstrual
diagnosed
. .
-Dysmenorrhea
vaginitis
problem

Laboratorydetected STD

Laboratorydetected
vaginitis

Severe
anemia

Any chronic
energy
deficiency

(291)
(94)

1.95’
1.00

1.64*
1.00

1.32
1.00

1.08
1.00

0.92
1.00

1.86**
1.00

0.66
1.00

0.33*
1.00

(339)
(46)

1.07
1.00

1.17
1.00

1.79
1.00

1.43
1.00

1.00

0.92
1.00

1.17
1.00

1.18
1.00

(214)
(45)
(126)

1.00
1.47
1.70*

1.00
1.72
1.10

1.00
0.99
0.96

1.00
1.17
1.11

1.00
0.96
1.02

1.00
1.65
1.04

1.00
0.71
0.50**

1.00
1.41
0.40*

(192)
(193)

1.00
0.84

1.00
0.92

1.00
0.90

1.00
1.15

1.00
1.45

1.00
0.84

1.00
0.92

1.00
0.65*

(47)
(168)
(170)

2.10
1.00
1.43

2.10*
1.00
1.35

2.18
1.00
1.40

1.49
1.00
0.95

0.32
1.00
0.75

1.62
1.00
1.06

0.98
1.00
0.68

0.77
1.00
1.09

(70)
(223)
(92)

1.30
1.00
1.90*

0.75
1.00
1.24

0.74
1.00
0.72

0.55*
1.00
1.11

1.67
1.00
1.43

0.38**
1.00
0.60

0.92
1.00
1.31

0.68
1.00
1.41

(351)
(34)

1.00
1.53

1.00
1.85

1.00
1.20

1.00
1.22

1.00
1.96

1.00
0.60

1.00
1.09

1.00
0.69

(46)
(146)
(193)

1.24
1.35
1.00

1.31
1.62*
1.00

0.87
0.87
1.00

0.94
1.22
1.00

1.16
0.81
1.00

0.68
1.97**
1.00

0.69
0.70
1.00

0.74
1.10
1.00

(200)
(185)

1.00
1.05

1.00
1.35

1.00
1.59

1.00
1.22

1.00
1.66

1.00
1.00

1.00
0.85

1.00
0.67*

(207)
(178)

1.00
0.91

1.00
1.20

1.00
1.08

1.00
0.87

1.00
1.16

1.00
1.21

1.00
1.11

1.00
0.94

(204)
(181)

1.00
1.00
1.22 __________ 1.42

1.00
0.88

1.00
0.73

1.00
1.24

1.00
0.92

1.00
0.84

1.00
1.08

Significant at *p<0.05; ’•psO.OI; •••psO.OOI.

(r) = reference category.

among women of the Muslim faith. Surprisingly, no ur­
ban-rural or socioeconomic variations are found in lev­
els of STD infection in this sample. Laboratory-deter­
mined vaginitis was observed to be significantly higher
among urban women and among those who had under­
gone sterilization than among their counterparts. In ad­
dition, the odds of a woman with only one pregnancy
having laboratory-detected vaginitis were only about
one-third of those among women with two or three
pregnancies.
The analysis of nutritional status indicates that the
odds of anemia and chronic energy deficiency occur­
ring among women with six or more years of educa­
tion were about 50 percent lower than among those with
no schooling. Chronic energy deficiency was signifi­
cantly lower among urban women, and somewhat lower
among those of higher economic status, and among those
with good environmental sanitation.

100

Any
Anyclinically
clinically
diagnosed RTI

Studies in Family Planning

•Fisher’s Exact Test (two tail) p< 0.01.

The results of the multiple logistic regression are
shown in Table 8. The net association between urban
residence and the reporting of menstrual problems and
dysmenorrhea is found to be statistically significant.
In addition, a significantly higher proportion of ster­
ilized women had complaints of painful menstruation
than did those who were not practicing contraception.
In the multivariate analysis, no characteristic of the
women included in the analysis was significantly asso­
ciated with clinically diagnosed vaginitis or RTIs, or
with laboratory-detected STDs. However, laboratorydetected vaginitis (bacterial vaginosis and Candida
albicans) was significantly higher among urban and
sterilized women. The prevalence of anemia and chronic
energy deficiency remained significantly lower among
women with six or more years of education. Chronic
energy deficiency was also significantly lower among
urban women.

Table 8 Multiple logistic regression of gynecological and related problems, by selected background characteristics of 385
women, according to type of problem, Karnataka, India, 1994
Characteristic

Any
Clinically
menstrual
diagnosed
problem Dysmenorrhea
vaginitis

Residence
Urban
Rural (r)
Education (years)
0(r)
1- 5
6+
Economic status
Low(r)
Middle
Pregnancies
1
2- 3 (r)
Current use of contraceptives
IUD
Tubectomy
Not using (r)
Constant
Significant at ’p<0.05; ”p<0.01: ’”p<0.001.

Any clinically
diagnosed RTI

Laboratory­
detected STD

Laboratorydetected
vaginitis

Severe
anemia

Any chronic
energy
deficiency

1.95*
1.00

1.64*
1.00

1.34
1.00

1.11
1.00

1.13
1.00

1.98”
1.00

0.82
1.00

0.43’”

1.00
1.00
0.76

1.00
1.86
0.95

1.00
1.03
0.95

1.00
1.23
1.13

1.00
0.72
1.04

1.00
1.37
0.85

1.00
0.74
0.55”

1.00
1.62
0.48”

1.00
0.73

1.00
0.87

1.00
0.91

1.00
1.05

1.00
1.79

1.00
0.88

1.00
0.93

1.00
1.20

0.68
1.00
1.26

0.71
1.00
1.40

0.46
1.00
0.77

0.60
1.00
1.15

1.38
1.00
0.74

0.52
1.00
0.69

0.84
1.00
1.27

0.78
1.00
1.49

1.61
1.17
1.00
0.72

1.33
1.67*
1.00
-0.41

0.78
0.77
1.00
-1.55

0.89
1.08
1.00
-0.42

0.60
0.76
1.00
-2.62

0.56
1.82’
1.00
-1.36

0.84
0.68
1.00
1.16

0.93
1.05
1.09
0.53

1.00

(r) = reference category.

Gynecological Conditions and Nutritional Status

Several women in the-study were found to have more
than one gynecological condition and low nutritional
status. Therefore, an examination of the interrelation­
ships between different conditions is of interest. The
likelihood that a woman with a clinically diagnosed re­
productive tract infection would have a laboratory-con­
firmed sexually transmitted disease is about three times
higher than that for women with no clinically diagnosed
RTI (not shown). However, no statistically significant
relationship was observed between anemia, chronic en­
ergy deficiency, and the gynecological conditions in­
cluded in the analysis. Even when severe anemia was
considered, the relationship between this condition and
the other seven broad categories of conditions did not
reach statistical significance. However, severe chronic
energy deficiency was observed to be significantly re­
lated to clinically diagnosed RTIs. Women with grade
ID CED were two times more likely to have a clinically
diagnosed reproductive tract infection than were those
not having this condition.
An attempt was also made to analyze relationships
between specific conditions included under each of the
broad categories. The results of this analysis indicate
that women with clinically diagnosed upper reproduc­
tive tract infections or pelvic inflammatory disease are
three times more likely than those who are not so diag­
nosed to report menstrual problems. Furthermore,
women having a laboratory-detected sexually transmit­
ted disease have a three times greater risk of having PID
than do women with no detected STDs.

Discussion
India is such a huge and varied country that no single,
localized study can lay claim to representing the whole.
Compared with many other states, Karnataka is ad­
vanced. Adult literacy is 56 percent compared with the
national figure of 52 percent, and infant mortality stands
at about 65 deaths per 1,000 live births, compared with
the all-India average of 78 deaths. In terms of health
and family planning services, the state is also progres­
sive. According to the 1992-93 National Family Health
Survey, 38 percent of deliveries in Karnataka are insti­
tutional, compared with 26 percent for the entire coun­
try. Similarly, contraceptive practice among married
couples is relatively high at 49 percent, compared with
an all-India estimate of 41 percent.
The site of the study was selected as being typical
of rural Karnataka and sufficiently far from Bangalore,
the capital city, to be shielded from the influence of
this rapidly growing, modern metropolis. The area is
predominantly agricultural, although it contains one
town of some 47,000 inhabitants. The study sample was
drawn from the town and from 48 villages, represent­
ing a distinct advantage over the two earlier population­
based inquiries into gynecological morbidity (Bang, 1989
and Younis, 1993), both of which were restricted to two
villages and included no urban areas.
One of the main lessons of the present study con­
cerns methods rather than substantive findings. The in­
corporation of intrusive clinical procedures into field
studies is usually plagued by noncompliance. In this in­
stance, more than 90 percent of respondents agreed to

Volume 28

Number 2

June 1997

101

vaginal and other forms of examination. This exception­
study were bacterial vaginosis and mucopurulent cer­
ally high rate of compliance came about for several rea­
vicitis. Medical opinion differs on the implications of
sons that have implications for the design of future stud­
these conditions, but some evidence exists that un­
ies. One reason concerns the familiarity of respondents
treated bacterial vaginosis may be a risk factor for pel­
with field investigators. By the time they were invited
vic inflammatory disease (Moi, 1990). However, approx­
for clinical examination, the women had been visited by
imately one-fourth of the women had clinical evidence
lay interviewers at monthly intervals over the span of a
of PID, cervical ectopy, and fistula. These conditions
year. Inevitably, a high degree of trust and rapport had
may represent a serious long-term threat to the health
developed. A further critical factor was proximity. A tem­
and fecundity of women. The contribution of STDs to
porary clinic was set up successively in each of the study
the high prevalence of reproductive tract infections in
villages and in the town, an arrangement that minimized
this sample appears to be modest. Evidence of STD in­
disruption of the respondents' daily lives.
fections was found in 10 percent of the women stud­
One of the purposes of the study was to assess child
ied, a proportion that appears to be much lower than
health, although results on this topic are not presented
that in the Maharashtra study, which may reflect dif­
here. Women were, therefore, invited to the clinic for
ferent sexual mores. A preliminary analysis of risk fac­
an assessment both of their child's and of their own
tors for RTIs found a significant link between tubectomy
health. In addition to the gynecologist, a pediatrician
and laboratory-confirmed vaginitis. A similar result was
attended all clinics. This emphasis on child health was,
obtained in Bangladesh (Wasserheit et al., 1989). In ad­
most likely, an important influence on attendance. Last­
dition, women who had undergone tubal ligation were
ly, treatment was offered both to women and to children.
much more likely to report painful menstruation. Al­
The relative contribution of these factors to the achieve­
though the etiology of these associations is uncertain,
ment of high compliance is impossible to assess, but the
the results potentially have far-reaching implications
experience does prove that problems of noncompliance
for India's family planning program, which has always
can be overcome by careful study design.
placed a primary emphasis on permanent methods
As noted above, this study collected information on
of birth control. Further biomedical investigation is a
gynecological and related conditions based on women's
priority.
perceptions and self reports within the context of a larger
A persistent urban-rural difference in the report­
cross-sectional study. (For detailed results, see Bhatia
ing of menstrual problems and dysmenorrhea and in
and Cleland, 1995a.) A comparison of estimates from
laboratory-confirmed vaginitis was also found. How­
different data-gathering methods is in progress. It will
ever, the prevalence of chronic energy deficiency is much
provide valuable information and will have larger im­
lower among urban women. Women's lack of school­
plications for determining the appropriateness of ser­
ing was unrelated to most gynecological conditions, but
vices that could be provided for screening for RTIs and
was strongly associated with anemia and CED. That the
other gynecological complaints.
lack of education is more strongly predictive of these
The single most important finding of this study, re­
conditions than economic status is yet another testi­
gardless of the method of assessment, is the high preva­
mony to the pervasive influence of this factor on health
lence of reproductive tract infections. By clinical crite­
and welfare.
ria, 36 percent of the women studied were diagnosed
The study has also highlighted the extreme diffi­
as having such an infection. The laboratory tests re­
culty of collecting information on menstrual disorders.
vealed evidence of reproductive tract infections in 56
When interviewed by experienced nonmedical staff, 15
percent of cases. The self-reports yielded a somewhat
percent of the respondents reported a menstrual prob­
lower estimate of symptoms of infection, suggesting that
lem. Under the more extensive probing of the gynecolo­
many infections are asymptomatic or produce symp­
gist, this estimate rose to 62 percent. The Maharashtra
toms that are regarded by women as normal. Some 29
study reported a “similar experience: 20 percent of re­
percent of respondents reported symptoms associated
spondents reported menstrual problems to a nonmedi­
with upper or lower reproductive tract infection. Esti­
cal interviewer and 58 percent reported them to a phy­
mates of infection in this study are much lower than
sician (Bang et al., 1989). Clearly, the propensity of wom­
those found in two villages in Maharashtra (Bang et al.,
en to report menstrual problems is highly sensitive to
1989), but are broadly similar to those found in rural
the context of the inquiry and, until more reliable meth­
Bangladesh (Wasserheit et al., 1989). The Bangladesh
ods of obtaining this type of information are devised,
study, however, was confined to symptomatic women,
little gain can come of adding menstrual problems to
which complicates comparison.
other conditions in the attempt to form total estimates
The two most common conditions found in this
of gynecological morbidity.

102

Studies in Family Planning

A final important result, albeit one that concerns the
general health of women rather than gynecological dis­
orders specifically, is the high level of anemia and
chronic energy deficiency found during the study. Al­
though Karnataka is a relatively advanced state, a large
majority of the sample were suffering from these dis­
orders. A study conducted by the Indian Council of
Medical Research (ICMR) in six centers in India during
1986-87 also found that 62 percent of pregnant wom­
en had a hemoglobin level of less than 1 Igm/dl (ICMR,
1992). This combination of widespread undemutrition
and malnutrition and reproductive tract infections in­
dicates the enormity of the task ahead, if radical im­
provements in women's health are to be achieved in In­
dia. The translation of the Cairo rhetoric into reality
clearly will require innovation, expense, and improved
diagnostic tools. Above all, it will require political will.

Indian Council of Medical Research Field Supplementation. (ICMR).
1992. Trial in Pregnant Women. An ICMR Task Force Study. New
Delhi: ICMR.
Ishizuka,
K. Oota, and K. Masyvycgu. 1972. Practical Cytodiagnosis. Philadelphia: J.B. Lippincott.

Jeffcoate, Norman. 1982. Principles of Gynaecology. London: Butter­
worth Scientfic.

Koss, Leopold G. 1968. Diagnostic Cytology and Its Histopathologic
Bases. Philadelphia: J.B.Lippincott.

Moi, H. 1990. "Prevalence of bacterial vaginosis and its association
with genital infections, inflammation, and contraceptive meth­
ods in women attending sexually transmitted diseases and pri­
mary health clinics." International Journal of STD and AIDS 1,2:
86-94.
Naib, Zuher M. 1970. Exfoliative Cytopathology. Boston: Little & Brown.

Shetty, P.S and W.P.T. James. 1994. Body Mass Index—A Measure of
Chronic Energy Deficiency in Adults. New York: United Nations
Food and Agriculture Organization.

United Nations. 1994. Report of the International Conference of Popula­
tion and Development, Cairo. Document A.Conf. 171/13.

References
Bah, P. and R.A. Bhujwela. 1969. "A pilot study of clinic-epidemio­
logic investigations of vaginal discharge by rural women." In­
dian Journal of Medical Research, 57, 12: 2,289-2,299.

Bang, R.A., A.T. Bang, M. Batule, Y. Choudhary, S. Sarmukaddam,
and O. Tale, 1989. "High prevalence of gynecological diseases
in rural Indian women." Lancet 8,629,1: 85-88.
Bhatia, Jagdish C. and John Cleland. 1995a. "Self-reported symptoms
of gynecological morbidity and their treatment in South India."
. Studies in Family Planning 26,4: 203-216.

-------- . 1995b. "Determinants of maternal care in a region of South
India." Health Transition Review 5: 127-142.
-------- . 1996. "Obstetric moibidity in South India: Results from a
community survey." Social Science and Medicine 43,10:1,507-1,516.
Brabin, L., J. Kemp, O.K. Obunge, J. Ikimalo, N. Dollimore, N. Odu,
C.N. Hart, and N.D. Briggs. 1995. "Reproductive tract infections
and abortion among adolescent girls in rural Nigeria." Lancet 345:
300-304.

Dewhurst, John (ed.). 1981. Integrated Obstetrics and Gynaecology for
Postgraduates. London: Blackwell Scientific Publications.

Howkins, John and Gordon Bourne (eds.). 1985. Shaw's Text Book of
Gynaecology. Delhi: BI Publications.

Wahi, P.N., U.K. Luthra, S. Mali, and M.B. Slijinkin. 1972. "Preva­
lence and distribution of cancer of uterine cervix in Agra dis­
trict in India." Cancer 30: 720-25.

Wasserheit, Judith N., Jeffrey R. Harris, J. Chakraborty, Bradford A.
Kay, and Karen J. Mason. 1989. "Reproductive tract infections
in a family planning population in rural Bangladesh." Studies in
Family Planning 20,2: 69-80.
Younis, Nabil, Hind Khattab, Huda Zurayk, Mawaheb El-Mouelhy,
Mohamed Fadle Amin, and Abdel Moneim Farag. 1993. "A com­
munity study of gynecological and related morbidities in rural
Egypt." Studies in Family Planning 24/3:175-186.

Acknowledgments
The authors would like to note the able assistance provided
by M.S. Sanath Kumar and S. Ramaswamy in data process­
ing. We are also thankful to Jinka Subramanya and J. Vasudev
Rao, pathologists at Bangalore Hospital, for their hard and
diligent laboratory work. Chris Elias of the Population Coun­
cil, Thailand, provided valuable comments on a draft of the
paper. Financial support for the study was provided by the
Ford Foundation, which is gratefully acknowledged. The
analysis of the data was partly funded by the Overseas De­
velopment Administration of the United Kingdom.

Volume 28

Number 2 June 1997

103

Estimates and Explanations of Gender
Differentials in Contraceptive Prevalence Rates
Alex Chika Ezeh and Gora Mboup

This article examines gender differentials in the reporting of contraceptive use and offers explanations
regarding the sources of these differences. Datafromfive countries where DHS surveys were conducted
recently among men and women are used in exploring these differences. The gap exists in all five
countries, with men (or husbands) reporting greater practice ofcontraception than women (or wives).
Results from the bivariate analysis suggest that the gap is attributable to polygyny and to gender
differences in how the purpose of contraception is understood, rather than to male extramarital sexual
relations. Additionally, gender differences in the definition ofcertain contraceptive methods and differences
in the interpretation ofquestions about contraception contribute to the observed gap. Thesefindings
are also consistent ivith results of the multivariate analysis. (Studies in Family Planning 1997; 28,2:
104-121)

The disparity between men and women (or husbands
and wives) in their reporting of contraceptive use has
long been documented in the demographic literature
(Yaukey et al., 1965; Stoeckel and Choudhury, 1969;
Green, 1969; Koenig et al., 1984; Hopflinger and Kuhne,
1984; Mitra et al., 1985). The nature of this discrepancy
is also well established, with men or husbands gener­
ally reporting greater use of contraceptives than do
women or wives.1 In a recent study, Ezeh and his col­
leagues (1996) found current contraceptive use to be
higher among currently married men compared with
currently married women in 12 of the 14 countries they
studied. In some of the countries in which men reported
greater contraceptive use than women did, current use
among men was found to be twice the level reported
by women. However, a difference of only two percent­
age points separates the reports of the men and the
women in the two countries where women reported
greater contraceptive use than men did. Most research­
ers have treated observed inconsistencies in contracep-

Alex Chika Ezeh is Research Associate, Applied Research
and Development and Gora Mboup is Country Monitorfor
francophone Africa, Demographic and Health Surveys,
Macro International, 11785 Beltsville Drive, Calverton, MD
20705.

104

Studies in Family Planning

tive-use status of husbands and wives as respondentrelated errors, and in particular, as underreporting by
wives (Koenig et al., 1984).
The treatment of inconsistencies in the reporting of
contraceptive use between married partners as misre­
porting by one partner assumes that spouses use con­
traceptives exclusively with their marital partners. An
extension of this assumption is that sexual activity
among married men and women occurs exclusively
within marriage. The treatment of wives' reports as misor underreporting assumes that spouses have equal
knowledge of the use of a method and that both part­
ners have the same definition of what constitutes con­
traceptive practice. In the paper on India by Koenig et
al. (1984), 67 percent of spouses in couples where at least
one partner reported use of contraceptives differed in
their response to the question on current use. Of these,
more than 84 percent of the discrepancies occurred
when husbands but not wives reported use of a method.
Three methods: condoms, periodic abstinence, and va­
sectomy, account for 79 percent of the absolute differ­
ence between the reports of husbands and wives and
for more than 93 percent of the difference associated
with only husbands reporting use of a method (see
Table 1 in Koenig et al., 1984). Except for prolonged ab­
stinence, a husband can use these methods without his
wife's knowing. Even when the wife knows, she may
not associate the use with contraception. Husbands may

Levels and Determinants of Gynecological
Morbidity in a District of South India
Jagdish C. Bhatia, John Cleland, Leela Bhagavan, and N.S.N. Rao

This article presents the results of an assessment of gynecological morbidity among 385 women
with young children residing in a district of Karnataka State, South India. All three main modes of
assessment (clinical examination, laboratory tests, and self-reports) reveal a high burden of
reproductive tract infections. The two most common conditions, identified by laboratory tests, were
bacterial vaginosis and mucopurulent cervicitis. Approximately one-fourth of the women had clinical
evidence of pelvic inflammatory disease, cervical ectopy, and fistula. The contribution of sexually
transmitted diseases to overall gynecological morbidity appears to be relatively modest; 10 percent
were so diagnosed. Associated conditions of anemia and chronic energy deficiency were common.
Severe anemia was found in 17 percent of cases and severe chronic energy deficiency in 12 percent.
These results indicate that radical improvements in women's health in India will require far more
than the diagnosis and treatment of reproductive tract infections. (Studies in Family Planning
1997; 28,2:95-103)

An ambitious program of action to make reproductive
health services universally available was drawn up at
the International Conference on Population and Devel­
opment held at Cairo, Egypt, in September 1994. Repro­
ductive health has been defined by the World Health Or­
ganization as a "state of complete physical, mental and
social well being and not merely the absence of disease
or infirmity, in all matters relating to the reproductive
system and to its functions and processes" (UN, 1994).
Such a broad definition does not lend itself readily
to the identification of priorities, but many experts
would agree that the three main dimensions of repro­
ductive health among adult women are: avoidance of
unwanted pregnancies, safe motherhood, and protec­
tion against infections and dysfunctions of the repro­
ductive tract, including sexually transmitted diseases.

Jagdish C. Bhatia is Professor of Health Management, Indian
Institute of Management, Bannerghatta Road, Bangalore,
560 076 India. John Cleland is Professor of Medical
Demography, London School of Hygiene and Tropical
Medicine, London, U.K. Leela Bhagavan is Obstetrician and
Gynecologist, Bangalore Hospital. N.S.N Rao is Consultant
in Biostatistics, Ford Foundation Project, Indian Institute of
Management, Bangalore.

$

The first dimension has been thoroughly researched, par­
ticularly in developing countries, and a growing body
of evidence exists on obstetric morbidity and mortality.
However, the third dimension has been badly neglected.
Gynecological morbidity can be defined as structural
and functional disorders of the genital tract not related
to pregnancy, delivery, or the puerperium. It includes
menstrual disorders, reproductive tract infections, cer­
vical cell changes, genital prolapse, and such other con­
ditions as syphilis, urinary tract infections, hypertension,
anemia, chronic energy deficiency (CED), and obesity.
Most knowledge in this area is based on hospital
statistics; information available at the community level
in developing countries is rare. A few community-based
studies have focused on specific problems, such as cer­
vical cancer, vaginal discharge, and reproductive tract
infections among selected groups of women (Wahi et
al.,1972; Bali and Bhujwela, 1969; Wasserheit et al., 1989;
Brabin et al., 1995), but only two such studies of over­
all gynecological morbidity have been reported: one
from India (Bang et al., 1989) and the other from Egypt
(Younis et al., 1993). The study in India covered two
predominantly tribal villages, and only 59 percent of
the sampled women participated in the clinical exami­
nation and laboratory investigations. Furthermore, little
information is available about the determinants of gy­
necological morbidity.

Volume 28

Number 2 June 1997 95

I'

More information is urgently needed to assess the
nature and burden of gynecological morbidity, if the
rhetoric of the Cairo Conference is to be translated into
realistic and cost-effective action. The present study ad­
dresses this need. Its objectives were first to determine
the level of gynecological morbidities and other related
conditions among women in a district of South India
using clinical, laboratory, and other methods, and sec­
ond, to analyze socioeconomic, demographic, and cog­
nitive correlates of gynecological morbidities and re­
lated conditions.

Materials and Methods
The study was conducted in the Indian state of Karna­
taka in the southern part of the country as part of a
wider inquiry, funded by the Ford Foundation, into
pathways through which the behavior and characteris­
tics of mothers influence the health and survival of their
children. A sample of 3,600 women (2,400 living in ru­
ral areas and 1,200 in town) was selected for the origi­
nal study. These women were asked about their symp­
toms of obstetric and gynecological morbidity, as well
as about other matters (Bhatia and Cleland, 1995a,
1995b, and 1996). Because the objective of the main
study was to examine the effect of mothers' behavior
on children's health and survival, all women with chil­
dren between the ages of six to 12 months (the most vul­
nerable age group) were included in the prospective
study. In the original sample, 440 women met this cri­
terion and were included in the study. All were mar­
ried and were younger than 35. They were interviewed
at monthly intervals for one year, at which point they
were invited to undergo a medical examination. A gap
of approximately 15 months occurred between the se­
lection of the sample for prospective study and the clini­
cal examination. A total of 23 women were lost to fol­
low-up prior to the clinical phase. Of the 417 women
remaining, 11 refused to participate or were temporarily
absent (this figure includes five women who appeared
at the examination center but refused vaginal exami­
nation), 17 could not be examined because they were
menstruating and did not appear for their rescheduled
appointments, and four women's specimens were re­
jected by the laboratory. Thus full gynecological exami­
nations and laboratory investigations were completed
for 385 women, representing 86 percent of the original
sample and 92 percent of those who completed the 12month prospective phase of the study.
A day before the medical examination was sched­
uled, women were interviewed by female social work­
ers who gathered information about the presence of

96

Studies in Family Planning

symptoms of gynecological problems. This information
was not made available to the gynecologist.
Clinics were held in the project office in town and
at schools in all the villages included in the study. An
experienced female gynecologist visited all these loca­
tions and conducted the examinations over a threemonth period in late 1994. She was assisted by a quali­
fied female technician. The gynecologist took menstrual
histories and anthropometric measurements. A general
medical examination that included blood-pressure mea­
surement and auscultation, and examination of breasts,
adiposity, and external genitalia was performed. Later,
a speculum examination was conducted during which
specimens were taken from the cervix and the poste­
rior fornix. After removal of the speculum, a bimanual
examination was performed to determine if tenderness
was present. A 5-milliliter sample of venous blood was
collected for hemoglobin and Venereal Disease Reference
Laboratory (VDRL) assessments. A midstream sample
of urine was also collected for laboratory assessment
of urinary tract infections. In addition, four slides wei
made from swab specimens of which two were fixed
immediately with propyl alcohol for a pap smear and
two were air dried. These specimens were stored in an
icebox and transported daily to one of the wellequipped laboratories in Karnataka, at Bangalore hos­
pital in the state capital. Because the laboratory where
the samples were sent was 70 to 100 kilometers (43 to
62 miles) distant from the villages where the clinics were
held, and because the time lag between the collection
of samples and their deposit at the laboratory ranged
from three to 12 hours, preparation of wet mounts for
the detection of trichomonal vaginosis and Candida
albicans was difficult.
The gynecological morbidities and related conditions
discussed in this report were measured, therefore, by sev­
eral different data-collection procedures: by means of soc­
ial workers' questions; by the gynecologist's careful ques
tioning, anthropometric measurements, and clinical
examination; and by laboratory investigations. The
precise diagnostic criteria are given in Table 1.
The clinical criteria were in accordance with those
given in standard textbooks of gynecology (Dewhurst,
1981; Jeffcoate, 1982; Howkins and Bourne, 1985). Fur­
ther details of the tests used in the study may be found
elsewhere (Koss, 1968; Ishizuka et al., 1972; Naib, 1970).

Findings
Characteristics of Respondents
One-fourth of the sampled women were residing in
town and the remaining three-fourths in rural areas. A

Table 1

Criteria for diagnosing gynecological morbidity and related conditions, Karnataka, India, 1994
Condition

Clinical
Vaginitis
Cervicitis
Cervical ectopy
Pelvic inflammatory disease (PID)
Genital prolapse

Diagnostic criteria
Inflammation (redness) of vaginal canal, with or without discharge
Presence of mucopurulent discharge in the cervix
Redness of cervix on speculum examination
Uterine or adnexal tenderness together with clinical cervicitis. Cervical motion not
tender in all cases. Acute cases had fever; chronic cases had history of fever
Descent of anterior or posterior vaginal wall together with uterine descent

Laboratory

Tnchomonal vaginalis
Gonorrhea
Chlamydia trachomatis
Syphilis
Bacterial vaginosis
Candida albicans
Mucopurulent cervicitis

Cervical ceil changes
Urinary trac t infection

Papanicolaou stain at 10 X 40 = 400 magnification
Gram’s stain at 10 X 100 = 1,000 magnification
Leishman's stain at 10 X 100 = 1,000 magnification
VDRL/KAHN (antigen, toxin) Rapid Plasma Reagin test
Presence of clue cells on Papanicolaou stain at 10 X 40 = 400 magnification
Presence of yeast cells on Papanicolaou stain at 10 X 40 = 400 magnification
Examination of slides on Gram’s and Papanicolaou stain. The cases where
numerous polymorphs of both viable and degenerated eutrophils with a background
of fibrous material were seen were diagnosed as having the condition
Presence of nuclear atypia
Microscopic examination of centrifuged specimen showing >8 pus cells/High Power

Severe anemia
Mild anemia

Hemoglobin < lOmg/dl
Hemoglobin < 12mg/dl

Chronic energy deficiency (CED)
CED grade I
CED grade II
CED grade III
Overweight
Obese
Hypertension

Body mass index (BMI) = Weight in kg/height in meters squared
BMI>17<18.5
BMI > 16 <17
BMI <16
BMI 25-30
BMI >30
Diastolic blood pressure £ 90mm hg

little more than one-tenth (11 percent) were Muslim, 1
percent were Christian, and the remaining 88 percent
were Hindu. Approximately one-half (56 percent) had
never gone to school, about one-third had more than
six years of schooling, and the remaining 11 percent had
finished schooling at the primary level (five years). The
age distribution of the sample was as follows: 12 per­
cent were younger than 20, 44 percent were between
20 and 24 years of age, and the remaining 44 percent
were older than 25. Regarding parity, 18 percent had
had only one pregnancy, 58 percent had had between
two and three pregnancies, and the remaining 24 per­
cent had had four or more pregnancies. Exactly onehalf were not using contraceptives at the time of their
interviews, about twelfths (38 percent) were sterilized,
and the remaining 12 percent were using an intrauter­
ine device (IUD). A little more than one-tenth (13 per­
cent) were pregnant and 8 percent were lactating at the
time of the study.
Complaints Reported During Interviews

A total of 152 women reported 226 gynecological com­
plaints to the social worker who interviewed them.
Thus, the average number of complaints is 1.5 among
those reporting any complaint. Fewer than one-fourth

(23 percent) of the women complained of excessive
weakness. The details of complaints reported are shown
in Table 2. The major gynecological cojnplaints men­
tioned were vaginal discharge with bad odor or itch­
ing or irritation (22 percent), lower abdominal pain or
vaginal discharge with fever (16 percent), and menstrual
bleeding disorders or painful menstruation or spotting
(15 percent).
Although the questions used to elicit symptoms
were identical to those administered earlier to the larger

Table 2 Percentage of 385 women who reported
gynecological and related symptoms to the social worker, by
type of complaint, Karnataka, India, 1994
Type of complaint

Percent

Feeling of excessive weakness, tiredness, or
breathlessness while performing normal household duties
White discharge from vagina with
bad odor or itching/irr itation
Lower abdominal pain or vaginal discharge with fever
Menstrual problems (heavy/light/ irregular
bleeding or painful menstruation or spotting
between periods)
Abnormally frequent urination or burning sensation
While passing urine, during the last three months
Pain or bleeding while passing stools
Some mass/swelling corning out of vagina
or leakage of urine when coughing or sneezing
Constant leakage of feces from vagina

Volume 28

Number 2

22.9

21.9
16.1

15.4

2.3
2.0
0.5
0.5

June 1997

97

sample of 3,600 women (Bhatia and Cleland, 1995a),
some marked differences as well as some similarities
are found in the reported prevalence of symptom cat­
egories. The prevalence of symptoms of anemia and
lower reproductive tract infections was broadly simi­
lar in the two inquiries, but menstrual problems and symp­
toms of possible pelvic inflammatory disease (lower ab­
dominal pain or vaginal discharge with fever) were
more commonly reported in the second, smaller inquiry.
One possible reason for these discrepancies concerns dif­
ferences in sample composition. Selection of respondents
for the smaller 12-month longitudinal study that culmi­
nated in clinical examinations was restricted to those
having infants between six and 12 months of age, where­
as the larger parent sample comprised respondents
whose youngest child was younger than five. A sec­
ond reason concerns the context of the two sets of in­
terviews. In the second inquiry, respondents were aware
of the invitation to undergo a medical examination on
the following day. This factor may have influenced their
propensity to report symptoms of ill health.

The Medical Examination
As Table 4 indicates, the gynecological examinations
revealed that 13 percent and 24 percent of the women
had clinical signs of vaginitis and cervicitis, respec­
tively. Cervical ectopy was diagnosed in 10 percent, pel­
vic inflammatory disease in 11 percent, and genital pro­
lapse in 3 percent of the women. Altogether, two out of
five women had at least one clinically diagnosed repro­
ductive tract infection.

The Laboratory Investigations

The results of laboratory tests showed that approxi­
mately one-tenth of the women suffered from sexually
transmitted diseases: trichomonal vaginalis (8 percent),
gonorrhea (1 percent), chlamydia trachomatis (1 per­
cent) and syphilis (2 percent) (see Table 5). More than
one-half of the women were found to have endogenous
infections: bacterial vaginosis (18 percent), Candida al­
bicans (5 percent), mucopurulent cervicitis (37 percent'
and cervical cell changes (1 percent). In addition, 7 per­
cent of the women had urinary tract infections.

Menstrual History
As shown in Table 3, a careful and detailed menstrual
history of the women taken by the gynecologist yielded
a far higher estimate of menstrual problems than was
obtained by the interview with the social worker. About
two-thirds of the women reported one or more men­
strual problems. The problems reported included ir­
regular periods (9 percent), long or short menstrual
cycles (9 percent), long or short periods (20 percent),
profuse discharge (15 percent), and passing of clots (8
percent). More than one-half of the women reported
painful menstruation (dysmenorrhea).

Table 3 Percentage of 385 women who reported menstrual
problems to the gynecologist, by type of problem, Karnataka,
India, 1994
Menstrual problem

Irregular periods
Long-cycle (>5 weeks)
Sbort-cycle (<3 weeks)
Long-lasting period (>5 days)
Short-lasting period (<3 days)
Profuse discharge
Scanty discharge
Passes dots
Dysmenorrhea: backache
Dysmenorrhea: abdominal pain
Premenstrual dysmenorrhea
Menstrual dysmenorrhea
Postmenstrual dysmenorrhea
Any dysmenorrhea
Any menstrual problem

98 Studies in Family Planning

Percent

9.4
7.5

1.1
16.0
3.7
14.7

6.7
8.3
36.6
15.5
5.9
5.9

0.8
54.5
62.3

Nutritional Status
As shown in Table 6, the results of the hemoglobin tests
indicated that approximately two-thirds (67 percent) of
the women were mildly anemic (hemoglobin <12gm/dl)
and 17 percent were severely anemic (hemoglobin
<10gm/dl). A body mass index (BMI) was generated
from height and weight measurements of the women.
Chronic energy deficiency (CED) was defined using the
criteria developed by the United Nations Food and Agri­
culture Organization (Shetty and James, 1994). CED
grade I (mild), II (moderate), and III (severe) was found
in 29 percent, 16 percent, and 12 percent of the cases, re­
spectively. Only 4 percent of the women were found to
be overweight or obese. In addition, blood-pressure mea
surements indicated that 2 percent were hypertensive.

Confidence interval
(95% confidence level)

[6.5. 12.3]
[4.9, 10.1]
[0.6. 2.1]
[12.3,19.7]
[1.8, 5.6]
[11.2, 18.2]
[4.2, 9.2]
[5.5,11.1]
[31.8,41.4]
[11.9,19.1]
[3.5, 8.3]
[3.5, 8.3]
[0.0, 1.7]
[49.5, 59.5]
[57.5,67.1]

Table 4 Percentage of 385 women with clinically diagnosed
gynecological morbidity, by type of morbidity, Karnataka,
India, 1994
Typ« of morbidity
Vaginitis
Cervicitis
Cervical ectopy
Pelvic inflammatory disease
Genital prolapse

Percent

Confidence Interval
(95% confidence level)

13.4
23.9

[19.6,28.2]

10.0

Fistula

10.7
3.4
0.3

Dyspareunia
Any clinically diagnosed morbidity

1.5
39.8

[9.9,16.8]
[7.0,13.0]
[7.6, 13.8]
[1.6, 5.2]

[0.0, 0.8]
[0.3, 2.7]
[34.9, 44.7]

Table 5 Percentage of 385 women diagnosed with
gynecological morbidity identified by laboratory tests, by type
of morbidity, Karnataka, India, 1994

which includes age of the woman, number of pregnan­
cies, and history of abortions, is demographic in nature.
The third group comprises current contraceptive sta­
Confidence Interval
tus, which is of interest because other studies have sug­
Type of morbidity
Percent
(95% confidence level)
gested
a link between certain contraceptive procedures
Any sexually transmitted disease
10.3
[7.3,13.3]
TrichomonaJ vaginalis
and infection of the reproductive tract. Some proce­
7.5
[4.9,10.1]
Gonorrhea
0.8
[0.0,1.7]
dures, if not properly performed, could cause infections,
Chlamydia trachomatis
0.5
[0.0,1.4]
and
also some reproductive tract infections are con­
Syphilis
1.5
[0.3, 2.7]
traindications for the use of certain contraceptives. The
Any endogenous infection
53.9
[48.9, 58.9]
Bacterial vaginosis
18.2
[14.3,22.1]
last group of independent variables consists of cogni­
Candida albicans
5.2
[3.0. 7.4]
tive and behavioral factors. For each respondent, scores
Mucopurulent cervicitis
36.6
[31.8,41.4]
on the following dimensions were derived: household
Cervical cell changes
1.0
[0.0, 2.0]
Urinary tract infection
6.5
[4.0. 8.9]
and environmental sanitation, personal hygiene, and ex­
posure to health education. (For the definition of these
variables, together with that of economic status, ^ee
Predictors of Gynecological Morbidity and
Bhatia and Cleland, 1995a.)
Nutritional Status
The analysis was carried out in two stages. First, in
a bivariate logistic regression, unadjusted odds ratios
For the purpose of identifying the predictors or corre­ and associated p-values were calculated for each of the
lates, gynecological conditions and nutritional status
above gynecological and related conditions and indi­
have been grouped into eight broad categories: any
vidual background variables. In the second stage, five
menstrual problem (as elicited by the gynecologist); any
independent variables (rural-urban residence, years of
dysmenorrhea; clinically diagnosed vaginitis; any clini­
schooling, economic status, number of pregnancies, and
cally diagnosed reproductive tract infection; any labo­ current use of contraceptives) were selected, and a mul­
ratory-detected sexually transmitted disease; labora­
tiple logistic regression was performed to estimate the
tory-detected vaginitis (bacterial vaginosis or Candida
net relationship of each of these factors to the preva­
albicans); severe anemia; and any chronic energy defi­
lence of gynecological conditions and nutritional sta­
ciency. In order to assess the relationships of selected
tus. All variables were grouped or were categorical in
socioeconomic, demographic, and cognitive variables to
nature; for each variable, one category was selected as
each of these broad categories, a logistic regression was
the reference category. The results of the bivariate analy­
performed. The dependent variables were dichotomized
sis are shown in Table 7.
as the presence or absence of each of these conditions.
Women residing in town, those with six or more
The independent or predictor variables selected for
years of schooling, and those having four or more preg­
the analysis fall into four main groups, similar to those
nancies were found to be significantly more likely to
used in the analysis of self-reported morbidity (Bhatia
report menstrual problems than were women in the cor­
and Cleland, 1995a). The first group, representing the
responding reference categories. With regard to painful
social and economic backgrounds of the women, in­
menstruation, the probability of reporting the problem
cludes rural-urban residence, religion, economic status,
was found to be higher among urban women and among
md years of schooling. The second group of factors.
women younger than 20. The prevalence of dysmenorrhea
was significantly higher among women who had been
sterilized than among those not using contraceptives.
Table 6 Percentage of 385 women observed to have
The prevalence of clinically diagnosed vaginitis and
conditions related to nutritional problems and hypertension,
reproductive tract infections showed little significant
by type of condition, Karnataka, India, 1994
variation. Indeed, the only statistically significant as­
Confidence Interval
sociation
was between clinically diagnosed RTIs and
Condition
Percent
(95% confidence level)
number
of
pregnancies. The odds of having this condi­
Anemia or chronic energy deficiency
88.1
[84.9,91.3]
Mild anemia: hemoglobin <12 gm/dl
tion
among
women with only one pregnancy were about
67.4
[62.7, 72.1]
Severe anemia: hemoglobin <10gm/dl
16.8
[13.1,20.5]
50 percent less than among those with two to three preg­
CED grade I (BMI > 17 < 18.5)
28.7
[24.2, 33.2]
nancies.
CED grade II (BMI > 16 < 17)
15.8
[12.1, 19.4]
Regarding laboratory-detected infections, sexually
CED grade III (BMI < 16)
12.2
[8.9, 15.5]
Overweight (BMI 25-30)
2.4
[0.9, 3.9]
transmitted diseases were found to be significantly low­
Obese (BMI > 30)
1.3
[0.1,2.4]
er among Muslim women than among Hindus (Fisher's
Hypertension
2.3
[0.8, 3.8]
Exact Test p<0.01). In fact, not a single case was detected

Volume 28 Number 2 June 1997 99

• .1
Table 7 Bivariate logistic regression of gynecological and related problems, by selected background characteristics of 385
women, according to type of problem, Karnataka, India, 1994__________________________
Characteristic

Residence
Urban
Rural (r)
Religion
Hindu
Muslim (r)
Education (years)
0(r)
1- 5
6+
Economic status
Low(r)
Fair
Age
<20
<20-24 (r)
25+
Pregnancies
1
2- 3 (r)
Abortions
No(r)
Yes
Current use of contraceptives
IUD
Tubectomy
Not using (r)
Environmental sanitation
Poor(r)
Fair
Personal hygiene
Poor(r)
Fair
Exposure to health education
Yes (r)
No

(N)

Any
menstrual
problem Dysmenorrhea

Clinically
diagnosed
vaginitis

Any clinically
diagnosed RTI

Laboratorydetected
vaginitis

0.92
1.00

1.86**
1.00

0.66
1.00

0.33***
1.00

Severe
anemia

Any chronic
energy
deficiency

(291)
(94)

1.95*
1.00

1.64*
1.00

1.32
1.00

1.08
1.00

(339)
(46)

1.07
1.00

1.17
1.00

1.79
1.00

1.43
1.00

1.00

0.92
1.00

1.17
1.00

1.18
1.00

(214)
(45)
(126)

1.00
1.47
1.70*

1.00
1.72
1.10

1.00
0.99
0.96

1.00
1.17
1.11

1.00
0.96
1.02

1.00
1.65
1.04

1.00
0.71
0.50**

1.00
1.41
0.40***

(192)
(193)

1.00
0.84

1.00
0.92

1.00
0.90

1.00
1.15

1.00
1.45

1.00
0.84

1.00
0.92

1.00
0.65*

(47)
(168)
(170)

2.10
1.00
1.43

2.10*
1.00
1.35

2.18
1.00
1.40

1.49
1.00
0.95

0.32
1.00
0.75

1.62
1.00
1.06

0.98
1.00
0.68

0.77
1.00
1.09

(70)
(223)
(92)

1.30
1.00
1.90*

0.75
1.00
1.24

0.74
1.00
0.72

0.55*
1.00
1.11

1.67
1.00
1.43

0.381.00
0.60

0.92
1.00
1.31

(351)
(34)

1.00
1.53

1.00
1.85

1.00
1.20

1.00
1.22

1.00
1.96

1.00
0.60

1.00
1.09

1.00
0.69

(46)
(146)
(193)

1.24
1.35
1.00

1.31
1.62*
1.00

0.87
0.87
1.00

0.94
1.22
1.00

1.16
0.81
1.00

0.68
1.97**
1.00

0.69
0.70
1.00

0.74
1.10
1.00

(200)
(185)

1.00
1.05

1.00
1.35

1.00
1.59

1.00
1.22

1.00
1.66

1.00
1.00

1.00
0.85

1.00
0.67*

(207)
(178)

1.00
0.91

1.00
1.20

1.00
1.08

1.00
0.87

1.00
1.16

1.00
1.21

1.00
1.11

1.00
0.94

(204)
(181)

1.00
1.22

1.00
1.42

1.00
0.88

1.00
0.73

1.00
1.24

1.00
0.92

1.00
0.84

1.00
1.08

Significant at *p<0.05; **p<0.01; —psO.OOI.

(r) = reference category.

among women of the Muslim faith. Surprisingly, no ur­
ban-rural or socioeconomic variations are found in lev­
els of STD infection in this sample. Laboratory-deter­
mined vaginitis was observed to be significantly higher
among urban women and among those who had under­
gone sterilization than among their counterparts. In ad­
dition, the odds of a woman with only one pregnancy
having laboratory-detected vaginitis were only about
one-third of those among women with two or three
pregnancies.
The analysis of nutritional status indicates that the
odds of anemia and chronic energy deficiency occur­
ring among women with six or more years of educa­
tion were about 50 percent lower than among those with
no schooling. Chronic energy deficiency was signifi­
cantly lower among urban women, and somewhat lower
among those of higher economic status, and among those
with good environmental sanitation.

100

Laboratory­
detected STD

Studies in Family Planning

.

*

0.68
1.00
1.41

•Fisher’s Exact Test (two tail) p< 0.01.

The results of the multiple logistic regression are
shown in Table 8. The net association between urban
residence and the reporting of menstrual problems and
dysmenorrhea is found to be statistically significant.
In addition, a significantly higher proportion of ster­
ilized women had complaints of painful menstruation
than did those who were not practicing contraception.
In the multivariate analysis, no characteristic of the
women included in the analysis was significantly asso­
ciated with clinically diagnosed vaginitis or RTIs, or
with laboratory-detected STDs. However, laboratorydetected vaginitis (bacterial vaginosis and Candida
albicans) was significantly higher among urban and
sterilized women. The prevalence of anemia and chronic
Energy deficiency remained significantly lower among
women with six or more years of education. Chronic
energy deficiency was also significantly lower among
urban women.

Table 8 Multiple logistic regression of gynecological and related problems, by selected background characteristics of 385
women, according to type of problem, Karnataka, India, 1994______ _______
Characteristic

Any
menstrual
problem Dysmenorrhea

Clinically
diagnosed
vaginitis

Any clinically
diagnosed RTI

Laboratory­
detected STD

Laboratorydetected
vaginitis

1.95*
1.00

1.64*
1.00

1.34
1.00

1.11
1.00

1.13
1.00

1.98**
1.00

0.82
1.00

0.43'
1.00

1.00
1.00
0.76

1.00
1.86
0.95

1.00
1.03
0.95

1.00
1.23
1.13

1.00
0.72
1.04

1.00
1.37
0.85

1.00
0.74
0.55“

1.00
1.62
0.48“

1.00
0.73

1.00
0.87

1.00
0.91

1.00
1.05

1.00
1.79

1.00
0.88

1.00
0.93

1.00
1.20

0.68
1.00
1.26

0.71
1.00
1.40

0.46
1.00
0.77

0.60
1.00
1.15

1.38
1.00
0.74

0.52
1.00
0.69

0.84
1.00
1.27

0.78
1.00
1.49

1.61
1.17
1.00
0.72

1.33
1.67*
1.00
-0.41

0.78
0.77
1.00
-1.55

0.89
1.08
1.00
-0.42

0.60
0.76
1.00
-2.62

0.56
1.82*
1.00
-1.36

0.84
0.68
1.00
1.16

0.93
1.05
1.00
0.53

Residence
Urban
Rural (r)
Education (years)
0(r)
1- 5
6+
Economic status
Low(r)
Middle
Pregnancies
1
2- 3 (r)
4+
Current use of contraceptives
IUD
Tubectomy
Not using (r)
Constant

Significant at *p<0.05; “p^O.OI; *“p<0.001.

Severe
anemia

Any chronic
energy
deficiency

(r) = reference category.

Gynecological Conditions and Nutritional Status

Discussion

Several women in the study were found to have more
India is such a huge and varied country that no single,
than one gynecological condition and low nutritional
localized study can lay claim to representing the whole.
status. Therefore, an examination of the interrelation­
Compared with many other states, Karnataka is ad­
ships between different conditions is of interest. The
vanced. Adult literacy is 56 percent compared with the
likelihood that a woman with a clinically diagnosed re­
national figure of 52 percent, and infant mortality stands
productive tract infection would have a laboratory-con­
at about 65 deaths per 1,000 live births, compared with
firmed sexually transmitted disease is about three times
the all-India average of 78 deaths. In terms of health
higher than that for women with no clinically diagnosed
and family planning services, the state is also progres­
RTI (not shown). However, no statistically significant
sive. According to the 1992-93 National Family Health
relationship was observed between anemia, chronic en­
Survey, 38 percent of deliveries in Karnataka are insti­
ergy deficiency, and the gynecological conditions in­
tutional, compared with 26 percent for the entire coun­
cluded in the analysis. Even when severe anemia was
try. Similarly, contraceptive practice among married
considered, the relationship between this condition and
couples is relatively high at 49 percent, compared with
the other seven broad categories of conditions did not
an all-India estimate of 41 percent.
reach statistical significance. However, severe chronic
The site of the study was selected as being typical
energy deficiency was observed to be significantly re­
of rural Karnataka and sufficiently far from Bangalore,
lated to clinically diagnosed RTIs. Women with grade
the capital city, to be shielded from the influence of
HI CED were two times more likely to have a clinically
this rapidly growing, modern metropolis. The area is
diagnosed reproductive tract infection than were those
predominantly agricultural, although it contains one
not having this condition.
town of some 47,000 inhabitants. The study sample was
An attempt was also made to analyze relationships
drawn from the town and from 48 villages, represent­
between specific conditions included under each of the
ing a distinct advantage over the two earlier population­
broad categories. The results of this analysis indicate
based inquiries into gynecological morbidity (Bang, 1989
that women with clinically diagnosed upper reproduc­
and Younis, 1993), both of which were restricted to two
tive tract infections or pelvic inflammatory disease are
villages and included no urban areas.
three times more likely than those who are not so diag­
One of the main lessons of the present study con­
nosed to report menstrual problems. Furthermore,
cerns methods rather than substantive findings. The in­
women having a laboratory-detected sexually transmit­
corporation of intrusive clinical procedures into field
ted disease have a three times greater risk of having FID
studies is usually plagued by noncompliance. In this in­
than do women with no detected STDs.
stance, more than 90 percent of respondents agreed to

Volume 28

Number 2

June 1997

101

vaginal and other forms of examination. This exception­
ally high rate of compliance came about for several rea­
sons that have implications for the design of future stud­
ies. One reason concerns the familiarity of respondents
with field investigators. By the time they were invited
for clinical examination, the women had been visited by
Hy interviewers at monthly intervals over the span of a
year. Inevitably, a high degree of trust and rapport had
developed. A further critical factor was proximity. A tem­
porary clinic was set up successively in each of the study
villages and in the town, an arrangement that minimized
disruption of the respondents' daily lives.
One of the purposes of the study was to assess child
health, although results on this topic are not presented
here. Women were, therefore, invited to the clinic for
an assessment both of their child's and of their own
health. In addition to the gynecologist, a pediatrician
attended all clinics. This emphasis on child health was,
most likely, an important influence on attendance. Last­
ly, treatment was offered both to women and to children.
The relative contribution of these factors to the achieve­
ment of high compliance is impossible to assess, but the
experience does prove that problems of noncompliance
can be overcome by careful study design.
As noted above, this study collected information on
gynecological and related conditions based on women's
perceptions and self reports within the context of a larger
cross-sectional study. (For detailed results, see Bhatia
and Cleland, 1995a.) A comparison of estimates from
different data-gathering methods is in progress. It will
provide valuable information and will have larger im­
plications for determining the appropriateness of ser­
vices that could be provided for screening for RTIs and
other gynecological complaints.
The single most important finding of this study, re­
gardless of the method of assessment, is the high preva­
lence of reproductive tract infections. By clinical crite­
ria, 36 percent of the women studied were diagnosed
as having such an infection. The laboratory tests re­
vealed evidence of reproductive tract infections in 56
percent of cases. The self-reports yielded a somewhat
lower estimate of symptoms of infection, suggesting that
many infections are asymptomatic or produce symp­
toms that are regarded by women as normal. Some 29
percent of respondents reported symptoms associated
with upper or lower reproductive tract infection. Esti­
mates of infection in this study are much lower than
those found in two villages in Maharashtra (Bang et al.,
1989), but are broadly similar to those found in rural
Bangladesh (Wasserheit et al., 1989). The Bangladesh
study, however, was confined to symptomatic women,
which complicates comparison.
The two most common conditions found in this

102

Studies in Family Planning

study were bacterial vaginosis and mucopurulent cer­
vicitis. Medical opinion differs on the implications of
these conditions, but some evidence exists that un­
treated bacterial vaginosis may be a risk factor for pel­
vic inflammatory disease (Moi, 1990). However, approx­
imately one-fourth of the women had clinical evidence
of PID, cervical ectopy, and fistula. These conditions
may represent a serious long-term threat to the health
and fecundity of women. The contribution of STDs to
the high prevalence of reproductive tract infections in
this sample appears to be modest. Evidence of STD in­
fections was found in 10 percent of the women stud­
ied, a proportion that appears to be much lower than
that in the Maharashtra study, which may reflect dif­
ferent sexual mores. A preliminary analysis of risk fac­
tors for RTIs found a significant link between tubectomy
and laboratory-confirmed vaginitis. A similar result was
obtained in Bangladesh (Wasserheit et al., 1989). In ad­
dition, women who had undergone tubal ligation were
much more likely to report painful menstruation. Al­
though the etiology of these associations is uncerta.
the results potentially have far-reaching implications
for India's family planning program, which has always
placed a primary emphasis on permanent methods
of birth control. Further biomedical investigation is a
priority.
A persistent urban-rural difference in the report­
ing of menstrual problems and dysmenorrhea and in
laboratory-confirmed vaginitis was also found. How­
ever, the prevalence of chronic energy deficiency is much
lower among urban women. Women's lack of school­
ing was unrelated to most gynecological conditions, but
was strongly associated with anemia and CED. That the
lack of education is more strongly predictive of these
conditions than economic status is yet another testi­
mony to the pervasive influence of this factor on health
and welfare.
The study has also highlighted the extreme diff
culty of collecting information on menstrual disorders,.
When interviewed by experienced nonmedical staff, 15
percent of the respondents reported a menstrual prob­
lem. Under the more extensive probing of the gynecolo­
gist, this estimate rose to 62 percent. The Maharashtra
study reported a similar experience: 20 percent of re­
spondents reported menstrual problems to a nonmedi­
cal interviewer and 58 percent reported them to a phy­
sician (Bang et al., 1989). Clearly, the propensity of wom­
en to report menstrual problems is highly sensitive to
the context of the inquiry and, until more reliable meth­
ods of obtaining this type of information are devised,
little gain can come of adding menstrual problems to
other conditions in the attempt to form total estimates
of gynecological morbidity.

A final important result, albeit one that concerns the
general health of women rather than gynecological dis­
orders specifically, is the high level of anemia and
chronic energy deficiency found during the study. Al­
though Karnataka is a relatively advanced state, a large
majority of the sample were suffering from these dis­
orders. A study conducted by the Indian Council of
Medical Research (ICMR) in six centers in India during
1986-87 also found that 62 percent of pregnant wom­
en had a hemoglobin level of less than 1 Igm/dl (ICMR,
1992). This combination of widespread undernutrition
and malnutrition and reproductive tract infections in­
dicates the enormity of the task ahead, if radical im­
provements in women's health are to be achieved in In­
dia. The translation of the Cairo rhetoric into reality
clearly will require innovation, expense, and improved
diagnostic tools. Above all, it will require political will.

Indian Council of Medical Research Field Supplementation. (ICMR).
1992. Trial in Pregnant Women. An ICMR Task Force Study. New
Delhi: ICMR.

Ishizuka, Y., K. Oota, and K. Masyvycgu. 1972. Practical Cytodiagnosis. Philadelphia: J.B. Lippincott.

Jeffcoate, Norman. 1982. Principles of Gynaecology. London: Butter­
worth Scientfic.

Koss, Leopold G. 1968. Diagnostic Cytology and Its Histopathologic
Bases. Philadelphia: J.B.Lippincott.
Moi, H. 1990. "Prevalence of bacterial vaginosis and its association
with genital infections, inflammation, and contraceptive meth­
ods in women attending sexually transmitted diseases and pri­
mary health clinics." International Journal of STD and AIDS 1,2:
86-94.
Naib, Zuher M. 1970. Exfoliative Cytopathology. Boston: Little & Brown.

Shetty, P.S and W.P.T. James. 1994. Body Mass Index—A Measure of
Chronic Energy Deficiency in Adults. New York: United Nations
Food and Agriculture Organization.

United Nations. 1994. Report of the International Conference of Popula­
tion and Development, Cairo. Document A.Conf. 171/13.

References

Wahi, P.N., U.K. Luthra, S. Mali, and M.B. Slijinkin. 1972. "Preva­
lence and distribution of cancer of uterine cervix in Agra dis­
trict in India." Cancer 30: 720-25.

Bali, P. and R.A. Bhujwela. 1969. "A pilot study of clinic-epidemio­
logic investigations of vaginal discharge by rural women." In­
dian Journal of Medical Research, 57,12: 2,289-2,299.

Wasserheit, Judith N., Jeffrey R. Harris, J. Chakraborty, Bradford A.
Kay, and Karen J. Mason. 1989. "Reproductive tract infections
in a family planning population in rural Bangladesh." Studies in
Family Planning 20,2: 69-80.

Bang, R.A., A.T. Bang, M. Bitule, Y. Choudhary, S. Sarmukaddam,
and O. Tale, 1989. "High prevalence of gynecological diseases
in rural Indian women." Lancet 8,629,1: 85-88.
Bhatia, Jagdish C. and John Cleland. 1995a. "Self-reported symptoms
of gynecological morbidity and their treatment in South India."
Studies in Family Planning 26,4: 203-216.

--------- . 1995b. "Determinants of maternal care in a region of South
India." Health Transition Review 5: 127-142.
---------. 1996. "Obstetric morbidity in South India: Results from a
community survey." Socul Science and'Medicine 43,10:1,507-1,516.
Brabin, L., J. Kemp, O.K. Obunge, J. Ikimalo, N. Dollimore, N. Odu,
C.N. Hart, and N.D. Briggs. 1995. "Reproductive tract infections
and abortion among adolescent girls in rural Nigeria." Lancet 345:
300-304.

Dewhurst, John (ed.). 1981. Integrated Obstetrics and Gynaecology for
Postgraduates. London: Blackwell Scientific Publications.
Howkins, John and Gordon Bourne (eds.). 1985. Shaw's Text Book of
Gynaecology. Delhi: BI Publications.

Younis, Nabil, Hind Khattab, Huda Zurayk, Mawaheb El-Mouelhy,
Mohamed Fadle Amin, and Abdel Moneim Farag. 1993. "A com­
munity study of gynecological and related morbidities in rural
Egypt." Studies in Family Planning 24/3: 175-186.

Acknowledgments
The authors would like to note the able assistance provided
by N.S. Sanath Kumar and S. Ramaswamy in data process­
ing. We are also thankful to Jinka Subramanya and J. Vasudev
Rao, pathologists at Bangalore Hospital, for their hard and
diligent laboratory work. Chris Elias of the Population Coun­
cil, Thailand, provided valuable comments on a draft of the
paper. Financial support for the study was provided by the
Ford Foundation, which is gratefully acknowledged. The
analysis of the data was partly funded by the Overseas De­
velopment Administration of the United Kingdom.

Volume 28

Number 2

June 1997

1 03

Estimates and Explanations of Gender
Differentials in Contraceptive Prevalence Rates
Alex Chika Ezeh and Gora Mboup

This article examines gender differentials in the reporting of contraceptive use and offers explanations
regarding the sources of these differences. Data fromfive countries where DHS surveys were conducted
recently among men and women are used in exploring these differences. The gap exists in all five
countries, with men (or husbands) reporting greater practice of contraception than women (or wives).
Results from the bivariate analysis suggest that the gap is attributable to polygyny and to gender
differences in how the purpose of contraception is understood, rather than to male extramarital sexual
relations. Additionally, gender differences in the definition ofcertain contraceptive methods and differences
in the interpretation of questions about contraception contribute to the observed gap. Thesefindings
are also consistent with results of the multivariate analysis. (Studies in Family Planning 1997; 28,2:
104-121)

The disparity between men and women (or husbands
and wives) in their reporting of contraceptive use has
long been documented in the demographic literature
(Yaukey et al., 1965; Stoeckel and Choudhury, 1969;
Green, 1969; Koenig et al., 1984; Hopflinger and Kuhne,
1984; Mitra et al., 1985). The nature of this discrepancy
is also well established, with men or husbands gener­
ally reporting greater use of contraceptives than do
women or wives.’ In a recent study, Ezeh and his col­
leagues (1996) found current contraceptive use to be
higher among currently married men compared with
currently married women in 12 of the 14 countries they
studied. In some of the countries in which men reported
greater contraceptive use than women did, current use
among men was found to be twice the level reported
by women. However, a difference of only two percent­
age points separates the reports of the men and the
women in the two countries where women reported
greater contraceptive use than men did. Most research­
ers have treated observed inconsistencies in contracep-

Alex Chika Ezeh is Research Associate, Applied Research
and Development and Gora Mboup is Country Monitor for
francophone Africa, Demographic and Health Surveys,
Macro International, 11785 Beltsville Drive, Calverton, MD
20705.

104

Studies in Family Planning

tive-use status of husbands and wives as respondentrelated errors, and in particular, as underreporting by
wives (Koenig et al., 1984).
The treatment of inconsistencies in the reporting of
contraceptive use between married partners as misre­
porting by one partner assumes that spouses use con­
traceptives exclusively with their marital partners. An
extension of this assumption is that sexual activity
among married men and women occurs exclusively
within marriage. The treatment of wives' reports as misor underreporting assumes that spouses have equal
knowledge of the use of a method and that both part­
ners have the same definition of what constitutes con­
traceptive practice. In the paper on India by Koenig et
al. (1984), 67 percent of spouses in couples where at least
one partner reported use of contraceptives differed in
their response to the question on current use. Of these,
more than 84 percent of the discrepancies occurred
when husbands but not wives reported use of a method.
Three methods: condoms, periodic abstinence, and va­
sectomy, account for 79 percent of the absolute differ­
ence between the reports of husbands and wives and
for more than 93 percent of the difference associated
with only husbands reporting use of a method (see
Table 1 in Koenig et al., 1984). Except for prolonged ab­
stinence, a husband can use these methods without his
wife's knowing. Even when the wife knows, she may
not associate the use with contraception. Husbands may

it-:

'ti?f

)

E

I

LEARN i

I

N G

ABOUT

i
»■<.

Iit



VM ■
A>..•;.<S
PlHV

•"*1

%

w



H‘l i'

y

'4

to
iw;

j

;

A PRACTICAL

6

Pl

It

BEGINNING

'/J

!4'

SONDRA ZEIDENSTEIN AND KIRSTEN MOORE
EDITORS

THE POPULATION COUNCIL
INTERNATIONAL WOMEN'S HEALTH COALITION
NEW YORK

I
0

*

‘1

.■

r
<

*

/4 Community Study of
*
Gynecological Disease
I
in Indian Villages
Some Experiences and Reflections

Rani Bang and Abhay Bang

l

I n third world countries, most
/ women tend to encounter the
'
health care system only when they' are the target of family planning programs. Little attention has been given to the reproductiveVealth of nonpregnant women. One reason for the relat

neglect of gynecological care is a failure to appreoate the extent
of unmet needs in rural areas (Bang, Bang, el al 198J).
In Gadchiroli, a remote, backward district in the central part o
India, where we run a nongovernmental orgamz.auon called
Society for Education, Action and Research m Community Health
(SEARCH), we did a community-based study of gynecological
problems of rural women, the first study of its kind m the• develop­
ing world. In the study, we sought to determine: (1) the pi eva­
lence, types, and distribution of gynecological diseases in rura
women; (2) the awareness and perceptions of the women abou
dieir gynecological and sexual disorders; and (3) the proportion
of women who have access to gy necological care.
,
The study team (a female gynecologist with ten years exper ence as a consultant, a physician, a pathologist, a aboratory tec inician, a nurse, and female social workers) visited the held camp
and conducted the study. Female soaal workers, village leade ,
223

Q

ft

!

224
*11
il

n
.1

Ij

ij

I •! '
| I* :

ri

;J i

/

YaeadrSVo1IdnteerVn?e.d 3,1 fema'eS Who
°,der tha" ^Ive
years old or who had reached menarche to i
participate in the
study, whether or not they had symptoms. A field r— - I camp was set up
in the village, with facilities for private interviews and pelvic
: examinauons, a pathology laboratory, and an operating theater
Informauon was obtained on personal history, socioeconomic st.

St eXnCeP
re8arding gynecological symptoms;
hhtorv Th nCe Care:,and obs^trical, gynecological, and sexual
incZnJsn W?men
had genera' P'1^"1 examinations
the nelJ P
Um examinations and bimanual examinations of
the pelvis, unmarried girls with intact hymens had rectal rather

were foTnd" '™adOnS (Bang’ Bang' et al. 1989). Women who
were found to have disease were offered treatment.
suffered fnd‘ngS Were. stardlng- Ninety-two percent of the women
If I
°m g{necol°glcal diseases. Each woman had an average

medttir b“‘ °nly 7 Per“M Of the

I

I !

I



This study generated interest in women’s gynecological dis­
ses as a public health problem in various parts of life world
Many groups interested in women’s health are now initiating simiar types of studies in their own areas (see the chapter “Involving
um^Bm e3
Morbidity Sl^y in Egypt” in this voh
ume). But epidemiologtcal and anthropological studies on reprobee'n"
reladvely -w, and research methods have no"
een perfected. Hence, along with hard data, observations and
shfred6?"5! abOi,Ut the Pr°CeSS neCd t0 be documented and
shared^ In th.s chapter, we recount some of our personal experi^nces dunng the process of conducting this study, with the hope
tha they may be of use to others. Methodology will gradually

Z nth11'

I

’COMMUNITY STUDY OF GYNECOLOGICAL DISEASE

h

°ft data' WC d° nOt’ h°WeVer’ ^end

-ggest

a h ? d u°U d necessanly rePeat what we did. There could be
a hundred other ways of reaching die same goal.
of thJeLe ,S one P°ssible excepdon, however. During the evolution
of the idea for this study, as well as while working on it, we realized
that reproductive health involves both males and females. To be
successful, any study of reproducdve health should involve a team
of male and female researchers. We have experienced the fruidulness of combined perspectives in our own work, time and again.

I

Rani Bang and Abhay Bang

/

225

Getting Started
The Story behind the Study
We, Rani and Abhay, were classmates during medical school
Subsequendy, Rani did her M.D. in obstetfics and gyneco^, and
Abhay4though inclined to pursue public health, gained an M.D
internal medicine. When we decided to marry, we didn t realize
the public healdi potential implicit in our particular combmation
of professional backgrounds and interests. But, as a couple sharing
the same profession, we often talked at die dmner table about vari­
ous types of padents seen during die day. Ram enjoyed telling sto­
ries and anecdotes about cases from die dime; Abhay enJ°yed
listening to diem and diinking diem over. One evening, Abhay
observed dial it seemed from Rani’s descripuon dial die majority
of women in society were suffering from one gynecolog.cal prob­
lem or another. How could it be such a frequent occurrence.
Clinic records did not reveal what proportion of women in
area had gynecological diseases. Could Rani subs^tia^OaU”^
picions? Rani counted die number of women residing ir a. hous­
ing colony whom she had recently treated for gynecological
problems. Surprisingly,
Surprisingly, it turned out that almost half of
women in
the colony
in the
colony were suffering from gynecological diseases.

That fact was astonishing.
Abhay said, “I don’t know of any estimate of the burden o
gynecological diseases in a community. Maybe no one has realize
Se magnitude and frequency of diis problem. If what you^have
experienced is representative, then gynecological diseases m y
turn out to be a major public healdi problem. We shoul1 inten­
sively and systemaucally study a community or a defined popul
tion to estimate the prevalence and incidence of gynecological
diseases. It may be an eye-opener!”

Preparation for the Study
j at Johns Hopkins for master’s
Later, when we were studying
did a computerized literature search
degrees in public healdi, we <
on populadon-based studies of gynecological disc ases in developing

*

226

/

Community Study of Gynecological Disease

countries and found that there were none! Thci e were many stud­
ies on screening populations for cervical carcinoma or syphilis, but
no estimates of die vast array of gynecological problems of women—
such as white discharge, vaginitis, cervicitis, pelvic inflammatory dis­
ease, or menstrual problems. Our belief that a study on the preva­
lence of gynecological diseases needed to be done became stronger.
We were fortunate to meet Professor Carl Taylor at Johns
Hopkins School of Public Health. He introduced us to established
research methods in public health and also guided us in develop­
ing a study protocol for the community study of gynecological dis­
eases. We learned from him the absolute prerequisites of clearly
stated, measurable objectives; a study design; sampling; definition
of various diagnostic entities; and data collection instruments,
including the creation of dummy tables.
The reasons that no one had conducted such a study soon
became obvious to us. Intending to help us develop the protocol,
Professor Taylor introduced us to a senior American gynecologist
who had headed obstetrics and gynecology departments in devel­
oping countries for years. He kept wondering why we wanted to
estimate prevalence and incidence. He thought that instead we
should take direct action by treating the gynecological diseases of
women who came to clinics or hospitals. He also said, “How does it
matter if many women have white discharge? Just as many people
get the common cold or have nasal discharge; is it of any conse­
quence? You are wasting your time measuring trivialities!”
As we attempted to raise funds over the next two years (1984—85),
we frequently heard similar discouraging comments. “Measuring
prevalence of gynecological disorders? Why? What for? Aren’t there
more serious diseases?” But Professor Ramalin- gaswamy, who was
then director general of die Indian Council of Medical Research,
and die Ford Foundation's New Delhi office decided to put trust in
us and in our ideas. We got die necessary financial support.

Selection of Villages
In 1986, we started SEARCH and moved to Gadchiroli, a district
that was relatively new to us. Our clinical work in the hospital
helped us rapidly gain the confidence of the local population,
and people from surrounding villages started contacting us about

9

r
Rani Bang and Abhay Bang

/

227

various problems. Gradually, the personalities of individual vil­
lages became known to us. We finally selected two villages—Wasa
and Amirza—for the study of gynecological diseases.
They were average, representative villages as far as socioeco­
nomic and demographic characteristics were concerned. They were
relatively distant from town, but not too remote. They had united
communities and backgrounds of collective action. For example,
the entire Amirza village had once boycotted voting in an election
because the government had not constructed tlie much needed
approach road. A community study of gynecological diseases
required a collective decision and die cooperation of the whole vil­
lage, since all women, whether diey had symptoms or not, were to
be examined in order to estimate true population prevalence. Wasa
and Amirza had educated village leaders who could understand die
need for such a study and acdvely mobilize the village community.
In Wasa, a small mission hospital was used as the base to con­
duct the study. In Amirza, Lambe Guruju, a bachelor teacher,
vacated his own house and made the necessary renovations to
enable us to conduct the study. Each woman participating was
required to visit five rooms: registration, interview by a female
social worker, history and examination by a female gynecologist,
pathology laboratory, and dispensary. The interview and examinadon were conducted in privacy. A woman cannot be at ease if
she feels that there is a direat of being overheard or watched dur­
ing such an interview and examination.

Interacting with the Community
i Community Participation
!

Why should villagers pardcipate in such a study? This was a million-dollar quesdon and one that we thought was important to
address directly with the community. Due to the aggressive family
planning program, people are always afraid of veiled family plan­
ning acdvity. It is necessary to remove this apprehension before
beginning any study of gynecological diseases.
With the help of village leaders, group meetings of villagers
were held, separately for women and men, to explain die need for

228

/

COMMUNITY STUDY OF GYNECOLOGICAL DISEASE

and the nature of the study. The village people, especially the lead­
ers, were informed that this was a research study that would help
their women and also have the policy impact of improving the
health of womankind elsewhere. Such meetings helped in two
ways. Villagers did not harbor suspicion about being used this
was a sort of informed consent of the community. And they felt
proud that they were helping to bring out information on
women's reproductive health to the whole world. It was not just
our study or research project; it was their research project.
The inauguration of the study was organized by the villagers
with tremendous enthusiasm and fanfare. The who e vi age
observed a holiday, and community dinners were served to about
a hundred guests in both villages. The idea and nature of the
study were again explained. This also helped remove the suspi­
cion and fear in women that our study could be a family planning
camp under some different garb.

Involving Men
It was essential to involve men for several reasons. They are the
decision makers in the family and, if not convinced, they might
not allow their women to participate in the study. However they
worry about the health of their wives, sexual relations, and the
health of their progeny, and most of the women communicate
with their husbands about their reproductive health problems,
our study, when we asked, “Have you spoken to anybody about
your gynecological complaint?” 80 percent replied yes. When
asked with whom, 76 percent responded that they had communi

cated with their husbands.
In another vivid example, a woman came to us for white discharge. When asked how she first noticed it, she respondec t la s le
and her husband both took baths every night and were very parucular about their sexual cleanliness. But they not.ced a foul odor at
the time of intercourse and were worried. Each thought drat it was
not his or her problem. To prove it, they smelled each other s geni­
tals and discharge and reached the conclusion that something w
wrong with die wife. Therefore, she came for the checkup.
But many men remain ignorant about reproductive health and
women's suffering. If diey are made aware, they can help and can

Rani Bang and Abhay Bang

/

229

persuade their women to seek health care. For example, when we
organized a camp for village chiefs and showed them slides on
reproductive health, they remarked, “Oh! We never knew that our
women suffer from such problems! Now we know the causes and
understand that these can be easily detected and treated.”

Village Volunteers
The village volunteers, who were all young men from the commu­
nity, were a great asset in conducting the study. Why did these
young men, unpaid for this work, take an interest in the study of
women’s gynecological diseases? Months later, they told us, “We
educated, unemployed youth while away our time and hence are
usually considered vagabonds in our own village. One day, we
were told that two America-returned doctors were coming to
Amirza lor some work. We awaited you with the idea of doing
some innocent mischief. When you came, we couldn’t believe
that you were such America-returned doctors—wearing simple
Indian-style dress of khadi. You spoke to us in local dialect. At that
moment, we decided to help you in whatever you wanted to do!”
What mattered most to them was not the topic of the study or the
potential significance of the research but how we dressed and
talked! Those young men, besides their contribution to the study,
have taught us a lot. Three of them later became our full-time
workers and took a leading part in a campaign against alcohol.
These volunteers did the population enumeration and made a
list of all women in the village eligible for the study. They sug­
gested the idea of giving each woman an appointment for her
examination. They gave several reasons. Since women have house­
hold work and also work in the field, they can’t abandon their rou­
tines and wait at the study site indefinitely for their turns. In
addition, women in the area have a notion that they should come
for abdominal and pelvic examinations with an empty stomach.
Hence, unnecessary waiting should be avoided.
Later, we learned another reason that women must be told
at least a few days before they will be examined. Some women
hesitated about participating in the study on the day of the
appointment because they were not “clean and ready.” We
observed that practically all rural women came with clean-shaved

?

230

/

Community Study of Gynecological Disease

pubic areas; rural women consider it indecent to expose their
hairy vulvas. Since women remove the hair by plucking it out with
their hands—quite a painful and time-consuming job—they defi­
nitely need prior notice to clean themselves.
The volunteers gave each eligible woman an identity card with
the date and approximate time to come to the study site written
on it. The study was conducted three days a week for five months.
The volunteers would go to each woman on the evening prior to
her appointment and remind her of the next day’s appointment
for the study. If they found any resistance, they tried to persuade
tlie women by getting help from other women or leaders, if nec­
essary. There was a waiting list, and if one woman could not come
for any reason, the next one on the waiting list would be called.
This was a skillfully handled, organized operation.

Providing Treatment
All the women participating in the study who had any disease
w'ere given treatment, free of cost. This was both an ethical and
a practical necessity. And although the study addressed only
gynecological diseases, people’s health needs know no such bound­
aries. Other types of diseases needed equal attention. Hence,
Abhay treated the medical problems of the women in the study
and also provided services to husbands and other family mem­
bers. This had a positive impact in eliciting cooperation.
For example, in Amirza, the landlord’s wife, an elderly woman,
had been sick and inactive for fifteen years. She had tried all sorts
of therapies without any relief. Abhay diagnosed her as suffering
from myxedema (hypothyroidism) and treated her. The response
to treatment was so dramatic that the woman’s total appearance
and life were changed. This gave us an instant reputation in the
village as “good doctors.” Such a reputation helped elicit wider
participation in the study, though it was not the sole determinant.
Women who were found to have gynecological problems were
asked to come back with their husbands and other close family
members. The problem was explained to them with the help of
sketches. Women and men were shown abnormal findings in the
pathology tests, such as worms in stools, microfilaria in blood
smears, and Trichomonas vaginalis and Candida in vaginal smears.

Rani Bang and Ab hay Bang

/

231

Because of these explanations, people were convinced about the
need for, and usefulness of, laboratory tests.

Nonparticipation
Since the study was population based, requiring the participation of
all or most of the eligible women (regardless of the presence of
symptoms), our two greatest challenges were full participation in all
aspects of the study and unbiased participation in die perception
study. Several issues hindered participation. Because we wanted to
do perception studies simultaneously widi die medical exams, we did
not explain to die women how die internal examinadon was done or
educate diem about the anatomy, physiology, or diseases of die
reproductive organs prior to the study. Since only 7.8 percent of the
women examined had received previous gynecological care, most
were totally ignorant about die process of gynecological history tak­
ing and pelvic examinadon. After die first few days of the study, a
rumor spread in die village: “Ohl It’s very vulgar! [Nanga nanga alu.]
They ask obscene questions; diat lady doctor wears a big gunny bag
[glove] and inserts die whole arm inside die vagina!’’ Women who
heard tiiese rumors were afraid to participate in the study.1
When we heard this, we sent our social worker to explain the
trutii. We showed die women leaders, die dais (traditional birtii atten­
dants), and die friends accompanying the women being examined
how the internal examination was done. We showed diem gloves and
explained why we needed to wear diem. 1 his, of course, helped a lot.
Many women had cervical erosion or cervicitis, and when other
women saw this, tiiey were convinced of die need for and importance
of an internal examination. We learned tiiat women did not have
much of a problem being observed by otiier trusted village women.
In retrospect, we believe tiiat we even could have fixed a mirror so
tiiat women could see tiieir own examinations in progress.
If we had provided adequate education and explanation before
starting the study, problems of misunderstanding would not have
become a major threat. Researchers who are interested in measur­
ing gynecological morbidity without doing perception studies
should first use various methods (slides, posters, group discus­
sions) to explain die whole process to women. If there is a risk of
modifying perceptions with these explanations, the perception

232

/

Community Study of gynecological disease

study should be done first, with the gynecological morbidity study
done only after an educational phase.
Another problem arose in Wasa, the first study site. Whenever
we found unmarried girls manifesting physical evidence of having
had sex, our own ideas of morality would compel us to preach to
the girl about the risks of such behavior. The result of our behav­
ior was that everyone, especially the unmarried girls in the village,
came to know that by examining them, the lady doctor could find
out the secrets of their sexual lives. The unmarried girls started
avoiding participation in the study.
In one family, there were two unmarried girls eligible for the
study. The mother came for her examination, but the two daughters
never turned up, despite repeated efforts of village leaders, volun­
teers, and social workers to convince them. In the end, the father of
those girls told our social worker, “See! 1 know my daughters are not
good. If they go for examination, the lady doctor is sure to know
about tlieir affairs, and my image in the doctor’s eyes will be low­
ered. So it is better that I don’t send my daughters for examination.
We had learned our lesson. In the next study village, Amirza,
.

«
r

mouths
shut and refrain
we had to train ourselves to keep our i..
v-----------from preaching, even though we saw evidence of premarital sex.
The participation of unmarried girls was quite high in Amirza.
Gynecological disease is a sensitive area of inquiry. In spite of our
best efforts, only about 60 percent of the eligible women in these
two villages participated in the study. One way of handling the non­
participation of the other 40 percent was to investigate whether
nonparticipants were in any way different from the participants. As
we described in a paper published in Lanc^ a random sample sur­
vey showed that this was not the case (Bang, Bang, et al. 1989).

Talking to Women
9

Problems of Language
In taking the women’s sexual and health histones, we learned
about the importance of knowing the local usage of language lor
effective communication. Women often present their complaints
in the form of symptoms that, in their culture, mean something

r
I
?
:■

Rani Bang and Abhay Bang

/

233

different from their literal meaning. In our area, if a woman has a
white discharge, she will say, “1 have weakness.” If the doctor does
not take this as a clue and ask her the leading question, “Do you
have white discharge?” the real complaint is missed.
We recollect now that when we were working as resident doc­
tors in the medical college, Nagpur, which is 200 kilometers from
Gadchiroli, at least 50 percent of the female patients in the busy
medical or gynecological outpatient department were diagnosed
as having “general debility,” and were treated with vitamins. No
disease could be detected in these women, whose main complaint
was “weakness.” Now we understand what they wanted to commu­
nicate. In retrospect, we see how futile were the vitamins dished
out to thousands of such women in one single hospital, day in and
day out, when they really sought care for the unspeakable symp­
tom of white discharge. Since we doctors learn medicine in
English and do not know the subtler meanings of the terms
women use, such as "weakness,” communication is faulty.2
Medical personnel and social workers need to know the local
terms for specific medical problems. Gadchiroli has a rich local
vocabulary, for example, balant dosh for puerperal psychosis,
padar or pandhara pani for white discharge, khaira dharane for
pain in one leg during pregnancy, or inavia-bhachich dukh for the
al lei pains of delivery.
It is important to understand not only the spoken language
' 5and its complexities but also “sign language." Women often don’t
i speak out their symptoms but make signs or gestures, by which the
health care provider is supposed to guess the symptom or the dis­
ease. For example, in a case of a prolapsed uterus, a woman may
not utter a single word but convey the problem with a particular
sign, using her hands. There are some symptoms that the patient
is not supposed to state at all. In the case of sterility, the sterile
woman signals other women to talk while she keeps mum.
Some words also have double meanings. For example, while
eliciting the medical histories, we asked, "How many children do
you have?” In the local language in Gadchiroli, it is constructed
as, “How many boys do you have?” (Tumhala kill mulan ahetl).
Because the word mulan means "boys” as well as "children,”
women often told us only the number of boys they had. There­
fore, we had to ask them separately (or the numbers of boys and
girls and for the numbers of dead children, stillbirths, and abor-

234

/

Community Study of Gynecological Disease

lions. Women generally avoided talking about pregnancy loss
because, as one woman expressed it, “What is the use of telling
that? It has gone to God!”

Underreporting of Disease
We also began to see die prevalence of underreporting or under­
recording of symptoms. There are a number of reasons for this.
Definite perceptions exist about what is “normal” and what is
“abnormal” in terms of g)necological health. For example,
women in our area believe that heavy menstrual flow is normal
and desirable and would not report it as a complaint. Since they
linked sterility with black-colored menses, instead of talking about
sterility, women complain of “black menses.” We medical profes­
sionals might ignore dark-colored menses, considering it normal.
We also learned that women often reveal one set of symptoms
or history to the doctor and another to the social worker. For
example, while examining one woman who came to us at the
study site, we found that she had fresh injury marks on her body,
which made it look as if someone had beaten her. When asked
about the marks, she said without any hesitation that her hus­
band had beaten her. We were annoyed and gave her a dose of
feminist reason: “It is criminal to beat a woman like this, why do
you tolerate it?” She very calmly said, “Oh! What is wrong in that?
He came home tired from the field in the evening, and my food
was not ready, so he beat me as he was very hungry. It was not his
fault. It was natural for him.”
As soon as we finished her examination and sent her for
pathology tests, the social worker came to us excited and said,
“Did she tell you her story?” When we answered yes, the social
worker said, “But do you know why he beats her?” We said, “Yes,
because she does not cook the food on time! But this is no way to
treat a wife. We will call the husband and talk to him.” The social
worker laughed and told us the story that the woman had told
her: that she had conceived in spite of her husband’s vasectomy,
through an illegitimate relationship. The pregnancy was quietly
terminated, but the husband came to know about it. From then
on, he suspected her fidelity and regularly beat her. However, the
woman continued her illegitimate relationship with her lover,

Rani Bang and Abhay Bang

/

235

who had accompanied her that day to the study site and was still
sitting in the corridor. Thus, it was important to compare both
histories in this measurement of reproductive morbidity.
There was always underreporting of stillbirths. Dais refused to
admit having ever conducted a delivery resulting in a stillborn
baby. We initially suspected that dais concealed this information
to keep up their reputations as successful dais. But later on we
realized that dais had no concept of uterine power in the mecha­
nism of labor. They believed that the babies came out as a result
of their own bodily movements. They asked us, “If the baby dies
inside the womb, how can it come out? It is impossible! It must be
alive to come out, and die later on.”
In general, women in traditional Indian society are condi­
tioned from childhood to tolerate pain and discomfort without
making any fuss. Therefore, they don’t report symptoms even if
they have them. They believe that it is women’s destiny to suffer
from these problems. Many women said, “Under any circum­
stances, we can't afford to take to bed as sick, because then who
will do the household work?” So they continue to suffer and work,
which makes family members believe that they do not have any
problems. And some women fear that if they reveal that they have
a gynecological problem or disease, their husbands may desert
them. Besides, when traditional Indian men or women go to med­
ical practitioners, they do not necessarily voice their symptoms or
complaints. When the doctor asks about their complaints, they say,
uYou tell us what problem we have!" People believe that it is the job
of the medical professional to recognize the complaints or disease
of the patient. The attitude is, if the patient has to tell his or her
complaints, what is the doctor there for?
Finally, in spite of all die precautions taken to record symp­
toms, there can still be women who have “silent disease.” In this
study, we found that 45 percent of the women with gynecological
diseases were asymptomatic.

Taking Sexual Histories
We saw that women, and men, too, are often ignorant about the
technical language for orgasm, foreplay, and other aspects of
their sexuality. We have to explain to them what we mean before

236

/

community Study of gynecological Disease

we inquire about such issues. Our experience has been that once
they do open up, people talk freely about their sexual lives and

problems.
We learned that women experienced anxiety and stress because
of disturbed marital relationships. For example, one young woman
from Wasa, aged twenty-two years, complained of severe weakness
and loss of weight. She looked quite ill and tense. On medical
examination, no abnormality was found. We called her back for a
special appointment and talked to her in private. She told her
story with tears in her eyes. She had been married seven years ago.
She and her husband loved each other. Their house had only one
room with no privacy, and the young woman’s mother-in-law slept
in the same room, with her head near the legs of the couple, so
they could not have sex without waking her. The mother-in-law
asserted her power by controlling the availability of sex to the
young couple. The husband would climax within two minutes,
and die young woman was never satisfied. (God knows about his
satisfaction!) But because of shyness and guilt, the young couple
would not protest to the old woman—despite their severe anxiety.
The young woman told us, “My mother-in-law openly says that she
never got any sexual satisfaction from her husband and this son of
hers would not give any to her daughter-in-law.’
Even old women gave histories of regular sexual relationships.
The old women would sheepishly justify, “Chikhalala pani sula’I some water in it.”
narach," which meant "Mud will always 'hold
when
old, the body will
This was their way of saying that even ’ ‘
and
women
will always
always have sexual desire, and both men

have some secretions.

After the study was completed, the findings of the study (except
those about premarital sex) were shared with the villagers through
group meetings of women, men, village leaders, and dais, as well
as through a health awakening, or jaha. The cost ol the study,
including laboratory tests and treatments, was told to the people
too. They were pleased about being taken into our confidence.

91

Rani Bang and Abhay Bang

/

237

J.

Notes
1. We found it perplexing that women considered our inquiry into their
gynecological and psychosexual problems vulgar, since their own gos' sip was far more open and direct about sexual matters. When we
pointed out this contradiction, many women said, Our vulgar talk has
an erotic function, it gives us pleasure, excitement. But why should a
doctor talk about these things?" So the problem was one of the
expected role of the doctor.
2. This failure of communication may be one reason that public health
professionals have not realized the magnitude of the gynecological
problems that exist. It may be possible that women all over the world
are being sent home from clinics with vitamins and tonics when they
have come seeking care for gynecological problems. This may also
partly explain our finding that even though 92 percent of the women
in our study had gynecological problems, only 7.8 percent had ever
undergone pelvic examinations and received treatment.

Reference
Bang, R. A., A. T. Bang, et al. 1989. High pievalcnce of gynaecological
diseases in rural Indian women. Lancet 1 (8629):85-88.

uoh \7-:

Self-reported Symptoms of Gynecological
Morbidity and Their Treatment
in South India
Jagdish C. Bhatia and John Cleland
This article presents an analysis of self-reported symptoms ofgynecological problems among 3,600
recent mothers in Karnataka State, India. Approximately one-third ofall women reported at least one
current symptom; the most common were a feeling of weakness and tiredness (suggestive ofanemia);
menstrual disorders; white or colored vaginal discharge (suggestive of lower reproductive tract
infection); and lower abdominal pain and discharge with fever (suggestive ofacute pelvic inflammatory
disease). Obstetric morbidity, associated with the last live birth, was strongly predictive of current
gynecological symptoms. Women who delivered their last child in a private institution were
significantly less likely to report symptoms than were those who delivered at home or in a government
hospital. Nonusers or users of reversible contraceptive methods were also less likely to report symptoms
of morbid conditions than were sterilized women. These associations persisted in analyses controlling
for potentially confounding economic and demographic characteristics, and havefar-reaching policy
implications. (Studies lx Family Planning 1995; 26,4: 203-216)

Information about obstetric and gynecological morbid­
ity in developing countries is meager. Most studies are
based on information from clinics or hospitals, but, be­
cause a large proportion of women typically does not
visit such facilities, results do not reflect the true mag­
nitude of the problem. In addition, these statistics pro­
vide information only on biomedical causes; investiga­
tions of social, economic, demographic, and behavioral
determinants are rare. The only dimension of women's re­
productive health that has been thoroughly investigated
by means of large representative surveys is fertility regu­
lation and the avoidance of unwanted pregnancy. Whereas
women are the main respondents in surveys on fertility
and family planning, effective ways of gathering infor­
mation about their health have not, until recently, been
explored.1
As a consequence, evidence concerning the levels
and causes of reproductive morbidity is scarce in South
Asia and in other developing regions. One of the few
community-based studies in India is that conducted by

Jagdish C. Bhatia, Ph.D. is Professor of Health Management,
Indian Institute of Management, Bannerghatta Road,
Bangalore 560076, India. John Cleland, M.A. is Professor
of Medical Demography, Centre for Population Studies,
London School of Hygiene and Tropical Medicine.

Bang et al. (1989) concerning women belonging to the
Gond tribe. In that study, 650 women aged 13 and older
from two villages in Maharashtra were interviewed, and
clinical examinations and laboratory tests were per­
formed. Of the total number of women studied, 55 per­
cent reported gynecological complaints and 45 percent
reported no symptoms. On clinical examination and test­
ing, 92 percent were found to have at least one gyneco­
logical disease. Infections of the genital tract contributed
half of this morbidity. Only 8 percent of the women re­
ported that they had undergone gynecological exami­
nation and treatment in the past. Gynecological diseases
were found to be more common among women with a
history of contraceptive use, and about two-thirds of the
women who had undergone tubectomy attributed their
symptoms to that procedure.2
A search of the literature reveals that knowledge
about reproductive morbidity and its determinants in
India is almost nonexistent. Concerted efforts are, there­
fore, needed to provide useful information to health
planners and policymakers so that appropriate strate­
gies can be designed to bring about an improvement in
the reproductive health of women. Any such improve­
ment may also enhance the acceptance and effectiveness
of family planning programs. With these objectives in
view, the present study on self-reported symptoms of
reproductive morbidity was undertaken in 1993 in
Karnataka, a southern state of India.

Volume 26

Number 4 July/August 1995

203

Materials and Methods
According to the 1991 census, the state of Karnataka has
a total population of 44.8 million, which accounts for a
little more than 5 percent of the national population. The
decennial (1981-91) population growth rate of the state
was almost 21 percent, as compared with an all-India
average of nearly 24 percent. The male, female, and com­
bined literacy rates in the state are 67, 44, and 56 per­
cent, close to the all-India averages of 63,39, and 52 per­
cent, respectively. Although many indicators place
Karnataka at a higher level of development than other
Indian states, considerable regional variation exists
within the state. A little more than two-fifths (44 per­
cent) of married couples are protected against preg­
nancy, mainly through female sterilization. In health and
family planning programs, the state maintains a progres­
sive approach. In fact, India's first family planning clinic
was established in Karnataka in 1935. Health and fam­
ily planning services in rural areas are available through
a large network of primary health-care centers and
subcenters. In the urban areas, maternity centers and
hospitals run by the government provide free curative
and preventive health services. In addition, private
health provision in the state is becoming increasingly
common.
The present study is part of a major research effort,
funded by the Ford Foundation, to investigate the path­
ways through which a mother's education influences her
child's survival. The main study has several compo­
nents: anthropological studies; investigation of primary
schools in three states of India; a cross-sectional survey;
and a prospective study. The in-depth anthropological
studies that were carried out over a period of more than
one year aimed to identify possible linkages and to de­
velop instruments and methods for subsequent quanti­
tative studies. During these in-depth investigations, the
mother's health was found to be intricately related to
that of her child. Therefore, detailed information was
collected in the cross-sectional and prospective studies
on different aspects of mothers' health.
The analysis contained in this article is based on a
cross-sectional survey conducted in a subdistrict of
Karnataka State, situated about 70 kilometers from the
capital at Bangalore. In 1991, the subdistrict had a total
population of 164,000 spread over 293 villages and one
small town with a population of 47,000. The study popu­
lation comprised women who were younger than 35 and
had at least one child younger than five. Because the
major purpose of the survey was explanatory rather than
descriptive, no attempt was made to design a strictly
representative sample of the subdistrict. Rather, the sam­
pling strategy was based on logistical considerations; all
eligible women living in the town and in the 48 villages
204

Studies in Family Planning

having a population of at least 500 persons were in­
cluded in the sample. The selection of only one urban
site and the omission of small villages implies that ru­
ral-urban differences in the results must be interpreted
with particular caution.
A complete listing was made of all eligible women
currently residing in these locations, and field-workers
attempted to interview all of them. To maximize the re­
sponse rate, the interviewers made at least three call­
backs—including visits early in the morning and late at
night—to contact respondents who worked away from
home. The achieved sample size was 3,600 (2,400 in ru­
ral areas and 1,200 in the town), with a contact and re­
sponse rate of more than 95 percent; only five women
refused to be interviewed.
The interviewers were women who had degrees in
the social sciences or related subjects. They were famil­
iar with the local culture and had prior experience in
conducting similar surveys. They underwent extensive
training to ensure that they thoroughly understood all
the questions and were able to administer them uni­
formly without hesitation. During the initial stages, sev­
eral senior project personnel accompanied them in the
field. The survey was initiated formally only when an
excellent rapport had been established with the study
population. An experienced survey specialist provided
supportive supervision to the interviewers, whose work
was checked on a day-to-day basis. The supervisor met
with the group every evening to discuss the problems
and discrepancies that had arisen during the day. Dis­
crepancies were rectified by making a repeat visit to the
household. The completed questionnaires were again
checked and edited in the office by qualified and expe­
rienced staff before they were coded and processed.
Most of the interview was devoted to identifying
pathways by which the mother's education influences
the health and survival of her child. Questions concern­
ing respondents' reported symptoms of reproductive
morbidity formed one of the 14 sections in the question­
naire. To compose the questions, an experienced female
obstetrician/gynecologist had been asked to prepare a
comprehensive list of reproductive morbidities (both
obstetric and gynecological), including complete details
of symptoms for each condition in everyday language
that the women being interviewed could understand.
This list was thoroughly pretested and translated into
Kannada, the local language. The interview lasted for
about one hour, of which about 10 minutes were devoted
to the section on symptoms of reproductive morbidity
and its treatment. The questions were not framed to as­
sess the occurrence of medically determined gynecologi­
cal problems. Instead, they elicited women's perceptions
that they were experiencing symptoms of biomedically
defined morbidities. The disorders associated with the

self-reported symptoms are as follows:
Menstrual disorders: Heavy or light irregular
bleeding, painful menstruation, or spotting be­
tween periods.

Dyspareunia: Pain during intercourse.

Hemorrhoids: Pain or bleeding while passing
stools.
Prolapse: Feeling of something (a mass or swell­
ing) coming from the vagina, or leakage of urine
when coughing or sneezing.

Fistula: Constant leaking of feces or urine from
the vagina.
Lower reproductive tract infection: White or col­
ored discharge from the vagina with bad odor,
itching, or irritation.
Urinary tract infection: During the three months
prior to the interview, abnormal frequency of
urination, with burning sensation while passing
urine.

Acute pelvic inflammatory disease (PID): Lower ab­
dominal pain or vaginal discharge with fever.
Infertility: Inability to become pregnant despite
attempts.
Anemia: Feeling excessively weak, tired, or
breathless during normal household activities.
No claims are made in this study about the diag­
nostic accuracy of these symptoms. The relationships
between self-reported symptoms and clinically verifiable
conditions in this study are uncertain. Some of the symp­
toms reported in the survey may not be gynecological
in nature. However, the symptom categories are sug­
gestive of the corresponding medical conditions and, in
terms of clinical practice, warrant referral for detailed
examination and laboratory testing. Regardless of the
imprecise correspondence between these women's re­
ported symptoms and medically verifiable conditions,
their perceived ill health is important in its own right,
because it determines their health-seeking behavior.
Moreover, the results of treatment strongly suggest that
most women do not regard these symptoms as trivial
or inconsequential.
The questions about pain during intercourse were
dropped because information on this topic proved dif­
ficult to collect. The questions about other symptoms
were included in the main questionnaire.
The prevalence of different types of symptoms in
the study population is first described by frequency dis­
tributions and then cross-tabulated by duration and by
the treatment or consultation sought for each. Subse­
quently, the four symptom groups most commonly re-

ported by the respondents were selected for further
analysis. This phase of the analysis was carried out in
several stages. Initially, statistically significant variations
in the reporting of symptoms between women of dif­
ferent socioeconomic, cultural, and demographic back­
grounds were documented by bivariate analysis. For
women reporting symptoms, a parallel analysis was
done on the proportions seeking treatment or consulta­
tion for each related condition. Finally, logistic regres­
sion analysis was performed in order to estimate the net
effect of each factor on the likelihood of reporting
specific types of symptoms and on the probability of
seeking treatment or a consultation. All variables were
grouped or categorical in nature; for each variable one
was selected as the reference category. Results are pre­
sented in terms of odds ratios and associated p-values.
The independent or predictor variables selected for
the analyses fall into four main groups. The first group,
representing the social and economic background of the
women, includes rural-urban residence; religion and
caste; years of schooling; and the economic status of the
household. No precise theoretical justification is given
for this selection, but clearly each factor might exert an
influence on the reporting of gynecological symptoms
and treatment, through proximate pathways.
The second group of factors is demographic in na­
ture, and most of the characteristics are derived from
the complete pregnancy history that formed part of the
interview. The current age of the respondent, her age at
first pregnancy, total number of pregnancies, and his­
tory of stillbirths and abortions are included. More than
90 percent of abortions were reported to be spontane­
ous rather than induced. Clinical reasons exist for ex­
pecting associations between these factors and gyneco­
logical morbidity, although the direction of causation
remains ambiguous in some cases. For instance, an abor­
tion may be the outcome of prior infection or a cause of
subsequent infection.
The third Hock of factors is related to the woman's
last live birth. Because of the nature of the study popu­
lation, all women in the sample had experienced at least
one live birth in the preceding 60 months. The median
time elapsed since the last birth was 29 months. Because
of the interval between the construction of the sampling
frame and the interviewing, no respondent was still in
the six-week immediate postpartum period when inter­
viewed, and thus the distinction between obstetric and
gynecological symptoms is maintained. In this group of
variables are duration since last birth, place of delivery,
and whether the respondent had experienced any ob­
stetric problem in connection with the last birth. Cur­
rent contraceptive use is also included in this group.
Place of delivery is relevant to the purposes of the analy­
sis because of possible variations in the risk of infection.
Volume 26

Number 4 July/August 1995 205

Obstetric problems are included because their relation­
ship to gynecological symptoms can be assessed, and
contraceptive use is of interest because other studies
have suggested a link between certain contraceptive pro­
cedures and infection of the reproductive tract.
The last group of independent variables consists of
cognitive and behavioral factors. For each respondent,
scores on the following dimensions were derived: per­
sonal hygiene; household environment and sanitation;
exposure to health information; and autonomy. The defi­
nitions of these variables, together with that of economic
status, are shown in the Appendix. Although the vari­
able "household environment and sanitation" was in­
cluded in the bivariate analysis, it was omitted from the
logistic regression because of its close statistical asso­
ciation with personal hygiene. Furthermore, the indica­
tor of autonomy is used only in the analysis of treatment.

Findings
Approximately one-third of the women included in this
study reported current symptoms suggestive of at least
one type of reproductive morbidity. As shown in Table
1, menstrual problems were reported by 7 percent, and
this percentage rises to 11 if currently pregnant and
amenorrheic women are excluded from the calculations.
Symptoms that may indicate lower reproductive tract
infections (17 percent) and anemia (23 percent) were also
commonly reported, but symptoms that may be associ­
ated with acute pelvic inflammatory disease (5 percent)
were less common. All other symptom categories were
reported infrequently: hemorrhoids (2 percent); urinary
tract infections (2 percent); prolapse (0.4 percent); and
infertility (0.2 percent). Because all women included in
this sample had at least one living child younger than
five years old, the six infertility cases reported were of a
secondary nature.
Large proportions of these women, ranging from 20

percent to about 50 percent, reported theiAymptoms
to have lasted for more than one year. The mean dura­
tions of specific symptoms lie between 12 months for
symptoms associated with prolapse and 26 months for
white/colored vaginal discharge. These estimated du­
rations should be interpreted cautiously, because mor­
bid episodes of short duration will be underrepresented
in any analysis of current perceived morbidity.
With the single exception of infertility, the propor­
tions who had sought treatment or a consultation are
surprisingly constant, ranging only from 43 percent to
55 percent for all other conditions, as shown in Table 2.
Again, interpretation is not straightforward, because
successfully treated episodes will not be represented and
because no attempt was made to assess women's per­
ceptions about the abnormality or severity of their symp­
toms. Nevertheless, the findings indicate that a large
proportion of women consider their symptoms to be suf­
ficiently serious to warrant medical treatment. A clear
majority of those women who had sought treatment
used private sources of medical care, while, typically,
about 30 percent had used the services available at a gov­
ernment hospital. A negligible proportion used the net­
work of health centers and subcenters available in rural
areas under the primary health-care system.
Further analysis of the determinants of reported
symptoms and their consultation or treatment is limit­
ed to the four main symptom categories—those associ­
ated with menstrual disorders, lower reproductive tract
infection, acute pelvic inflammatory disease, and ane­
mia. The results of bivariate and multivariate analyses
are shown in Tables 3A, 3B, 4, and 5, and are described
below.

Correlates of Self-reported Symptoms of Morbidity
Socioeconomic Factors
The bivariate analysis reveals a pattern of socioeconomic
differentials in self-reported symptoms of morbidity that

Table 1 Percentage of women reporting current symptoms associated with gynecological morbidity, and percentage distribution
of duration of specific symptoms, Karnataka, India. 1993
Duration of symptom (percent)

Symptoms
associated with_______
Menstrual problems
Lower reproductive tract infections
Acute pelvic inflammatory disease
Anemia
Hemorrhoids
Urinary tract infections
Prolapse
Fistula
Infertility

206

Reporting symptom®
Percent
(N)
7.3
16.9
5.2
23.4
1.8
1.5
0.4
0.3
0.2

(263)
(606)
(187)
(841)
(63)
(54)
(15)
(9)
(6)

<3
months

4-6
months

7-12
months

13+
months

Total

Mean
duration
(months)

20.1
15.7
19.3
22.7
36.5
35.2
53.3
22.3
33.3

22.1
14.7
11.2
18.4
15.9
18.5
13.4
11.1
16.7

15.2
17.3
23.5
22.1
15.9
9.3
13.4
44.4
0.0

42.6
52.3
46.0
36.8
31.7
37.0
19.9
22 2
50.0

100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0

20.8
26.5
23.4
17.4
16.4
24.4
11.9
14.9
22.1

Studies in Family Planning

- - - ' - —r

Table 2 Percentage of women who sought treatment or consultation for specific symptoms, and percentage distribution of type
and place of treatment or consultation, by symptom, Karnataka, India, 1993
Se«klng consultation,
among those
with symptoms

Symptoms associated with

Menstrual problems
Lower reproductive tract infections
Acute pelvic inflammatory disease
Anemia
Hemonhojds
Urinary tract infections
Prolapse
Fistula
Infertility

Percent

(N)

43.0
52.8
50.3
48.3
44.4
42.6
53.3
55.5
16.7

(113)
(320)
(94)
(406)
(28)
(23)
(8)
(5)

Type/place of consultation (percent)
Govern­
ment
hospital

Primary
health
center/unlt

Subcenter

Private
doctor

Traditional
and others

Total

29.2
29.0
36.2
27.6
14.3
26.1
37.5
0.0
0.0

3.5
3.8
2.1
3.7
3.6
8.7
0.0
0.0
0.0

6.2
2.5
0.0
2.7
0.0
0.0
0.0
0.0
0.0

56.4
62.8
58.5
64.5
75.0
56.5
62.5
80.0
100.0

2.7
1.9
3.2
1.5
7.1
8.7
0.0
20.0
0.0

100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0

(D

is broadly consistent for all four conditions. As can be
seen in Table 3A, women from households of low eco­
nomic status, with fewer than six years of schooling, and
of lower-caste backgrounds are more likely to report
symptoms of illness than are other women. Urban-ru­
ral differences, however, are inconsistent in direction
and small in magnitude.
When the influence of other factors is controlled in
the multivariate analysis, most of these socioeconomic
effects lose their statistical significance, though the di­
rections of the adjusted differences remain the same.
Only two significant results were found: Higher-caste
Hindus are less likely to report symptoms suggestive
of acute PID and anemia than are women of the refer­
ence category, non-Hindus (mainly Muslims). In addi­
tion, urban-rural residence emerges as a significant
correlate of experiencing symptoms associated with
menstrual problems and anemia. In both cases, an urban
setting is associated with higher perceived morbidity.
Demographic Factors
Among the demographic predictors, age at first preg­
nancy and total number of pregnancies are consistently
related to all four symptom groups, also shown in Table
3A. The bivariate analysis indicates relationships that
are strong, monotonic, and in the expected direction. The
younger her age at first pregnancy and the greater her
number of pregnancies, the more likely a woman will
report symptoms of gynecological problems. However,
after controlling for other factors, the only link to per­
sist is that with feelings of tiredness and weakness,
though, in addition, women experiencing their first
pregnancy at the age of 25 or more are significantly less
likely to report menstrual problems. The respondent's
current age appears to be unrelated to any of the re­
ported symptoms.
Only about 10 percent of women reported a still­
birth, and a similar proportion indicated experience of
an abortion over their lifetime. Neither factor emerges

as a strong predictor of current symptoms. This nega­
tive result may reflect, in part, severe underreporting of
abortions. In the logistic regression for these character­
istics, only one significant result is found: Women who
have had at least one stillbirth are more likely to report
menstrual disorders.
Factors Related to Last Birth
As Table 3B indicates, a total of 1,474 (or 41 percent) of
all women reported disorders or problems associated
with their last live birth. Current problems were much
more common among this group of women than among
others; for instance, 22 percent reported symptoms that
may be associated with lower reproductive tract infec­
tion and nearly 9 percent, symptoms of possible acute
PID. The corresponding figures for the group who re­
ported no obstetric problems are about 13 percent and
3 percent, respectively. Given this tentative suggestion
of a strong link between obstetric and gynecological
symptoms, women with recent births might be expected
to have reported more current disorders. However, the
opposite appears to be the case: With the exception of
menstrual problems, respondents with the longest open
interval (25+ months) were more likely than women
with shorter intervals to report symptoms.
The majority (61 percent) of this study population
delivered their last birth at home. The remainder are
about equally divided between those who delivered at
a government and at a private institution. In general,
the probability of reporting current symptoms is high­
est among those who had their last baby at home, inter­
mediate among those using a government facility, and
least for women who used a private hospital or clinic.
The difference between the first two groups is small,
while the differential between private facility users and
the rest is more substantial.
To examine the relationship between contraceptive
use and gynecological symptoms, respondents were
split into three categories: those who had undergone
Volume 26 Number 4 July/August 1995

207

Table 3A Percentage of women who reported symptoms associated with gynecological morbidity and percentage who,reported
seeking treatment or consultation, by socioeconomic and demographic characteristics, according to symptom category,
Karnataka, India, 1993
Symptom categories

Characteristic

(N)*

Menstrual problems

Lower reproductive
tract Infectlona

Acuta pelvic
Inflammatory disease

Percent who
Percent
reported
who
seeking
reported treatment or
symptom consultation

Percent who
Percent
reported
who
seeking
reported treatmentor
symptom consultation

Percent who
Percent
reported
who
seeking
reported treatment or
symptom consultation

Anemia

Percent who
Percent
reported
who
seeking
reported treatment or
symptom consultation

Socioeconomic

Residence

Urban

(1.197)

12.5

39.6

15.7

54.0

Rural (r)

(2.396)

9.5

45.8

17.5

52.9

4.8
5.4

47.5

25.1

53.5*

52.0

22.6

45.5

Religion/caste

High caste

(182)

8.3’

38.5

12.1 —

56.5

1.6—

33.3

19.2—

51.4—

Middle caste

(2.127)

9 8

44.0

14.7

55.3

4.5

55.2

21.1

52.9

Lower caste

(950)

36.8

23.1

52.1

27.3

38.2

(333)

63.6

15.9

44.4

7.6
4.5

43.7

Non-Hindu (r)

14.2
10.4

53.3

28.8

52.6

45.3*
46.7
69.2

25.3—
26.8

41.2—

19.0

59.9

27.2"*
23.9
19.2

41.2—

Years of schooling
No schooling (r)

(1.887)

12.0—

38.4

19.7’**

49.7

1 -5 years

(504)

13.0

44.4

17.5

56.8

6.2—
6.2

6* years •

(1.203)

8.3

51.3

12.2

59.9

3.3

Low (r)

(1.194)

13.9”*

39.4

20.4'”

46.3—

('. -S?)

9.3

40.0

15.7

61.5

High

(1.207)

9.3

51.2

14.7

54.0

6.056
4.0

45.7

Middle

53.0

Economic status*
56.1

49.0

53.0
52.3

Demographic

Age

(370)

9.3

36.8

17.3

43.8—

3.2

41.7’*

22.7

41.0

20-24 years (r)

(1.476)

10.6

40.4

15.7

47.4

4.5

37.3

22.4

50.5

25» years

(1.747)

11.0

46.2

17.8

59.5

6.2

59.8

24.4

48.1

< 19 years

Number of pregnancies
(674)

8.8

42.9

13.5—

44.6

2.8—

42.1

15.0—

43.3

(2.516)

10.8

41.5

16.8

53.3

5.3

49.3

24.6

50.0

(403)

14.0

52.8

22.7

61.5

8.5

60.6

30.3

44.3

(1.123)
(2.304)

13.0’”

38.4

20.7—

50.9

6.1’”

49.3

31.2—

45.3

10.1

45.2

15.5

55.2

5.1

50.9

20.1

50.4

(166)

4.6

71.4

9.6

43.8

0.6

100.0

16.9

48.3

No (r)

(3.305)

10.2—

43.2

16.6

49.1

23.3

47.9

(2S8)

17.3

43.8

19.8

50.9’”
75.4

5.1

Yes

6.3

64.7

25.1

52.8

48.3—
78.6

22.7—

47.5

30.1

54.6

1

2-4 (r)
5*
Age at first pregnancy

<18 years

18-24 years (r)

25* years
History of stillbirths

History of abortions

No (r)

(3.271)

10.4

43.7

16.4

53.0

5.3

Yes

(322)

13.8

40.6

21.4

55.1

4.3

’ p< .05; ** p< .01; *”p< .001.

(r) = Reference category.

•304 women were pregnant and 838 women were amenorrheic. Thus, 31.8 percent of the women surveyed have been excluded from the analysis of menstrual
problems.

• (N) is fewer because some data are missing for this variable.

tubectomy, a much smaller group who were using re­
versible contraceptive methods, and nonusers. As the
table indicates, sterilized women were more likely to re­
port all types of symptoms, except menstrual disorders.
No clear differences, however, emerged between nonus­
ers and those using reversible methods of contraception.
These patterns of association, based on the bivari­
ate analysis, change little when they are reassessed in a
multivariate context. Thus the experience of obstetric
problems during the most recent birth remains a strong
predictor of all four types of gynecological symptoms,

208

Studies in Family Planning

with odds ratios ranging from 2.15 for menstrual prob­
lems to 4.78 for lower abdominal pain/discharge with
fever. Delivery at home or in a government institution
also remain significant risk factors compared with de­
livery in a private institution. This finding suggests that
place of delivery is not acting merely as a proxy for so­
cioeconomic status. Similarly, sterilization retains its as­
sociation with three of the four symptom categories, with
odds ratios (relative to nonusers) of 1.44 for feelings of
weakness, 1.57 for vaginal discharge, and 2.50 for lower
abdominal pain or discharge with fever.

was dropped because its close association with personal
hygiene and the effects of the remaining two factors are
much diminished. While personal hygiene remains sig­
nificantly related only to menstrual disorders and symp­
toms of possible lower reproductive tract infections, ex­
posure to health information is not a net predictor of
any of the four symptoms.

Cognitive and Behavioral Factors
The results with regard to the cognitive and behavioral
factors presented in Table 3B may be summarized suc­
cinctly. In the bivariate analysis, all three—personal hy­
giene, household environment and sanitation, and ex­
posure to health education—are related to reported
symptoms. Nearly all the associations are statistically
significant and monotonic in direction. Those classified
as having poor personal hygiene, or unsanitary house­
hold conditions, and with a low exposure to health in­
formation are more likely to report current gynecologi­
cal problems than are their counterparts. In the logistic
regression presented in Table 4, household environment

Correlates of Treatment
For reasons of symmetry, the same bivariate analyses
were performed for treatment as for reported symptoms.
Most of the results are of minor interest. Moreover, rela­
tively few relationships attain statistical significance.

Table 3B Percentage of women who reported symptoms associated with gynecological morbidity and percentage who reported
seeking treatment or consultation, by characteristics related to last pregnancy and cognitive/behavioral characteristics, according
to symptom category, Karnataka, India, 1993
Symptom categories

Menstrual problems

Characteristic

(N)‘

Lower reproductive
tract Infections

Percent who
reported
seeking
treatment or
consultation

Acute pelvic
Inflammatory disease

Percent who
Percent
reported
who
eeeklng
reported treatment or
symptom consultation

Anemia

Percent who
Percent
reported
who
seeking
reported treatment or
symptom consultation

Percent who
Percent
reported
who
seeking
reported treatment or
symptom consultation

Percent
who
reported
symptom

37.3

13.2’”

51.1

2 g...

54.1

17.5’”

44.6*

8.6

48.8

32.0

51.3

Last-pregnancy-related
Problems
8 2—

(2.119)
(1.474)

14.4

48.3

22.1

55.0

(2.212)

11.1*

41.0

18.8’”

51.1

6 2”*

46.7”

25.0*”

44.1”’

(647)

12.9

41.7

16.1

52.3

4.0

46.2

24.1

46.8

(733)

8 1

53.3

11.9

64.4

3.1

78.3

18.0

68.2

(1.172)

10.1

51.2

4.6*”

44.4

21.5*”

43.2*

13-24 months

(951)

40.0
44.9

14.3”’

10.5

15.2

48.3

3.5

53.1

20.2

46.1

25* months

(1.470)

11.1

43.5

19.9

56.8

6 8

53.0

27.0

52.8

(1.263)

12.1

44.7

20.3’”

57.3

8.3’”

50.5

28.3*”

51.5*

(258)

7.9

65.0

13.9

55.6

2.3

83.3

16.2

60.0

(2.072)

10.1

38.1

15 1

49.5

37

48.0

21.3

44.7

Poor (r)

(1.317)

12.8’”

37.9

6.3”

46.4

25.5”

38.8”*

(1.100)

11.4

43.2

56.2

5.5

50.0

23.7

50.8

Good

(1.176)

8.4

50.6

20.7’”
17.0
12.5

48.5

Fair

58.1

3.7

59.5

20.7

59.0

No (r)
Yes

Place of delivery
Home

Government institution
Private institution (r)

Time since last birth

<13 months (r)

Contraception

Tubectomy
Users of reversible methods

Nonusers (r)
Cognitive/behavioral
Personal hygiene of woman

Household environment
and sanitation
Poor (r)

(1.344)

13.0—

37.4’

20.8’”

49.1

7.1”*

49.5

27.6*”

39.7”*

Fair

(1.049)

11.3

17.8

46.0

23.0

52.9

Good

8.3

11.7

55.1
58.7

5.0

(1.200)

40.0
54.4

3.3

59.0

19.1

57.7

Poor(r)

(1.191)

12.4

31.8’

19.7”*

45.7"*

6.8 —

36.3’”

26.3**

36.2—

Fair

(1.212)

10 8

45.5

16.6

49.5

4.6

63.2

22.3

50.4

Good

(1.190)

9.4

52.3

14.4

67.8

4.2

59.2

21.7

61.9

Poor (r)

(1.268)

na

41.5

na

45.3

na

41.6—

(1.196)

na

46.3

na
na

45.1'

Fair

55.9

na

52.0

na

48.0

Good

(1.129)

na

42.3

na

59.9

na

53.4

na

58.4

Exposure to health education

Autonomy

* p< .05; ” p< .01; *”p< .001;
(r) = Reference category.
na = Not applicable.
•304 women were pregnant and 838 women were amenorrheic. Thus, 31.8 percent of the women surveyed have been excluded from the analysis of menstrual
problems.

Volume 26

Number 4 July/August 1995

209

Table 4

Logistic regression of common symptom categories, by women's characteristics, Karnataka. India, 1993
Odds ratios of symptom categories

Characteristic

Reference category

Menstrual
problems

Lower
reproductive
tract
Infections

Acute pelvic
Inflammatory
disease

Anemia

Socioeconomic

Residence

Rural

Urban

1.60—

1.01

1.06

1.11

0.82
0.83

0.39’

0.70

1.05

0.88

1.36

1.28

1.43

0.68—
0.83

1-5 years

1.21

0.98

6* years

1.23

1.09

0.89

0.77

0.85

0.86

Middle

0.74

High

0.85
0.84

1.26

0.86

0.85

0.99
0.84

< 19 years

0.80

1.14

25* years

0.92

1.06

1.14

1.03

1.18

0.99

Caste/religion

Higher caste
Middle caste
Lower caste

Education

Economic status

None

Low

Demographic
Age

Age at First pregnancy

20-24 years

18-24 years

< 18 years

1.22

25* years

1.09

0.87

1.53”

0.42'

0.68

0.15

0.90

Number of pregnancies

2-4

1

5+
History of stillbirths

1.14

1.02

0.95

0.65—

0.93

1.16

1.47

1.07

l^*’

1.12

1.04

0.97

1.24

1.27

0.62

1.12

2.15—

2.33—

4.78—

2.88—

1.42

1.55—

2.19"’

1.68—

1.66’

1.37’

1.37

1.55—

No

Yes
History of abortions

No

Yes
Last-pregnancy-related

V

1.22’

Non-Hindus

Problems

No

Yes

Place of delivery

Private institution

Home
Government institution

Months since last birth

< 13 months

13-24 months

1.05

25* months

1.19

0.89

1.01

0.99

1.55—

1.50'

1.34—

1.31

1.57'”

0.95

1.28

2.50—
0.94

1.44 —
0.83

1.00
0.94

0.94

Contraception

Nonusers

Sterilized (tubectomy)

Users of reversible methods

Behavloral/cognltive

Personal hygiene

Poor

Fair
Good

Exposure to health education

0.86

0.87

0.67’

0.74'

Poor

0.94

Fair

1.02

1.03

Good

0.77

1.04

1.08

0.75

0.93
0.93

-3.10

-2.54

-4.80

-2.07

Constant
*P< .05; ”p<.01; •••p<.001.

partly because of the small number of observations.
However, more marked variations are seen in the pro­
portions seeking treatment for the symptoms associated
with anemia (tiredness, weakness, breathlessness) than
for the other three categories.
The multivariate analysis of treatment is contained
in Table 5, where the main interest is focused on the in­
fluence of socioeconomic and related factors. Two mod­

210 Studies in Family Planning

els were used. In the first, the effects of the four socio­
economic factors are assessed. In the second model, an
additional four variables are added. Two of these (du­
ration of problem and respondent's age) may be viewed
as controls. The other two, exposure to health education
and autonomy, represent possible pathways through
which the socioeconomic factors might influence the
probability of seeking treatment. Extent of formal school-

Table 5 Logistic regression of probabilities of seeking treatment or consultation for common symptom categories, by women’s
characteristics , Karnataka, India, 1993___________ _
Odds ratios of symptom categories

Raference
category

Lower
reproductive
tract Infections

Menstrual
problems

Acute pelvic
inflaiT'*-jtory disease

Anemia

Model 2

Model 1

Model 2

Model 1

Model 2

Model 1

Model 2

Model 1

0.70

0.58

1.15

0.97

0.62

0.86

1.25

1.09

High caste

0.26’

0.13*’

1.52

1.09

1.73

1.52

1.15

1.06

Low caste

0.30*

Q.23”'
0.22—

0.66
0.85

0.79

0.34’

0.56
0.87

0.89

Middle caste

1.69

1.67

0.61

0.59

0.76

0.74

1-5 years

1.21

1.07

1.34

1.34

0.97

0.61

1.48’

. 1.26

6* years

1.63

1.25

1.40

1.09

2.47’

1.81

1.74”

1.32

Middle

0.96

0.82

1.79*

1.46

1.43

1.05

1.41’

1.21

High

1.52

1.33

1.28

1.13

0.90

0.83

1.44’

1.29

na

2.08’

na

1.04

na

3.19—

na

1.40

na

2.92”’

na

2.20—

na

2.21

na

1.88—

Medium

na

1.33

na

1.43

1.23

na

na

1.58

na
na

1.23

na

1.28
1.47

na

High

< 20 years

na

1.13

na

1.16

na

0.89

na

1.27

na

1.70—

na
na

1.06

25* years

2.79’”

na

0.85

na
na

0.98

na

1.54

na

1.81

na

1.05

1.89

na

2.35’”

na

2.07

na

1.90’”

0.57

-1.0

-0.77

-0.13

-1.44

-0.57

Characteristic

Socioeconomic
Residence

Rural

Urban
Casta/reiigion

Education

Economic status

Non- Hindus

None

Low

Other

Exposure to health
education
Medium

Low

High

Autonomy

Age

Duration of the problem
13-24 months

25* months

Constant
• p< .05; ’* p< .01; *"p< .001.

Low

1.35

20-24 years

< 13 months

-1.56

-0.79

na = Not applicable.

ing, for instance, may be related to exposure to health
information and to autonomy, which in turn may affect
the likelihood of seeking medical treatment. Compari­
son of the results of the two models thus allows a de­
composition of the effect of the socioeconomic factors.
With regard to the results, the dependent variable,
treatment, is a simple dichotomy: that is, any treatment
versus no treatment. The most common source of treat­
ment is a private medical practitioner. These practitio­
ners are not necessarily qualified in allopathic or ayur­
vedic medicine. In India, as elsewhere, practitioners vary
widely in terms of qualifications, skills, and therapeutic
principles. Most large villages in India will possess a pri­
vate practitioner of some sort (Chuttani el al., 1973). In
this study, no attempt was made to elicit information
about the qualifications of practitioners.
The wide dispersion of sources of medical advice
and treatment, in addition to the limitations of the
sample, may account for the mostly small urban-rural
differences in treatment-seeking; these are generally in­
consistent in direction and not statistically significant.

The influence of religion is particularly fascinating.
All the Hindu castes are more likely to seek treatment
for white/colored vaginal discharge (suggestive of lower
reproductive tract infections) than are Muslims. Al­
though the Model 1 odds ratios are large, they do not
attain statistical significance. For menstrual disorders,
the differences are large and highly significant, and ac­
tually increase when the additional four variables are
added (Model 2). All Hindu castes are much less likely
to seek treatment for this type of disorder than are Mus­
lims. For other symptom categories, differences are
smaller and, again, not significant.
The results for education and economic status are
broadly similar. As expected, better-educated women
from more affluent households are more likely to seek
treatment for symptoms of gynecological problems than
are their less privileged counterparts. The effects, how­
ever, are not pronounced, and, with the exception of
symptoms associated with anemia, few of them are sta­
tistically significant. Odds ratios systematically attenu­
ate when exposure to health education and other fac-

Volume 26

Number 4 July/August 1995

211

tors are entered into the logistic regression. Indeed, none
of the effects of education or economic status remains
significant in Model 2.
Exposure to health education emerges as a major
predictor of therapy-seeking behavior, though interpre­
tation should be cautious. Reverse causation is a possi­
bility: Women with problems may be more likely to seek
out, or be alert to, relevant information presented on the
radio or in the press.
The relationships between women's autonomy and
the probability of their seeking treatment are all in the
expected direction, but are not pronounced nor of sta­
tistical significance. Finally, older women are more likely
than younger women to seek treatment (except for the
symptoms of possible anemia), and women who have
been suffering from the specified symptoms for a long
time have a higher probability of seeking treatment than
do those for whom the onset of symptoms is more recent.

Discussion
Despite the localized nature of this study, the socioeco­
nomic profile of survey respondents is similar to that of
Karnataka as a whole, and no grounds are found for be­
lieving that the study population is atypical of rural cen­
tral-south India.
In the past, the usefulness of data on self-reported
morbidity has been questioned. Though medical scien­
tists naturally prefer to estimate the prevalence of spe­
cific diseases through clinical examinations and labora­
tory tests, such procedures are usually too expensive to
be carried out on a large scale in community settings. In
the Danfa project in Ghana, for instance, medical exami­
nations were found to be eight times more costly than
interviews (Belcher et al., 1976). Furthermore, in most
developing countries, women are generally reluctant to
subject themselves to gynecological examination, par­
ticularly when they have no apparent symptoms of dis­
ease. In the few small-scale studies of reproductive mor­
bidity that have been conducted, the refusal rate was
found to be high, even when clinical examination was
limited to symptomatic women (Bang et al., 1989; Wasserheit et al., 1989). The multicountry study conducted
under the auspices of the World Health Organization
also encountered high refusal rates (Omran and Standley, 1976 and 1981). A rare exception is a recently re­
ported study in Egypt (Younis et al., 1993). High refusal
rates introduce severe selection biases and, in such cir­
cumstances, comparisons of the responses from inter­
views with the results of clinical examination and medi­
cal tests are difficult to interpret. Traditionally, doctors
have depended to a large extent on a patient's history
for reaching a diagnosis, but, with advancements in

212

Studies in Family Planning

medical technology, more reliance is now placed on so­
phisticated tests requiring expensive instruments that
can be performed only in clinical settings. Until simple,
inexpensive, and accurate screening procedures are de­
veloped that can be used in field situations, the contri­
bution of clinical examination and laboratory tests for
detecting reproductive morbidity among representative
samples will remain limited.
Health-care facility records are valuable because
they indicate what services are being sought and by
whom. However, they are much less suitable for esti­
mating the magnitude of reproductive health problems,
because a large proportion of women having such prob­
lems does not visit these facilities. In the face of these
limitations, the interview survey remains one of the few
options for gauging the extent of women's health prob­
lems. If the instruments for these surveys are developed
through the joint efforts of medical professionals, social
scientists, demographers, and biostatisticians, much use­
ful information can be gathered for measuring perceived
morbidity at the community level. The review by
Campbell and Graham (1990) comes to the same con­
clusion. Even though the correspondence between self­
reports and clinical diagnosis is far from exact, the rel­
evance of symptoms reported by women is in no doubt,
because the recognition of a disorder, its perceived
causes, and feared outcome, rather than a sophisticated
medical diagnosis, determines a patient's reaction and
treatment-seeking behavior. Certainly, some conditions
and infections are asymptomatic or exist in so early a
stage that the patients do not recognize them; equally,
conditions arise that are psychosomatic in origin or that
medical professionals do not recognize. But if people
seek care for imaginary illnesses and are willing to pay
for it, medical professionals abound who are willing to
treat them, even when no biomedical justification is
found for such care. As a measure of potential demand
for services, self-reported symptoms of morbidity are,
therefore, more applicable than clinical diagnoses.
Health education may increase, in the long run, the con­
sistency between an individually determined and medi­
cally determined need.
Approximately one-third of women in this study re­
ported the symptoms suggestive of at least one kind of gy­
necological morbidity'. The exclusion of older women with­
out young children may have acted to depress this
estimate. Nevertheless, the results indicate a high preva­
lence of perceived illness or disorder. While no infor­
mation was collected on the severity of the reported prob­
lems, the finding that about half of those reporting
symptoms had sought treatment clearly suggests that
the majority of symptoms were not regarded as minor
or inconsequential.

Approximately one-tenth of menstruating women
reported menstrual problems. Menstruation is, of course,
a normal occurrence and an integral part of women's
lives. Presumably, women would report menstrual prob­
lems only if they perceived them to be severe. Quoting
data from the National Center of Health Statistics, Koblinsky et al. (1993) claim that menstrual dysfunction and
other abnormal vaginal bleeding are also the primary
diagnoses for 350,000 hospitalizations per year, and dys­
functional uterine bleeding has been found to be one of
the most important reasons for hysterectomy in the United
States.
The incidence of infection of the reproductive tract
in developing countries has been found to be remark­
ably high, owing to a combination of biomedical, behav­
ioral, and societal factors (for example, see Wasserheit,
1989). These infections, if not diagnosed early and trea­
ted promptly, may represent a serious long-term threat
to women's health, fertility, and productivity. The sur­
vival and health of infants born in these countries may
be seriously compromised, and such infections may ad­
versely affect the appeal of family planning programs.
In this study, approximately one-sixth of the women re­
ported white or colored vaginal discharge, which is of­
ten associated with lower reproductive tract infections
as well as with no diagnosable problem. A further 5 per­
cent reported symptoms possibly indicative of acute
PID, namely lower abdominal pain or vaginal discharge
with fever. Self-reports of these latter symptoms in our
study are low, compared with their reported prevalence
in other studies (Bang et al., 1989). One of the important
risk factors for PID is sexually transmitted diseases
(STDs). Because, in India, premarital and extramarital
sexual relations probably are not as widespread as they
are in many other countries, the prevalence of STDrelated pelvic inflammatory disease may be low. Con­
versely, PID related to postabortion complications may
be relatively high among older women who have ter­
minated unwanted pregnancies. The incidence of abor­
tion, both spontaneous and induced, is difficult to esti­
mate because of reporting problems. A few studies
indicate that a large number of abortions in India are
performed by unqualified medical practitioners (Bhatia
and Ramaiah, 1971; Bhatia, 1973). Such abortions not
only result in a great number of maternal deaths but also
may cause severe long-term disabilities, including PID,
among surviving women.
The other symptoms most commonly reported by
women were excessive weakness and a feeling of breath­
lessness and tiredness during normal household activi­
ties. These may be symptoms of anemia, and were re­
ported by about one-fourth (23 percent) of respondents.
A multicenter study conducted by the Indian Council

of Medical Research (ICMR) in six centers in India dur­
ing 1986-87 found that 62 percent of pregnant women
had hemoglobin levels of fewer than 11 grams per deci­
liter (ICMR, 1992). World Health Organization estimates
for 1985 suggest that 59 percent of pregnant women and
47 percent of nonpregnant women in developing coun­
tries (excluding China) suffered from anemia, and the
prevalence of this condition was found to be very high
in South Asia and Africa (DeMaeyer and Adiels-Tegman, 1985). Although precise measurement of anemia
is not possible without laboratory tests of hemoglobin
or hematocrit levels, the symptoms the women reported
during their interviews are suggestive of this condition,
and have been labeled here as anemia for illustrative
purposes only.
Logistic regression analysis of the determinants of
reported symptoms of gynecological morbidity revealed
the existence of three strong, pervasive influences, each
of which has potentially far-reaching policy implications.
The first of these factors is the experience of obstetric
problems and complications associated with the last live
birth. Respondents were questioned in detail about prob­
lems that arose during pregnancy, delivery, and the puerperium. The most commonly reported problems were
excessive bleeding and lower abdominal pain in the six
postpartum weeks. Forty-one percent of all respondents
reported at least one problem. This group had a much
higher likelihood of self-reported gynecological symp­
toms at the time of the survey than did women who re­
ported no obstetric problem. Several explanations may
be presented for this strong link. Both recent obstetric
and current gynecological disorders may have a com­
mon etiology. Alternately, the results may be an artifact
of different reporting styles; some women may be par­
ticularly sensitive to minor problems and are thus more
likely to mention them during an interview, whether the
question relates to the last pregnancy or the current situ­
ation. Finally, obstetric problems may be a direct or con­
tributory cause of subsequent morbidity. On the as­
sumption that differential reporting is unlikely to be the
major explanation, these results carry the important im­
plication that obstetric problems can act as a warning
sign of more persistent problems of reproductive ill
health. In some settings, targeting follow-up diagnosis
and treatment for these women may be feasible.
The second major risk factor for reported gyneco­
logical complaints concerns the place of the last deliv­
ery. Institutional deliveries were common in this study
population: More than one-third (38 percent) of last de­
liveries had taken place in a hospital or clinic, with ap­
proximately equal proportions in public- and in private­
sector institutions. The analysis revealed a clear-cut
advantage to delivering at private institutions in terms

Volume 26 Number 4 July/August 1995 213

J

of the likelihood of subsequent self-reported problems.
Differences were not large but were consistent for all
four main symptom categories, and most were statisti­
cally significant. For home deliveries, the adjusted odds
ratios range from 1.42 (for menstrual disorders) to 2.19
(for lower abdominal pain or discharge with fever), rela­
tive to private institutional care. For women delivering
in government facilities, the odds ratios are in the range
of 1.37 to 1.66. Thus, little difference appears in the risks
associated with home or public-sector delivery. As noted
earlier, these results were obtained after adjusting for
place of residence, for socioeconomic status, and for the
occurrence of obstetric problems associated with the last
birth. While a direct causal link cannot be proved, the
results strongly suggest that the quality of care, and, in
particular, hygienic conditions, may be lower in govern­
ment hospitals than in private hospitals and clinics. In­
deed, the data show that delivery in a government hos­
pital may offer little advantage over home delivery in
terms of protection against infection.
The third, perhaps the most important, finding of
this study is that reporting of all symptom categories,
except menstrual problems, is significantly higher among
women who have undergone tubectomy than among those
who are not using a method of contraception or who
are using a reversible method. This finding is consistent
with other studies conducted in South Asia (for example,
see Bang et al., 1989, and Wasserheit et al., 1989). In the
case of reported s^Tnptoms of possible anemia, a direct link
is biomedically implausible, and the results may reflect a
widespread belief in South Asia that the sterilization of
women (or men) leads to weakness. The statistical asso­
ciation between tubectomy and the symptoms of infec­
tion are more plausible, because procedures associated
with tubectomy may introduce infections if high stan­
dards of hygiene are not maintained. One alternative in­
terpretation is that sterilized women are more prone
than nonsterilized women to report minor symptoms;
this possibility was assessed by a special analysis of the
link between sterilization and obstetric problems asso­
ciated with the last live birth. Women who had been ster­
ilized subsequently were found to be no more likely to
report symptoms of obstetric problems than were other
women. Although no general psychological difference
is seen between the two groups in the reporting of symp­
toms, the result may, nevertheless, be an artifact of dif­
ferential reporting.

Conclusion
The results of this study, if substantiated by more de­
tailed biomedical studies and by clinical examinations.

214

Studies in Family Planning

have far-reaching implications for India's family plan­
ning program. The program emphasizes feiTiale steril­
ization, and the prevalence of reversible methods of con­
traception is very low. Since 1983, more than four million
sterilizations have been performed annually, most of
them with minilaparotomy and laparoscopic techniques.
In Karnataka, minilaparotomy accounts for about 80 per­
cent of such procedures. Because these sterilizations can
be performed as outpatient procedures, even in rural set­
tings, and because a large number may be performed
by a single doctor in a day, maintaining proper sanitary
conditions is often difficult (Mehta, 1989). Furthermore,
the cases of sterilized women are seldom followed up,
especially on a long-term basis, and thus little is known
about the incidence of reproductive problems among these
women. Further studies in this area are urgently needed.
Approximately half of the women reporting symp­
toms sought treatment or consultation for these prob­
lems. A majority of such consultations were with pri­
vate medical practitioners. The number of unqualified
private medical practitioners practicing allopathic (west­
ern) medicine is rapidly increasing in the rural areas and
small towns of India. People willingly pay for their ser­
vices rather than availing themselves of free services at
the government health facilities (Bhatia et al., 1975).
Women rarely use the primary health centers and sub­
centers. This underutilization has been described in many
other studies (Johns Hopkins University, 1976; Chuttani
et al., 1976; Kanitikar and Sinha, 1989). A radical review of
facilities available under the primary health-care system
is required, along with a more systematic evaluation of
the private medical sector.

Appendix
Construction of Variables
Economic Status
This variable was determined on the basis of the im­
puted financial value of consumer durables, such as ra­
dios, televisions, fans, refrigerators, furniture, washing
machines, bicycles, two-wheeled and four-wheeled mo­
tor vehicles, and agricultural implements, such as trac­
tors and thrashers. After calculating the total monetary
value of these possessions in each household, the house­
holds were divided into three categories of approxi- »
mately equal size.

Household Environment and Sanitation
Information was collected on the availability and type
of several household facilities used for bathing, cook­
ing, washing, drainage, and so forth. Data were also col­
lected about household practices, including laundering.

disposal of waste, storage of drinking water and food,
and frequency of cleaning places that are occupied by
people and animals. Based on responses, a numerical
environment and sanitation score was derived, and the
households were divided into three categories.

1

Personal Hygiene
Each respondent was asked about her personal hygiene
practices, including bathing, washing and combing her
hair, washing and changing her clothes, clipping her
nails, and washing her hands after defecation and be­
fore meals. She was asked what materials she uses for
bathing and washing. The responses were numerically
scored, and the respondents divided into three groups.
Exposure to Health Education

The extent of the respondent's exposure to the media
and to informal adult and health-education programs
was ascertained by asking her a series of questions about
the type, duration, and content of the programs she has
attended and about her preferred radio and television
programs and reading habits. A numerical score was as­
signed to each response, and a total score for the respon­
dent was computed. The respondents were divided into
three groups.

Autonomy
Each respondent was asked several questions to obtain
information about her status in the household. These
questions pertained to 1) her freedom to make economic
and financial decisions; (2) her mobility; (3) her commu­
nication with her husband about sensitive matters; and
(4) her active involvement in important household af­
fairs. The responses elicited were numerically scored,
and the women were divided into three categories.

Notes
i

The major exception to this generalization is the World Health
Organization multicountry study of family-formation patterns
and health conducted more than 20 years ago (Omran and
Standley, 1976 and 1981).

2

In a similar study in rural Bangladesh, self-reported gynecologi­
cal morbidity was much lower, but, again, a link with contracep. tive use was found. Users of intrauterine devices and sterilized
women were each four times more likely than nonusers to re­
port symptoms and se’. en times more likely than nonusers to
have infection confirmed by examination (Wasserheit et al., 1989).

Acknowledgments
Financial support for this study was provided by the Ford
Foundation, which is gratefully acknowledged. The authors
would also like to put on record the hard and diligent work

of N.S.N. Rao in the organization and supervision qf the field­
work, and the able assistance provided by N. Sanath Kumar
and S. Ramaswamy in the data processing. The interviewers,
who collected data under difficult field conditions, deserve
our appreciation and commendation. Thanks are also due to
the Karnataka Health and Integrated Child Development Ser­
vice officials for their full support and cooperation in the con­
duct of this study.

References
Bang, R., A. Bang, M. Baitule, Y. Chaudhary, Y. Sarmukaddams, and
O.T. Tale. 1989. "High prevalence of gynecological diseases in
rural Indian women." The Lancet 1: 85-88.

Belcher, D.W., F.K. Warupa, A.K. Neumann, and I.M. Lourie. 1976.
"Comparison of morbidity interviews with a health examination
survey in rural Africa." American Journal of Tropical Medicine and
Hygiene 25,5: 751-758.
Bhatia, Jagdish C. 1973. "Abortionists and abortion seekers." Indian
Journal of Social Work 34,3: 275-285.

Bhatia, Jagdish C. and T.J.Ramaiah. 1971. "Incidence of induced abor­
tions in a community development area." Social Action 21,3: 224232.
Bhatia, Jagdish C., Vir Dharam, C.S. Chuttani, and A. Timmappaya.
1975. "Traditional healers and modem medicine." Social Science
and MedicineS: 15-21.
Campbell, Oona M.R. and Wendy J. Graham. 1990. "Measuring Ma­
ternal Mortality and Morbidity: Levels and Trends." Maternal and
Child Epidemiology Unit Research Paper No. 2. London School of
Hygiene and Tropical Medicine.

Chuttani, C.S., Jagdish C. Bhatia, Vir Dharam, and A. Timmappaya.
1973. "A survey of indigenous medicine practitioners in rural ar­
eas of five different states in India." Indian Journal of Medical Re­
search 61,6: 962—967.
-------- . 1976. "Factors responsible for underutilisation of primary
health centres: A community survey in three states of India." Jour­
nal of Health Administration (NIHAE Bulletin) 9,3: 229-237.

DeMaeyer, E. and M. Adiels-Tegman. 1985. "The prevalence of
anaemia in the world." World Health Statistics Quarterly 38: 302316.
Indian Council of Medical Research (ICMR). 1992. Field Supplementa­
tion Trial in Pregnant Women. An ICMR Task Force Study. New
Delhi: ICMR.
Johns Hopkins University. 1976. Functional Analysis of Health Needs
and Services. New Delhi: Asia Publishing House.

Kanitikar, T. and R.K. Sinha. 1989. "Antenatal care services in five
states in India." In Population Transition in India. Vol. 2. Eds. S.N.
Singh et al. Delhi: B.R. Publishing.

Koblinsky, Marjorie A., Oona M.R. Campbell, and D. Harlow. 1993.
"Mother and more: A broader perspective on women's health."
In The Health of Women:A Clobal Perspective. Eds. Marjorie A.
Koblinsky, J.Timyan, and J.Gay. Oxford, England: Westview
Press.
Mehta, P.V. 1989. "A total of 250,136 laparoscopic sterilizations by a
simple doctor." British Journal of Obstetrics and Gynaecology 96:
1,024-1,034.

Volume 26

Number 4 July/August 1995 215

Omran, A.R. and C.C. Standley. 1976. Family Formation Patterns and
Health. Geneva: World Health Organization.
-------- . 1981. Family Formation Patterns and Health—Further Studies.
Geneva: World Health Organization.

Wasserheit, Judith N. 1989. “The significance and scope of reproduc­
tive tract infections among third world women." International
Journal of Gynaecology and Obstetrics Supplement 3:145-168.

Kay, and Karen J. Mason. 1989.- ?.a!?rodoctive tract Weetions in
in Fama family planning population in r-ra- Bangladesh.’
Uy Planning 20,2: 69-80.
Younis, Nabil, Hind Khattab, Huda Zurayk. Mau-aheb B-Mouelhy,
Mohamed Fadle Amin, and AWe! Mor.edm Farag.
’A community study of gynecological a.-d relaced morbidities in rural
Egypt." Studies in Family Planning 24,3: l>a-lS6.

1
!

Wasserheit, Judith N., Jeffrey R. Harris, J. Chakraborty, Bradford A.

i

1
r
7

t
1

1

216

Studies in Family Planning

■i

CO H \

Gynecological Morbidity in Rural Gujarat:
Some Preliminary Findings

Leela Visaria

Gujarat Institute of Development Research
Ahmedabad, India

S

Introduction
After the International Conference on Population and Development held at Cairo in September 1994, the
Government of India removed the method-specific contraceptive targets from its family planning
programme in some districts beginning April 1995 and throughout the country in April 1996. The Ministry
of Health and Family Welfare is also geared up to adopt the reproductive health approach, articulated in the
Plan of Action of Cairo Conference, with an emphasis on the quality of services. However, the available
data base on the status of reproductive health and morbidity patterns among Indian women (and men) is too
limited to be used as a background when designing and also implementing appropriate programme
interxentions.
The reasons for lack of a comprehensive knowledge and understanding of the reproductive health
needs of Indian women are many. The Indian family welfare programme until recently had been too pre­
occupied with achieving the method-specific contraceptive targets to address the issues of women’s health.
The field-based non-governmental organizations (NGOs), many of whom have been working in the area of
health, do provide reproductive health services along with general health care to the community, but rarely
document their efforts or disseminate their distilled understanding in a systematic manner. The hospital­
based statistics have their serious limitations but very little effort is made even to compile and analyze these
data. The Indian academic community has also by and large kept away from conducting research in an area
which is considered too sensitive to yield meaningful results.
In the 1990s the situation has started changing. The Government of India has not only decided to
remove the method-specific contraceptive targets but has also decided to "revamp" its family welfare
programme. Instead of the earlier singular concern of reducing the birth rate and population growth, efforts
are being made to find ways to address the health needs of the women and emphasis is being laid on
improving the quality of services provided by the programme (MOHFW, 1996). Also, preliminary results
from a few community-based studies undertaken in two tribal villages of Maharashtra state by an NGO
(Bang and Bang. 1996), a few studies undertaken in small pockets in the states of Gujarat, West Bengal and
Maharashtra. (Bhatia and Cleland, 1995; Bhattacharya, 1994; Latha and Shah, 1994; Kannani, Latha and
Shah. 1994; Mulgaonkar, et al., 1994; Parikh et al., 1996) are pointers to the culture of silence as far as the
health of Indian women is concerned. Even today, not much is known about the reproductive health of
Indian women at the national or state level.
The few studies undertaken so far raise many methodological issues and questions about the
validity and reliability of the data (Koenig, et.al., 1996). And yet, research must go on to enhance our
knowledge base not only about the incidence or prevalence of various reproductive tract infections and
sexually transmitted diseases, but also about the health-seeking behaviour of women, the socio-economic
correlates of their health status and other issues. The findings from carefully conducted research will go a
long way in providing meaningful programmatic interventions. Also, some of the methodology,
comparability and consistency issues will also get resolved with further research.
Gujarat Study : Objectives
A large prospective study to understand the interrelations between mother's education and the health
seeking behaviour with regard to her children as well as herself has been undertaken in a rural area of
Kheda district in Gujarat state by the Gujarat Institute of Development Research in Ahmedabad with the
financial support of the Ford Foundation. Among various research components of the study, one has dealt
with the issue of gynaecological morbidity among rural women in this region.
The objectives of the study were to estimate the level of gynaecological morbidity among different
socio-economic groups, to assess the extent of medical care sought by women and to assess whether and to
which extent self-reported morbidity matches with the clinical examination of women. This paper presents
the preliminary findings related to the first two objectives of the study. The analysis based on the matching
of self-reported morbidity with that based on the clinical examination of the women will be reported at a
later date.

Research Methodology and Data
The field work for this study was undertaken during 1995-96 in four rounds. In the first round, we
conducted five group discussions with women in the region to understand the description and terminologies
used by them to describe the various illnesses associated with the reproductive tract. The information
provided and the language used by the women were incorporated in the interview schedule designed to
collect information on morbidity as reported by women. The schedule was also shared and discussed with a
gynecologist to check whether the questions would yield meaningful information on various illnesses. The
schedule was prepared both in Gujarati, the language of the area in which the study was undertaken, and in
English.
The second round of data collection involved canvassing of the schedule in the community. The
schedule was canvassed among 800 women from five contiguous villages in Kheda district. The sample
was drawn from the population enumerated in the baseline survey conducted eighteen months prior to the
survey. They represent nearly half of all the ever-married women aged 19-59 in the population of the five
villages under study.
Although while the field investigators engaged in the collection of data had several years of
experience, their understanding of women's reproductive system and functions and various reproductive
tract infections was enhanced by holding two sessions by a gynecologist who held two sessions to explain
these with diagrams, discussions, etc. The investigators had also participated in the initial focus group
discussions held with the rural women.
The third round of data collection involved an actual clinical examination ot the women. The data
collected from the respondents on their self-reported illnesses, were partly verified by a clinical examination
conducted by gynecologists in the field. The examination also included obtaining cervical smear for early
detection of cancer among women in the communities. The smears were sent to the Civil Hospital in
Ahmedabad for analysis. When an abnormality was detected, the concerned woman was contacted and
arrangements weie made for further investigations. At the time of the examination also, the medical staff
provided individual counseling to women, free drugs to treat minor ailments and referral advice. We had
established links with the nearby hospitals to provide care to the women from the communities.
In the fourth round, the results of the study were shared with the community. We also followed up
on the women who were provided care during the visits of the gynecologists to the field as well as those
women who needed further care. All these activities were documented. The research study thus evolved into
an action-research project.
This paper presents data collected during the survey of self-reporting of reproductive tract
infections among women. The findings of the medical examination of women were manually compiled very
quickly and the results of the pap smears were obtained from the hospital. They were used to identify the
women who needed to be followed up for further investigations. Matching and linking of these results with
the survey data and the analysis will take some more time. The analysis of the data is in progress and will be
reported later.
Background Characteristics
As shown in Table 1, the total population of the five villages was close to 11 thousand according to the
1991 census. The average household size was 5.7 members and the average age of the respondent women
(the age ranged between 16 and 60 years) was 31 years. The estimated mean years of schooling received by
the women was 3.8. However, nearly 51 percent of the women'were illiterate, 25 percent had undergone
seven years of schooling and the remaining 24 percent had eight or more years of schooling.

All the study villages were multi-caste villages. In terms of the caste/religious composition, a little
over a quarter (27 percent) of the population belonged to upper Hindu castes (mostly Patels), 44 percent

2

belonged to the lower Hindu castes, such as Baria and Thakor, 17 percent belonged to scheduled castes (the

category’ includes Christians who were converted from the scheduled caste of Vankars) and 12 percent were
Muslims. This region of Gujarat state has virtually no tribal population in rural areas.
Almost 75 percent of the households had electricity. On the other hand, 31 percent had a separate
bathroom and 30 percent had toilet facilities within their premises. Majnrily of these households had both

the facilities^. ?) In this region of Gujarat piped water is nearly universally available and practically every

household had a water tap within the premise.
Prevalence of Reported Gynecological Morbidity
Overall. 75 percent of the respondent women reported some disorder or problem with their reproductive
functions, as shown in Table 2. The average number of episodes of illness reported was 3.6. A little over
half the w omen reported that they suffered from lower backache just before or during their menstrual

period. About 22 percent described their menstrual periods as very painful, such that they had to take rest or
lie down for a few hours during the day. Some women also indicated that they felt feverish during this
period.

Excessive \ aginal discharge was reported by 36 percent of the women. The discharge was
considered excessive by women when their clothes (petticoat and sari) were soiled, or when they had to use
pads to protect themselves. About 14 percent of the women complained about itching sensation in the
vaginal area and/or bad odour. Nearly the same proportion indicated painful sexual intercourse.
Incontinence or involuntary urination and genital prolapse were reported by 8 and 3 percent of the

respondents, respectively.
Treatment Seeking Behaviour
In order to understand the extent to which women are silent sufferers, we asked them a series of questions
about communication between them and other members of the household about each of their problems, and

about seeking treatment for their reported problems. As indicated in Table 2, only about a third of the
women ever discussed their problems related to discharge, itching and odour and pain experienced during
intercourse with the family members. The rest endured their pain in silence. Only 8 percent of the women
who had experienced involuntary urination, had ever discussed their problem with others in the family.

There is a sense of shame attached to such conditions.
Nearly 90 percent of the women who talked about their problems, did so only with their husbands
rather than with anyone else. It has been our observation that in rural areas the majority of women hardly
ex er confide in other women about their pains and sufferings. The practice of village exogamy leaves them
with virtually no female friends in the villages of their husbands. A certain distance is maintained with the
female relatives w ho belong to the kin group of the in-laws.
From amongst those who reported the ailments, 88 to 98 percent of the women had not sought any
treatment. The percentage varied between the types of illness. When asked why they did not seek treatment,
w omen reported that most of their problems were chronic and were a part of their being women and bearing

children. Endurance of pain is an integral part of their living. Women also felt that treatment would be
expensive if they visited the private practitioners. Many women indicated that the private doctors generally
ask women to bring their husbands also for consultation and treatment, but most men would never agree to
that. Women also thought there was no point in visiting the government operated primary health centres
because these have no doctors to attend to women's health problems or to infections related to the
reproductive tract or functions.
An important dimension of these findings is that 97 percent of the women w ho reported painful
sexual intercourse did not seek any treatment for it although a third had conveyed their problem to their
husbands. This partly reflects the attitude of men towards their partners. Women cannot go alone to the

3

health care providers for treatment and husbands do not see the need to accompany their wives. According
to the respondents, men see no reason to visit the health care provider when they have no problem. By
agreeing to pay for the treatment of their wives, many men absolve themselves of their responsibility.
Table 3 presents the results of our effort to understand whether vaginal discharge is accompanied
by any other symptoms and see whether there are any differences in incidence of reporting about the
problems and treatment seeking behaviour. Although this list was prepared in consultation with two
gynecologists, I have refrained from using medical terminology or labeling them as medical conditions but
have retained the symptomatic descriptions at this stage. The differences in incidence of reporting of the
problem and treatment seeking behaviour, however, suggest that the more severe the problem, the less are
the chances that women will seek treatment, until it reaches beyond their endurance limit and needs surgical
or other drastic interventions.

Correlates of Gynaecological Morbidity
Tables 4 and 5 show the percentage of women reporting a few illnesses classified by their caste and the
level of education, respectively. The middle panels of the Tables show the percent of women with a
problem, who had talked about their condition to some one within or outside the family. The lower panels
show the percent of women who sought treatment for their problems.
Except among the scheduled caste women, there was very little difference between women
belonging to the other groups in the incidence of the three conditions (vaginal discharge, itching and/or bad
odour and painful sexual intercourse) shown in Table 4. The incidence of all the three problems shown in
the Table is significantly higher among the women belonging to the scheduled castes compared to all the
other women. It is difficult at this stage to assess whether the differences are due to differentials in the
reporting of morbidity or in the actual load of morbidity among women of various communities. The results
of the gynecological examinations, when fully analyzed, may throw more light on these issues or clear
some of the queries.
However, women's willingness and the space available to them to discuss some of their problems
varied between women belonging to different caste/religious groups. Although not statistically significant,
scheduled caste women discussed problems related to vaginal discharge or sensation of itching or bad odour
with the other members of the household more than the women belonging to the other caste groups. The
caste differences with regard to informing someone about painful sexual intercourse were significant and
clearly reflect the differences in social and cultural mores of the various caste groups. It is often observed
that women belonging to scheduled castes (and tribes) are much more vocal, that they joke about and are
willing to discuss sex-related issues with others, whereas women belonging to the other communities are
much more subdued. Only 18 percent of the lower Hindu caste women had informed someone about the
pain experienced during sexual intercourse but more than half the scheduled caste women had done so.
As indicated in the lower panel of Table 4, a very small percent of women (ranging between 3 and
9) sought treatment among those who experienced the stated problems. One cannot, therefore, discern with
confidence any inter-caste differences in the treatment seeking behaviour of women.
In Table 5 the data on the incidence and the treatment seeking behaviour of women by the level of
education are presented. Education does not seem to be an important factor determining the incidence of
various problems. Yet, the incidence of the three problems was lower among women with no education than
among women with primary education as well as those who were educated up to and beyond high school.
Again, at this juncture, it is difficult to say whether this is a reporting problem. However, a higher
proportion of the better educated women tended to talk about their problems (mainly to their husbands)
compared to those with no education. The treatment seeking behaviour, on the other hand, indicated no
clear education related pattern.

4

Pending further analysis of the data and matching of the results of the medical examination with the
data, the results obtained thus far do unequivocally suggest that the majority of the rural Indian women
suffer pain, discomfort and absence of well being in silence. A strong programme which addresses their
needs and empathetic health care providers will go a long way to alleviate suffering among millions of
women.

5

Table 1: Study Area and Profile of Respondents

Household Characteristics







Total Population of five villages according
to the 1991 Census
Number of Households
Average Household size
% of Households having separate bathroom
% of Households having toilet facility
% of households having electricity for domestic use



Caste Composition



10,787
1,892
5.7
31
30
75
% of households

Upper caste Hindus
Lower Caste Hindus
Scheduled castes
Muslims

27.5
44.0
16.5
12.0

Characteristics of Respondent Women






Number of respondent women (ever-married women
aged 16-60)
Average age of Respondent women (years)
Mean Years of Schooling of Respondent women
% Distribution of women by level of Education
Illiterate
Up to 7th standard
Sth to 10 standard
Above 10 standard

803
31
3.8
51
25
17
7

*

6

Table 2: Percent of Women Reporting Gynecological Morbidity and
Their Treatment Seeking Behaviour, Rural Gujarat, 1995

Nature of
Problem

% of women
reporting

% talking about
it to others among
those who report
the problem

% seeking
treatment among
those who report
the problem

Backache before
or during menses

51.6

Painful menses

22.0

31.6

21.0

Excessive vaginal
discharge

35.7

35.5

9.4

Itching or bad odour

13.5

36.1

8.3

Painful sexual intercourse 11.7

35.1

3.2

Involuntary urination

7.6

8.3

3.3

Genital prolapse

3.1

29.2

9.5

4.9

Percent of women reporting
at least one problem

75.0

Average number of
problems reported

3.6

No. of respondent women

803

Note : The questions on whether and to whom women talked about the problem
related to backache before or during menstruation were not asked in the
siuwey. However, questions on the treatment seeking behaviour were asked.

7

Table 3: Types of Vaginal Infections and Treatment Seeking Behaviour
Reported by Women in Rural Gujarat, 1995
Type of infection

% distribution
of incidence
reported

% talking
about the
problem to
others

% seeking
treatment
among those who
report the problem

18.4

68.0

30.1

8.5

36.1

14.8

Vaginal discharge w ith burning
during urination and fever

11.3

7.8

4.7

Vaginal discharge w ith severe
backache

6.3

10.5

2.3

287

(36.3)

Vaginal discharge w ith fever
Vaginal discharge w ith itching,
bad odour

No. (and %) of w omen
reporting vaginal discharge
No. of problems reported
Ax erage no. of problems per
woman reporting a problem

1.28

No. of respondent women

803

368

8

Table 4: Percentage of Women Reporting Selected Gynaecological
Problems by Caste/Religion, Rural Gujarat, 1995

Caste/
Religion

Total Number
of respondents

Vaginal
discharge

Itching and/
or bad odour

Painful
sexual
intercourse

32
35
46
31
36

16
9

216
358

21

9
11
22

16

8

14

12

97
803

(a) Percent Reporting

Upper caste Hindus
Lower caste Hindus
Scheduled caste
Muslim

All

131

(b) Percent of Women Informing Someone About Their Problem

Upper caste Hindus
Lower caste Hindus
Scheduled caste
Muslim
All
(c)

37
18
55
38
35

38
34
41

27
36

Percent of Women Seeking Treatment Among
Those who Reported the Problem

Upper caste Hindus
Lower caste Hindus
Scheduled caste
Muslim
All

Note:

35
35
38
30
36

8
9
12
3
9

5
3
3
0
3

9
13
7
0
8

The percent of women seeking treatment among those who
experienced painful sexual intercourse was very small, hence the
figures may not be very stable.

9

Table 5: Percentage of Women Reporting Selected Gynecological
Problems by Level of Education, Rural Gujarat, 1995
Les el of
Education

Vaginal
discharge

Itching and/
or bad odour

Painful
sexual
intercourse

Total Number
of Respondents

11

(410)
(201)
(136)
(803)

(N)

(a) Percent Reporting

Illiterate
Primary up to Std. 7
Standard 8-10
Above Std. 10
All

32
38
43
37

11
15
17

1 1
15
12

36

14

12

17

(56)

(b) Percent of Women Informing Someone About Their Problem

Illiterate
Primary up to Std. 7
Standard 8-10
Above Std. 10
All

(c)

32
41

35

25
39
44

23
48
40

36
36

60
36

57
35

Percent of Women Seeking Treatment Among Those who
Reported the Problem

Illiterate
Primary up to Std.
Standard 8-10
Above Std. 10
All

10
8
7
18
9

9
13
0

0
9
5
0

10
8

3

Note: The per cent of women seeking treatment among those who experienced painful sexual intercourse
was very >mall, hence the figures may not be very stable.

10

REFERENCES

Bang. R. and A. Bang. 1996. A community study of gynecological disease in Indian villages: some
experiences and reflections, in S. Zeidenstein and K. Moore (eds.) Learning About Sexuality, New York,
The Population Council: 223-237.
Bhatia. J.C. and J. Cleland. 1995. Self-reported symptoms of gynecological morbidity and their treatment in
South India. Studies in Family Planning. 26(4): 203-216.

Bhattacharya, S. 1994. Gynecological morbidities among women in West Bengal, Child In Need Institute,
Unpublished report. Cited by Koenig, et al. 1996.
Kannani. S„ K. Latha and M. Shah. 1994. Application of qualitative methodologies to investigate
perceptions of women and health practitioners regarding women's health disorders in Baroda slums, in: J.
Gittlesohn, et al. (eds.). Listening to Women Talk about Their Health, New Delhi, Har-Anand Publications:
116-130.

Koenig. M. et al. 1996. Undertaking community-based research on the prevalence of gynecological
morbidity: lessons from India, paper presented at the IUSSP Seminar on Innovative Approaches to the
Assessment of Reproductive Health, held at Manila, the Philippines, September 1996.
Ministry of Health and Family Welfare (MOHFW). 1996. Action plan for Revamping the family welfare
programme in India, in: Anthony R. Measham and Richard A. Heaver, (eds.) Supplement to India's Family
Welfare Program: Moving to a Reproductive and Child Health Approach, Washington D.C., The World
Bank: 1-12.

Mulgaonkar, V. B. et al. 1994. Perceptions of Bombay slum women regarding refusal to participate in a
gynecological health programme, in: J. Gittlesohn, et al. (eds.) Listening to Women Talk about Their Health,
New Delhi. Har-Anand Publications: 145-167.

Parikh. I. et al. 1996. Gynecological morbidity among women in a Bombay slum, Streehitakarini, Bombay,
Unpublished report.

11

OOH

Undertaking Ccm muni tv-based Research on the Prevalence of Gynecological Morbidity: Lessons
from India

Michael Koenig (1)
Shireen Jejeebhoy (2)
Sagri Singh (I)
S. Sridhar(3)

Paper presented at the IUSSP Seminar oil Innovative Approaches to the Assessment of Reproductive

Health. Manila, the Philippines
September 24-27, 1996

(1) Ford Foundation, New Delhi

(2) Independent Consultant, Bombay

(3) SEW A-Rural, Gujarat

Undertaking Ccmniunitv-based Research on the Prevalence of Gynecological Morbidity: Lessons
from India

Michael Koenig (1)

Shireen Jejeebhoy (2)
Sagri Singh (I)

S. Sridhar (3)

Paper presented at the IUSSP Seminar oil Innovative Approaches to the Assessment of Reproductive
Health. Manila, the Philippines
September 24-27, 1996

(1) Ford Foundation, New Delhi

(2) Independent Consultant, Bombay

(3) SEWA-Rural, Gujarat

4

I!

Introduction

The last decade has witnessed growing international recognition of the scope and significance of
gynecological moibidity among poor women in developing countries (Pixon-Muellcr and Wasserheit,

1991: Germain, et al. 1992). In documenting that gynecological problems are not confined to special

clinic-based populations but are instead widespread within the community at large, scientifically rigorous
and carefully implemented research has played a key role in increasing the visibility of this issue. India

has been a particularly fertile site for such research, with a total of seven community-abased prevalence

studies of gynecological morbidity completed to date.1 These data collectively constitute some of the
most important available empirical evidence internationally on the magnitude of this public health

problem." and ha\ e also played an important catalytic role in the recent policy and programmatic shift
w ithin India toward a reproductive health approach.3

The studies also offer many important insights into the methodological challenges associated
with undertaking community-based research on this critical component of reproductive health. The Ford

Foundation funded or co-funded all seven of these community-based studies. In our respective roles as
donors, consultant, and grantee, the authors have had the opportunity to work closely with many of the

researchers involved with the implementation of these projects. The study investigators have generously
shared their experiences regarding the numerous challenges and complexities associated with
undertaking research in this sphere. Our objective in this paper is to distill and summarize the collective

lessons from these projects in carrying out community-based research on the prevalence of gynecological

morbidity, and the implications for future research efforts both in India and other developing countries.

In the following sections, the study designs and methodologies employed in the seven
community-based studies are initially reviewed and compared. The findings of the studies with respect to
prevalence and patterns of gynecological morbidity are then briefly reviewed. We subsequently

summarize some of the important methodological issues and lessons for research which have been
learned o\ er the course of this effort. The paper concludes with a discussion of the implications of these

lessons for future community-based research on gynecological morbidity in developing settings.

Community-Based Surveys Of Gynecological Morbidity In India
The seven studies represent a diverse range of geographical and cultural settings within India (Figure 1).
All of the studies undertook community-based surveys of gynecological morbidity among women,

consisting of both self-reported morbidity as well as clinical examinations. In five of the studies.

1

Figure 1

Location of Seven Community-based Studies of Gynecological
Morbidity in India

/

$

0 Rajasthan
Studv

y.est Bengal Study

^qroda 5

mraL' Mahar
• ?°Bombay St

°Kar

'uudy

a Study

ntra

laboratory investigations were also carried out.4 However, as laboratory tests and procedures varied
widely across sites and were the least comparable data collected by projects, we have not considered this
component in the present paper.

Organizational Settings:

Six of the studies were carried out under the auspices of highly respected voluntary organizations, with

long-standing records of addressing the health needs of the communities which they studied (Table 1).

The rural Maharashtra study was conducted by SEARCH and the rur^l Gujarat study by SEWA-Rural,

both highly respected voluntary organizations which provided comprehensive health services to
communities from which the study populations were drawn. The Rajasthan study was conducted by an
independent researcher, under the auspices of a well-established non-governmental organization, the

URMUL Trust, which had provided health care in a remote district of rural western Rajasthan since
1984. The West Bengal study was conducted by the Child in Need Institute, a voluntary organization on

the periphery of Calcutta, which had provided family planning and health services to women and
children in 40 villages for more than a decade prior to the study. The two urban slum studies were also

conducted by NGOs which had established themselves as major providers of both health information and
services in the project areas; the Baroda Citizen’s Council in Baroda and Streehitakarini in Bombay, with

a record of three decades of addressing the health needs of poor women and children. The only exception

to this rule was the study in Karnataka, which was undertaken by a team of researchers from the Indian
Institute of Management-Bangalore, unaffiliated with an NGO and with no prior record of service
provision or engagement with the population studied.

Sampling Procedures:
Sampling designs and procedures varied considerably across the seven studies (Table 1). Several rural
studies enumerated and sought to obtain information upon all eligible women in.a limited number of

field sites; these included rural Maharashtra (2 villages), Rajasthan (2 villages) and rural Gujarat (10
villages). In contrast, the West Bengal study adopted a quota sampling approach, whereby it continued

data collection until it obtained the requisite number of respondents from selected project areas. Only
three studies adopted study designs based upon representative sampling procedures. The two urban slum

studies, Bombay and Baroda, selected random samples of populations residing in their project areas; a 50
percent random sample of married women from two slum areas in the case of the Baroda study, and a 10

percent random sample of women in the Bombay slum study.5 The Karnataka study was a prospective

3

Tabic 1. Description of Community-based Gynecological Morbidity Studies in India
Location
and Year of
Study

NGO
Affiliation

Sample

Data Collection
Procedure

Study
Enrollment
Period

Study Population

Ever married and
unmarried women
aged 13 years and
older
Ever married and
unmarried women
aged 13-45 years

100% sample
in 2 rural
villages

Concurrent interview
and examination at
village level clinics

5 month period, 3
days per week

Quota sample
of 500 women
in 8 rural
villages
10% random
sample of
project
households in
one urban
slum area

Concurrent interview
and examination at
central medical facility

18 month period,
two clinics per
week
2-3 repeat visits
12 month
duration; open
enrollment for
clinical
examination over
entire period

Rural
Maharashtra
(1989)

SEARCH

Rural West
Bengal
(1990-91)

Child in Need
Institute

Bombay
(1988-89)

Streehitikarini

Baroda
(1990-91)'

Baroda
Citizens
Council

50% random
sample from
two urban
slum areas

Karnataka
(1992)

No affiliation

Rural Gujarat
(1988-89)

SEWA Kura!

Rural Rajasthan
(1994)

URMUL Trust

Subsample of
a random
sample of 48
villages and
one town
100% sample
in 10 rural
villages
100% sample
of two rural
villages

Interviews conducted at
home and concurrently
with clinical
examination;
examination conducted
at NGO clinic or near
home
Interviews conducted at
home and concurrently
with clinical
examination;
examination conducted
at NGO clinic
Interview conducted
monthly at home;
examination conducted
at project office clinic

Concurrent interview
and examination at
village health fairs
Interview conducted at
home; examination
conducted at village
health clinics

Ever married
women aged 15
years and older

Sample for
Clinical
Morbidity
Survey
Examination
(% participating)
654 women
650 women and
(includes 92
unmarried girls
unmarried girls)
(59%)
(59%)________
500 women
500 women and
unmarried girls
(includes 237
(44%)
unmarried girls)
(44%)________
756 women
756 women
(72%)
(72%)

12 month
duration; within
2-3 days of
interview

Ever married
women aged 18-45
years

840 women
(100%)

548 women
(65%)

12 month
duration; day
following last
interview

Ever married
women less than 35
years of age with a
child < 6 mos.

440 women
(100%)

385 women
(87%)

2-4 days during
village health fair

Ever married
women aged 15
years and older
Ever married
women aged 15
years and older

1103 women
(49%)

324 women
(29%)

274 women
(100%)

51 women
(19%)

2 month
duration; 1-2
days in third
month

I

i

r

follow-up of all women with a child under 6 months of age (a total of 440 in number), drawn from a

larger representative cross-sectional survey of 3595 women wifh a child under five years of age residing


I

»

in 48 villages and one town.
I

Two studies found it necessary to expand their original sampling frames in order to obtain the

sample size desired, given significant levels of non-participation by women. The West Bengal study, for
example, expanded the number of study villages from two to eight, with the sample comprised of all
women who subsequently attended health clinics. In the Rajasthan study, low participation among

women from the two originally selected study villages prompted the investigators to subsequently add
data for women from 38 adjacent villages who attended project health clinics.

Study Populations:

In all of the studies, the primary study population of interest was ever married women of reproductive

age. However, at least two of the studies also collected data from nonrinarried women. While the
majority of the 650 female women respondents in the rural Maharashtra study were either currently or

previously married,6 the sample also included data from an additional 92 unmarried girls. Similarly, in
the West Bengal study, clinical and survey data from 105unmarried adolescent girls were collected in

addition to data for 395 married women.7 While some studies limited investigation to women within the
reproductive ages (West Bengal, Baroda), other studies extended their samples to include women beyond

age 45 as well (the rural Maharashtra, Gujarat, Rajasthan, and Bombay studies).

Data Collection Approaches:

Given the reluctance of many women to undergo gynecological examinations, data collection strategies
in each site were tailored to local conditions in order to enhance participation (Table 1). In several
studies, the structured interview and clinical examination were conducted simultaneously— that is,

before women were clinically examined, they were administered a survey questionnaire. This sequence

occurred in the rural Maharashtra study, where a field camp was set up in each village, with facilities for
both the interview and the internal examination, assuring privacy. It also occurred in the West Bengal
study, where the clinic was based in a central place, and the clinical examination was preceded by a

structured interview. The rural Gujarat study, although conducting the structured interview and the

clinical examination simultaneously, adopted a somewhat different approach. Fearing that cultural
proscriptions would deter women from attending a women's health clinic, general village health fairs

5

open to all men and women were organized in each village. Each fair lasted between two and four days

and included general check-ups as well as health exhibitions. Extensive publicity preceded the fairs and
the entire population was invited to attend; it was repeatedly indicated that the camp was intended for

-

healthy as well as those sick persons. In the Rajasthan study, while women from the two main study

villages were interviewed at their home using a socioeconomic and demographic survey, only women
who appeared for the clinical examination were interviewed about morbidity by the physician

immediately prior to the clinical examination.
In the other three studies, there was a clear demarcation between the survey and the clinical
examination, with the survey preceding the clinical examination by one or more days. For example, in
the Karnataka study, a day before the medical examination was scheduled, women were interviewed by
female interviewers and information was elicited on symptoms of gynecological problems. Physical

examinations were held on the following day in clinics set up at the project office in town or at primary
schools. Similarly, in the two urban slum studies in Bombay and Baroda, interviews were conducted in

the home setting by social workers and respondents were invited to attend established clinics for

examination at a subsequent date.
The studies also differed considerably in terms of overall duration, periods for enrollment, and
examination sites for the clinical component. The rural .vlaharashtra study extended over five months,

with three clinic days scheduled per week in the two study villages. In the Rajasthan study, the survey
phase was completed in two months, followed thereafter by a one month clinical phase in nine different
village sites, with each clinic lasting for 1-2 days. The clinical phase of the Karnataka study lasted three

months, with separate examination sites set up in each sampled village, usually m a primary school
building. Significant effort was made to enroll women who missed the clinic scheduled in their own

village, by transporting them to clinics organized in other villages and through arranging a central clinic
near the end of the study. In the rural Gujarat study, enrollment in the clinical component was tied to the

organization of the multi-day health fairs held in each village, and the availability of the examining
physician on 1-3 of these days, with only limited efforts at subsequent enlistment. The enrollment period

for the clinical component of the Bombay study extended over an entire 12 months, with sampled
respondents able to enroll in the study at any time during this period at a NGO clinic adjacent to the
slum. The West Bengal study also followed an extended study period of 18 months, with clinic sessions
scheduled for two afternoons each w eek. Generally 2-3 repeat visits were made to encourage respondents

to attend a clinic to undergo the physical examination, which was situated at a medical facility located 35 kilometers away. In the Baroda study, the enrollment for sampled women was within the week
J

6

following the home interview, w ith only limited attempts at subsequent recruitment for the study among
women who initially refused or were unavailable.

Main Findings
As shown in Table 2. marked variation is evident across the seven studies in terms of patterns and levels

of both self-reported and clinically diagnosed gynecological morbidity. Menstrual disorders constituted
an important gynecological problem in all studies, with the percentage of women reporting this problem

ranging from a low of 33 percent in West Bengal to 60 percent or more in rural Maharashtra and

Karnataka. Excessive discharge was also a frequently reported problem, although wide variation between
studies was once again evident—ranging from 13 percent in the rural Maharashtra study to as high as 78
percent among highly self-selected women in the Rajasthan study. Less common but still important

problems were lower abdominal pain (ranging from 9 to 21 percent) and lower backache (from 5 to 39
percent) in studies which collected and reported these d$ta separately. Lastly, while dyspareunia (pain
during intercourse) was an infrequently cited problem in three of the studies which included this

condition, in the Rajasthan study which probed extensively on this issue, fully 48 percent of women
reported this problem. Overall, the percentage of women reporting one or more gynecological problems

ranged between 55 to 74 percent of the four studies with high participation rates, and 84 and 100 percent

for two rural studies in Gujarat and Rajasthan, where self-selection is likely to have been a particularly

serious concern.
Widespread variation was also evident across studies with respect to the principal gynecological

morbidities diagnosed through clinical examination. The prevalence of vaginitis ranged from a low ol 4
percent in the West Bengal study to a high of 62 percent in the rural Maharashtra study, with rates

ranging between 10 to 15 percent in four out of the other five studies. The percentage of women with
cerx icitis ranged from as low as 8 percent (rural Gujarat study) to 40 percent or higher (Bombay and

rural Maharashtra studies), and cervical erosion from as low as 2 percent in the West Bengal study to 46
percent in the rural Maharashtra study. Levels of diagnosed pelvic inflammatory disease also varied

widely, ranging from 1 1 percent or under in four of the studies, but as high as 24 and 36 percent in the
rural Maharashtra and Rajasthan studies, respectively. The percentage of women with diagnosed

prolapse was also highly variable, ranging from under 1 percent in the rural Maharashtra study, to 3-5
percent in the Karnataka and Baroda studies, to just under 20 percent in the West Bengal and Bombay

studies, and as high as 27 percent in the Rajasthan study. Overall, the percentage of women with one or
more clinically diagnosed conditions ranged from a low of 26 percent in the Baroda study, to 43 percent

7

*

i

L
Table 2. .Prevalence of Self-Reported and-Clinical Gynecological Morbidity Among Community-based Studies in India

Morbidity

Rural
Maharashtra

Rural West
Bengal

Bombay

Baroda

Karnataka

Rural
Gujarat

Rural
Rajasthan*

% of Women with
SELF-REPORTED:

Menstrual problems

60

33

41

58

65

59

48

Excessive discharge

13

50

31

22

22

57

78

Lower abdominal pain

13

17

21

9

16

NA

10

Lower backache

30

5

39

24

NA

30

16

Dyspareunia

7

2

NR

NR

1

NR

48

55**

65

74

65

NA

84

100

Vaginitis

62

4

15

11

13

10

30

Cervicitis

48

14

40

13

24

8

19

Cervical erosion

46

2

21

5

10

20

30

Pelvic inflammatory disease

24

1

16

8

11

8

36

Prolapse

0.5

17

18

5

NA

27

One or more conditions

NA

43

74

26

43

77

One or more conditions

CLINICALLY DIAGNOSED:

NR = Not recorded
NA = Could not be ascertained from available data
• * Data from community and clinic-based samples combined
** Does not include lower backache and I
abdominal pain

70

in the rural Gujarat and West Bengal studies, to 70 percent or more in the three other studies where data
8

was available.. >

h •
Lessons Learned

1.

The studies demonstrate the central importance of close engagement with the community
for the successful completion of such studies.

A common element of almost all of the studies was the high level of engagement and rapport between
the investigators with the communities studied. Almost all studies were carried out through affiliations

with highly respected voluntary organizations with long-standing records of service to the communities,

either directly by the organizations themselves or by researchers working under their auspices. The one
important exception was the Karnataka study, which achieved very high study participation rates despite

being unaffiliated with any particular voluntary group.
It is also e\ idem from a review of these studies that despite having an entree into communities
through voluntary groups, exhaustive preparatory efforts and engagement with communities was

nonetheless required in order for researchers to gain their cooperation and active support. Given the
subordinate position of most Indian women within the family, convincing male community and family

members of the rationale and need for the study assumes central importance in successfully enlisting the

participation of women in the study. The rural Maharashtra study, for example, held an extensive series

of meetings with \ullage leaders as well as men and women villagers separately to inform them of the

rationale for the study: male village volunteers also played a critical role in organizing the survey and
clinical field components. Similarly, both the rural Gujarat and Rajasthan studies undertook extensive
advance publicity efforts in attempting to inform and persuade all members of the community— both

male and female— of the merits of their study. Although not well documented, it is clear that the other
three projects— West Bengal, Baroda, and Bombay—also engaged in intensive mobilization efforts to
promote women's participation in their studies, using the infrastructure from their existing community

health programs.

2.

The studies illustrate the importance of providing appropriate medical care to respondents,
since for many women, this may represent the only opportunity they will have to address
their reproductive health problems.

The provision of medical care to address respondent’s reproductive health problems may be considered
almost mandatory in settings such as India, where few poor women have access to high quality

gynecological care.9 It is evident from the studies considered that the arrangement of required care for

9

study participants is important not only from an ethical perspective, but may also be a pre-requisite for

enlisting women’s cooperation and participation in studies of this nature. From a pragmatic standpoint, it
seems doubtful that many women would willingly acknowledge sensitive problems or agree to submit to

*what for many constitutes an embarrassing medical procedure, if Effective recourse or trCatme’nt was not
r

also provided.



Studies varied considerably in terms of the extent to which they included medical intervention

directly within their study designs. The rural Gujarat study, through health fairs in each village, offered

detailed health check-ups to all married women in the villages who attended, and less comprehensive

check-ups tor other community members. The Bombay, Rajasthan, rural Gujarat and West Bengal
studies provided medicine to women with readily treatable conditions, and in the case of more

complicated cases, referred women to known hospitals for further investigation and treatment. In the
Rajasthan study, the URMUL Trust under whose auspices the study was conducted, has continued to
follow up women who participated in the study, and went so far as to establish a fund to defray the costs

of needed surgical procedures for women for conditions such as prolapse. The rural Maharashtra study

provided treatment free of charge for all health problems to all participating women, and extended this

service to other members of the women's family as well. The Karnataka study also provided medicines

for simpler problems, and referred more complicated cases either to local hospitals or to the examining
physician in Bangalore, 70 kilometers away. The Baroda study treated manageable cases through their

clinic located in the slum under study, and referred more complicated cases to the municipal government
medical college.

Although not well-documented, there would appear to be considerable variation in the extent to
which projects assumed continued responsibility for resolving women’s reproductive health problems,
and with more complicated cases, in guiding them through what to many poor women constitutes an

intimidating system of higher order medical care.10 What is clear from this experience, however, is that

without a strong commitment to appropriate and effective medical treatment for identified health

problems, it is unlikely that projects would have been able to enlist the degree of cooperation and
participation by women in their studies which they did, especially for the clinical component.

10

V

i

3.

The studies illustrate the importance of ascertaining and using local terminology and
language when querying about gynecological conditions.

In almost all of the studies, considerable emphasis was placed upon understanding women’s

perceptions of morbidity and in establishing local terms and expressions fordescribing gynecological

complaints which were understandable and salient by the women themselves.11 As an example,
investigation by the rural Maharashtra research team revealed that among rural women in their project
area, there existed as many as twelve different expressions for, and five major categories of, white

discharge (Bang and Bang, 1994). Similarly, exploratory in-depth research by the West Bengal team
underscored both the diversity of terms women used to describe conditions such as discharge, as well as
the reluctance of women to classify certain conditions as problems (e.g., dysmenorrhea, pain during

intercourse) or e\ en to view other conditions as indicative of good health (excessive bleeding). Detailed

ethnographic research in the Bombay study provided insight into the range of terms used by slum women
for conditions such as vaginal discharge and prolapse (Streehitikarini, 1995). A detailed understanding of

local language and terminology also assisted investigators in understanding the subtle and more indirect
ways in which women describe some gynecological problems. As the investigators in the rural

Maharashtra study describe, a woman experiencing white discharge will often describe this as


‘weakness’. Without this prior knowledge, such important conditions are likely to go undetected by both

researchers and providers (Bang and Bang, 1996).

4.

Underreporting of gynecological problems by women remains a significant methodological
problem in self-reported responses on gynecological morbidity.
Two sources of data provide strong evidence of significant levels of underreporting of

gynecological problems in response to survey questions. The first set of data come from the previously

described study from Karnataka by Bhatia and colleagues. As one component of the larger study, a
subsample of 440 mothers were interviewed monthly over the course of one year, and were asked about
their children's and their own health problems, including gynecological problems, during the preceding

month. As shown in Figure 2, the percentage of women reporting gynecological problems rose steadily
over the course of the study, increasing from 8 percent in Round 1 to 33 percent in Round 12. The
increase in the number of women reporting the problem of excessive discharge over the course of the
study was particularly notable, increasing from 5 percent in Round I to 19 percent in Round 12. Similar

results were evident from a study in Haryana in North India, which as part of a larger prospective study
on the impact of women’s employment on health,12 conducted six bi-monthly visits to a sample of 276

11

Figure 2: Women's Self-Reported Gynaecological Problems According to Round of Data Collection:
and Haryana
40
37.9

35
332

30

28.9

25
226

20
17.3

14.5

15

13 1 --------

17.3 17.3
__________ 16.5

Ii s

14.2

13

‘■rf-

¥

10.7

10

8.9
7.8 ____
6.5

6.2

f
4.9 r" ■

5

0
1

2

3

4

5

6

7

Round

Karnataka

Sources: Bhatia, et al., unpublished data
Kumtakar and Roy, unpublished data

8

9

10

TT

12

!'

H

I

T

Round
Haryana

•.:r
5

6

Karnataka

women to inquire about health problems, including reproductive morbidity, during the preceding 15

da> s. A sharp increase in the percentage of women acknowledging gynecological problems was also

evident in this study, with the percentage increasing from 5 percent in Round 1 to 38 percent in Round

6.13
While the possibility of some overreporting of gynecological problems cannot be entirely ruled
out. given the highly sensitive nature of these issues, a more likely explanation is significant

underreporting of gynecological problems, especially in the early rounds of both studies. Several factors

may contribute to women’s increased willingness to acknowledge gynecological problems, including
women becoming increasingly comfortable with the interviewers and more confident that the
confidentiality of such information is safeguarded; interviewers becoming more proficient in eliciting
such information: and women increasingly coming to view such conditions as legitimate problems, rather

than ’a woman's lot in life' which they have been conditioned to tolerate.14

The experience of several studies also suggests the level of reported gynecological morbidity

may be strongly influenced by the background of the interviewer. Morbidity levels reported to medical
practitioners or to interviewers shortly preceding the clinical examination tend to be significantly higher
than those reported to interviewers in home visits, a finding also reported elsewhere.15 For example, in
the Karnataka study, while only 15 percent of all women reported menstrual problems in interviews in

their homes one day preceding the clinical examination, as many as 65 percent reported a problem the
next day to the examining clinician (Bhatia, et al. 1995). In the rural Maharashtra study, while 21 percent
of w omen reported menstrual disturbances to a non-medical interviewer, this percentage rose to 58

percent when interviewed by the examining physician (Bang, et al. 1989). Similar differences were

obsen ed in the Baroda and Bombay studies; morbidity rates reported to investigators (usually with
social work backgrounds) were so markedly lower than those reported to the clinician immediately prior

to the examination that both studies chose to rely only upon responses made to the clinician. One
contributing factor may be the greater propensity of physicians to probe thoroughly for possible
morbidity, an issue discussed in the following section. The Bombay study also observed that although

investigators had already developed rapport with the community, and although they explained at great

length the need for this study, they faced some reluctance from respondents to respond to their questions,
given their inability to resolve medical problems.

13

5.

Specificity in question wording dnd detailedi'prs'liing ain significantly increase the
likelihood of women reporting actual gynecological morbidity.

A major reason why gynecological morbidity may be underreported could be due to a widespread
perception among women that such conditions are normal, and therefore do not merit acknowledgment
or complaint. In light of this, special efforts are required in order to elicit such information from women.

1 he experience of these studies illustrates the importance of both question specificity and probing

beyond respondents’ initial responses in ascertaining the true extent of gynecological morbidity among
w omen. Considerable variation w as evident in terms of the specificity with which interviewers presented
questions to women regarding gynecological complaints and problems. In the Baroda and West Bengal

studies, tor example, women were asked open-ended questions concerning problems they were
experiencing, and their answers were checked against a list of symptoms pre-recorded on the

questionnaire. The rural Gujarat, rural Maharashtra, and Bombay studies, in contrast, queried
respondents using a checklist of individual gynecological symptoms, and asked women about each one.

I lie Rajasthan study adopted a two-stage approach, whereby women were initially queried about any

problems which they had at the time, and subsequently about a series, of other specific conditions and
symptoms which had not been mentioned earlier.
1 he experience of several studies also strongly illustrates the importance of detailed probing. In
the Rajasthan study, after women had listed their gynecological problems, they were prompted for

symptoms which they had not reported earlier using a checklist type question: "You have mentioned
symptoms. Do you have any of the following problems?" If women reported additional symptoms
on probing, they were asked why they had omitted this problem: a common reason was embarrassment.

As shown in Table 3. prompting significantly raised the percentages of women reporting specific

conditions: from 28 percent to 48 percent for menstrual problems, and from 10 percent to 20 percent for
prolapse. I he most dramatic increase w as in the percentage of women reporting problems during sexual

intercourse, from 1 percent to 48 percent. Significant increases were also evident in the percentages of
women reporting vaginal discharge and infertility.16

In other studies, probing was less systematic, yet there is evidence that where it was built in,
reported rales of morbidity are correspondingly higher. For example, the rural Gujarat study probed

specifically tor dysuria (burning upon urination) while others did not. likely accounting for the relatively
high reported rates of this condition in this population (26 percent). Similar findings emerged from the

\\ est Bengal study, where while only 2 percent of respondents from the larger survey spontaneously

14

Table 3. Women’s Unprompted and Prompted Self-Reported Gynaecological Problems:
Rural Rajasthan, 1994

Unprompted

Both Unprompted and
Prompted

%

%

Menstrual Problems

27.9

47.9

Prolapse

10.0

19.6

Problems during Sexual
Intercourse

1.3

47.7

Vaginal Discharge

68.0

78.8

Infertility

9.8

14.9

I

Source: Oomman, 1996, p. 198



V

mentioned pain during intercourse as a problem, 15 percent of women who participated in the separate
i

focus group sessions acknowledged this problem upon probing. As one woman in the stilly responded:
"Does anyone discuss this even if she suffers?’'17 These results suggest that while all gynecological
problems should be probed, those that are equated most directly with.sexual behavior are likely to be
most sensitive in terms of reporting.

A more generic problem facing all such studies is the absence of agreed-upon definitions of
morbidity, and resulting variation in specificity with which individual conditions are described. While

several studies (rural Maharashtra, Rajasthan, and Bombay) inquired at length about vaginal discharge

(e.g., color, odor, consistency, quantity), others (rural Gujarat, Baroda, and Karnataka) did not, and this

may have contributed to the relatively low reported fates in at least some of the studies. In at least two of

the studies (rural Gujarat and Baroda), women were intentionally not asked about dyspareunia. A more
general and problematic issue is that questions tended to be vague and open to subjective
interpretation— for example, ‘scanty' periods or ‘excessive/abnonnal’ vaginal discharge. Although some

studies attempted to qualify morbidity questions— for example, "Are your periods scanty? For how
many days do you bleed?"— others did not. making comparisons problematic.

6.

Significant sample loss due to women’s refusal or reluctance to undergo clinical
examinations represents a serious problem in almost all of the community-based studies of
gynecological morbidity.

Significant sample loss for the clinical component of the study was a common feature in almost all of the
studies considered (Table 1). At one extreme were the two rural studies in Rajasthan and rural Gujarat,

where despite intensive efforts at recruitment and a long-standing presence in the communities studied,

only 19 percent and 29 percent, respectively, of sampled women were successfully examined. In the
other two rural studies, while the West Bengal study was successful in enlisting less than one-half of
selected respondents (44 percent) for the clinical phase, the rural Maharashtra study successfully enlisted

59 percent of all reproductive-aged women from the two sample villages. The two urban slum studies
achieved higher participation rates for the clinical component; 65 percent for the Baroda study and 72

percent for the Bombay study, with the latter high participation rate significantly assisted by the 12
month open enrollment period for the clinical examination. In terms of participation rates, the Karnataka

study stands out as exceptional, with the investigators successfully enrolling 88 percent of sampled
women in the clinical phase. It should be noted that the clinical component of this study tcok place at the

16

. end efan in.ensiye o„e year period ofeen.ae. -i.b sampiod —. listing of moadriy interview,

regarding their health and the health of their children.
Non-compiitn.ee was prim.rii, due re women’, mfusai or inabiiity to undergo a eim.c.
examina>ion. Several studies highlight the importance of male opinion in th.s area (West Beng. .
Bombay, and Rajasthan) and the inability to blunt opposition from male comnmn.ty members as a

primary facto, behind the lower-than-hoped-for study enrollment rates. A fluently mem.oned reason

for such opposition was the lingering distrust of the govemmem family planning program, and the fear

mat the gynecologic! exam,...ions wo.M provide a guise for conducting sterilization procedures. In
some oases, restrictions on the availability of trained medic! pemonnel a!so contributed to lower

participation rates. In the rural Gujarat study, for example, while almost half of the women who attended
the health fairs expressed a willingness to undergo the clinical exam, a much smaller percentage of

women did so. largely due to the availability of a gynecologist on only certain day, of the fair and the.,
reluctance or inability to make a return visit for the exam.

7.

Selection bias for the clinical examination by women with pre-existing gynecological
p Xs "presents a major methodo.ogical problem in commun.ty-based s ud.es of
gynecological morbidity. Self-selection would lead to an overestimation of the true
prevalence of gynecological morbidity in the community, particularly when overall
enrollment rates are low.

There is evidence indicating that women with reproductive health problems may be significantly more

likely to consent to clinical examination compared to women without apparent symptoms. A number of
the studies reported that a primary explanation given by sampled women for refusing the gynecolog.cai
examination was that they had no apparent or significant health problems (rural Gujarat, Bombay,

Baroda, and Rajasthan).
!„ several studies, it was possible to assess the extent of selection bias by comparing reported
morbidity of women who underwent clinical examination with those who refused (Table 4).18 In the rural

Gujarat study, where clinical sample loss was pronounced, the percentages of women reporting
menstrual ptoblems was somewhat higher (85 versus 70 percent), and reported vaginal discharge

significantly higher (56 versus 22 percent) among examined as opposed to non-examined women. In the

Baroda study, although no differences were evident between these two groups in reported levels of
menstrual problems (58 and 56 percent), 22 percent of examined women but only 15 percent of non­
examined women reported the presence of white discharge. Similar findings were reported in the rural

Maharashtra study, where while reported overall menstrual problems were similar or even lower in the

17

9

*

Table 4. Percentage of Women Reporting Gynecological Problems Among Women who
Undenvent and did not Undergo Clinical Examinations

Study

Women who underwent
a gynecological
examination

Women who did not
undergo a gynecological
examination

% Reporting

Rural Gujarat
Vaginal Discharge

55.9

22.5

Menstrual Problems

84.9

70.4

(324)

(779)

Vaginal Discharge

22.3

15.1

Menstrual Problems

58.0

56.2

(548)

(292)

Vaginal Discharge

13.5

8.2

Menstrual Problems
- scanty/irregular
- profuse/dysmenorrhea

19.5
20.0

31.3
17.9

(N)

(650)

(105)

(N)

Baroda

(N)

Rural Maharashtra

Source: Baroda Citizens Council, et al. (1995) and R.Bang et al. (1989)

examined group reported vaginal discharge was higher among women-who participated in the study
compared to a random sample of women who refused the clinical examination (14 versus 8 percent).

Selectivity was most apparent in the Rajasthan study, in which among the only 51 women who agreed to
undergo a physical examination, qu.t of the 27.4 wonipn in the two villages originally selected for

interview, all were symptomatic. These results suggest that the overall prevalence of gynecological

morbidity may be biased upward, as a result of sample selectivity, with the effect most marked in studies

with higher rates of sample loss.

8.

Significant variation was evident across studies in terms of clinical definitions and criteria
employed for gynecological morbidity.

A less widely recognized but important additional source of variation between studies relates to
variability in clinicians’ diagnoses of gynecological morbidity. Clear differences were evident among
studies with respect to clinical definitions which were employed (Table 5). For example, for diagnoses of

vaginitis, while most studies used inflammation of the vaginal wall plus discharge as their basic

definition, the rural Maharashtra and West Bengal studies were more specific in using laboratory
evidence of infection. A further possible point of variation could have been in clinicians’ assessments of
what constitutes ‘abnormal' vaginal discharge. For cervical erosion, while five studies included all

erosions, two studies excluded small, asymptomatic erosions (Baroda and Karnataka). Clinical
definitions also differed significantly with respect to cervicitis. While some studies included any cases of

cervical inflammation with pus discharge and/or evidence of prior or present scarred cervix (rural
Gujarat. Baroda, and Bombay), others considered only current inflammation with pus (Rajasthan and

Karnataka), still other studies also required lab evidence of infection to confirm the presence of cervicitis
(West Bengal and rural Maharashtra). In diagnosing pelvic inflammatory disease, different combinations
of a wide range of criteria were employed, including abdominal pain or tenderness, cervical motion

tenderness, uterine tenderness, adnexal tenderness or fixity, palpable adnexae, thickened vaginal
fornices, clinical cervicitis, and/or laboratory evidence of systemic infection. Lastly, among the six
studies for which data on prolapse are available, descent below the ischial spine served as the basic

criteria, with four studies utilizing both straining and cervical traction in diagnosing this condition, and
two studies using straining only. In the Baroda study, mild cases of cystocoele and uterine prolapse have

not been recorded.

19

Tabic 5. Clinical Definitions of Morbidity Employed in Community-based Studies of Gynecological Morbidity

Study

Composition of
Clinician Team

Vaginitis

Cervical Erosion

Cervicitis

Rural
Maharashtra

I female
gynecologist

Inflammation of vaginal
wall and lab evidence

All erosions

Cervical
inflammation with
pus discharge and lab
evidence of infection

Rural West
Bengal*

1 female
gynecologist

I n fl a m m at io n o f va g i n a I
wall with or without
discharge and lab
evidence

All erosions

Cervical
inflammation with
pus discharge and lab
evidence of infection

Bombay

2 female
gynecologists

Inflammation of vaginal
wall with discharge and
itching/pain

All erosions

Baroda

3 female
gynecologists

Inflammation of vaginal
wall

All erosions except
small, asymptomatic
erosions

Karnataka

1 female
gynecologist

Inflammation of vaginal
wall with discharge

All erosions except
small, asymptomatic
erosions

Cervical
inflammation with
pus discharge or
cases with inflammed
scarred cervix
Cervical
inflammation with
pus discharge or
cases with inflammed
scarred cervix
Cervical
inflammation with
pus discharge

Rural Gujarat

2 female
gynecologists

Inflammation of vaginal
wall with abnormal
discharge

All erosions

Inflammation of vaginal
wall with abnormal
discharge

All erosions

Rural Rajasthan

1 female
gynecologist

NA = Could not be ascertained from avaibu,e data

Cervical
inflammation with
pus discharge or all
cases with scarred
cervix
Cervical
inflammation with
pus discharge

Pelvic
Inflammatory
Disease
Palpable and tender
tissues around uterus
and cervical motion
tenderness and lab
evidence of infection
Tenderness or fixity
of tissues around
uterus and lab
evidence of infection

Prolapse

Based on traction and
straining

Based on straining

Lower abdominal
pain and tenderness
or thickening of
tissues around uterus

Based on traction and
straining

Tenderness or fixity
of tissues around the
uterus

Based on traction and
straining; excludes
mild cystocele and
Grade I descent

Uterine tenderness
and tenderness of
tissues around uterus
and clinical cervicitis
Tenderness or fixity
of tissues around
uterus

Based on straining

Abdominal pain and
uterine tenderness

Based on traction and
straining

NA

The fact that the studies Were independently designed and conducted only served to increase the

likelihood of variation in clinical definitions of gynecological morbidity employed. While we have
focused upon inter-study variability, the possibility of significant intra-study variation in clinical
diagnoses also cannot be discounted, given that several studies involved two or more gynecologists.

9.

While these studies have made significant contributions to knowledge of the prevalence of
gynecological morbidity within the community, significant lacunae persist in our
understanding of the determinants and factors influencing such morbidity, as well as its
consequences for women’s lives.

In almost all of the studies from India, the primary objective was estimation of the prevalence of

gynecological morbidity within the communities studied. Reflecting the predominantly biomedical or
epidemiological orientation of the researchers involved, comparatively less emphasis was placed upon
understanding the social, behavioral, and biomedical antecedents of such morbidity. The collection of
data on possible determinants of morbidity therefore tended to be parsimonious, and confined largely to
standard demographic or socioeconomic variables. In the subset of studies which did investigate
determinants, some significant (although not entirely consistent) associations existed between

demographic and socioeconomic characteristics and delivery practices and specific self-reported or

clinically diagnosed gynecological problems (Bhatia and Cleland, 1995; Parikh et al., 1996).

Perhaps

the most striking finding to emerge from these analyses has been the strong association between use of

contraception— either female sterilization or in some cases the IUD — and reported or diagnosed
gynecological morbidity (Bang, et al. 1989; Bhatia and Cleland, 1995; Oomman, 1996; Parikh, et al.

1996). This finding has also been reported elsewhere, but on the whole remains poorly understood."0

These findings aside, it is clear that our understanding of the range and importance of
determinants of gynecological morbidity — behavioral, environmental, iatrogenic, and biomedical—
remains at a very rudimentary stage in India and elsewhere. In particular, the roles of potentially key but

difficult-to-research determinants such as sexual behavior and practices, and other iatrogenic factors such

as unsafe abortion, in influencing women’s vulnerability to gynecological morbidity remain largely

unexplored. Similar important gaps in knowledge exist with respect to the role of male partners— in
terms of their own sexual and reproductive health, their importance as a source of transmission to other

partners, and their role in assisting or impeding women’s ability to address and resolve reproductive
health problems.

21

The studies also collectively shed little light on the consequences and implications of
gynecological morbidity for poor women’s daily lives. Little is known about how such morbidity
impacts upon women's ability to fulfill a diverse range of expected domestic and familial roles— from
economic productivity, to domestic responsibilities, to economic productivity, to marital and sexual

relationships— as well as upon their own mental health and psychological well-being. Moreover,
gynecological problems include a continuum of conditions which range from inconvenience to the

extreme of disimpairment or even death, and in the absence of more detailed research on severity, there
has been a tendency to view' all such conditions as equivalent. It is clear, however, that from the

perspectives of both individual women as well as broader public health, these conditions have widely
different implications in terms of severity, priority, and required intervention.

Conclusions
Among the seven community-based studies which have been considered in this review, wide variation is
evident in both self-reported as well as clinically-diagnosed levels and patterns of gynecological
morbidity. Given the geographical and cultural heterogeneity which characterizes India, it is plausible
that innate differences exist across population groups, and that no single set of estimates apply for India

as a w hole. However, it is also evident upon close examination of these studies that fundamental
differences exist across studies with respect to study designs and research methodologies. Indeed,
differences among studies are so marked with respect to such basic components as study participation

levels, possible selection bias, questionnaire wording and extent of probing, study enrollment periods,
and clinical definitions of morbidity that it would appear to make little sense to try and directly compare

results. These differences notwithstanding, the studies also collectively leave little doubt that
gynecological morbidity levels are unacceptably high, and constitute a major public health problem

among poor Indian women.
These studies also provide a number of important broader lessons for carrying out community­
based research in this area. The review has served to highlight the numerous and frequently formidable

methodological challenges inherent in undertaking community-based research on women’s
gynecological morbidity in settings such as India. It is very evident that simple adaptation of

conventional Knowledge. Attitude, and Practice (KAP) survey approaches—although useful for
studying a range of other health and family planning issues— is unlikely to prove effective for
investigation of an issue as complex and culturally sensitive as gynecological morbidity.

22

Experience from these studies also illustrates the extent of engagement with the communities—

before, during, and even subsequent to the study— which is required in order to elicit their active
cooperation *and participation. One strategy’addpted by many of the studies was to work with or through

voluntary organizations with long-standing ties and records of service to the communities being studied.
While desirable, it is important to recognize that simple affiliation with a voluntary organization is
unlikely by itself to assure women’s participation in such research, especially if the organization has no

previous record of addressing reproductive health issues within its community programs.21 An extended

and intensive period of rapport and confidence building with members of the community

both men as

well as women— has been shown to be essential for the successful implementation of such field
research. Equally important is an explicit commitment by the researchers from the onset to resolve, or

facilitate resolution, of participants’ gynecological problems which are detected over the course of the
study. Without assuming this obligation, it seems very doubtful that many respondents would consent to

actively participating in such studies.

The experience of these studies indicates that even if high participation rates can be attained,
serious methodological concerns remain regarding the reliability and validity of data on women’s

gynecological morbidity collected through self-reports. The studies highlight the importance of advance
in-depth, qualitative research, which assists in identifying the range of local terminology and more
nuanced indications of gynecological morbidity which women employ to describe such conditions. The

importance of building in detailed probes for morbidity within questionnaires has also been highlighted,
as has the need for greater standardization in areas such as questionnaire design, morbidity definitions
and reference periods.22 The experience of these studies also provide a strong argument in favor of

multiple interviews, and indicate that single round, cross-sectional surveys of gynecological morbidity

undertaken in largely unfamiliar communities— or even through voluntary organizations who have

worked closely with the respondents— are unlikely to adequately capture the true levels of gynecological
morbidity in the communities under study. Finally, several of the studies raise important questions about
the most appropriate choice of interviewers, and specifically whether clinicians may not often be more

effective in eliciting information from women about sensitive reproductive health concerns.
A central lesson to emerge from this review is that despite the best efforts of researchers,

significant sample attrition and some degree of selection bias for the clinical component of community­
based studies are almost inevitable in conservative settings such as India. It is therefore essential that

future studies anticipate such attrition, and build in appropriate mechanisms to learn more about

morbidity patterns among women who refuse to undergo clinical examination, in order to be able to

23

more accurately gauge the effects of selection bias upon reported results.23 Experience has underscored
the importance.of a longer study window and the need for repeat visits to initially reluctant or*
unavailable respondents in order to achieve scientifically acceptable levels of study participation. At the

same time, care must be taken to avoid extending the period of enrollment so widely

as was done in

some of the studies— that the results can no longer be considered to reflect true estimates of morbidity
prevalence.

Of commensurate but largely unrecognized concern is the widespread variation across, or even

within, studies in clinicians’ assessments of gynecological morbidity, and the need to move toward more
standardized clinical definitions and diagnoses. Although this paper did not consider the issue of

laboratory tests, this component undoubtedly adds another level of complexity and potential variability
across studies.
Another important lesson is that women's compliance in undergoing gynecological examinations

is likely to be most problematic in settings where women’s sexuality remains closely guarded and rarely
discussed, where women’s reproductive health problems are most widespread, and where low status and
autonomy constrain women’s ability to effectively resolve them. Within the context of India, and it
should therefore not be surprising that the studies which experienced the greatest difficulty were based in
rural areas and/or the North (e.g., the rural Gujarat and Rajasthan studies), and the most successful study
in terms of participation was from the southern state of Karnataka.

Finally, both in India and elsewhere, a general consensus is rapidly emerging on the magnitude
of the problems of reproductive tract infections and gynecological morbidity for women in the

developing world. Reproductive morbidity prevalence studies such as those considered have played a
major role in raising awareness of this neglected issue and placing it on the international and national
policy agenda. It is essentia), however, that the next generation of research studies move ‘beyond
prevalence’ in terms of primary focus. Research must begin to illuminate the socioeconomic, behavioral,

and biomedical causes of reproductive ill-health, the factors which impede women from prioritizing such
problems and seeking appropriate care, as well as the profound consequences of reproductive morbidity
for the quality of women’s and men’s lives. This will require more broad-based, multi-disciplinary
research initiatives and expertise which few projects have been able to successfully assemble thus far.

Equally important, in settings such as India, the research agenda must necessarily broaden to incorporate
action research which pilot tests the feasibility of replicable models of women-centered health care

which includes the prevention and treatment of gynecological morbidity.

24

ACKNOWLEDGMENTS

We are indebted to members of the seven research projects for taking the time and effort to
provide additional details and information regarding aspects of each field study. Particular gratitude is

extended to current or former members of the following projects: Rani and Abhay Bang of SEARCH,
Indumati Parikh, VijayalaxmiTaskar, Veena Mulgaonkar and Neela Dharap of Streehitakarini, Nandini
Oomman and Saral Dabir with URMUL Trust, K. Lath a Sasi and Shubuda Kanani of the Baroda
Citizen’s Council , Suttapa Battacharya and Prabha Dutta of the Child in Need Institute, Prodipto Roy

and Sarojini Kumtakar of the Council for Social Development, Gayatri Giri of SEWA-Rural, and Jagdish

Bhatia, S. Ramaswamy, and Leela Bhagavan of the Indian Institute of Management-Bangalore, whose
collective contributions made the preparation of this overview possible. We also gratefully acknowledge

the helpful comments of Leela Visaria and the assistance of Zarina Rao in the preparation of the
manuscript. The authors assume all responsibility for any factual errors in the text.

*

25

References

'

Bang. R. and A. Bang. “Women’s Perceptions of White Vaginal Discharge: Ethnographic Data from
Rural Maharashtra.” Pp. 79-94 in J. Gittelsohn, et al. (eds.) Listening to Women Talk about Their Health.

New Delhi: Har-Anand Publications, 1994.

Bang, R. and A. Bang. “A Community Study of Gynecological Disease in Indian Villages: Some
Experiences and Reflections”. Pp. 223- 237 in S. Zeidenstein and K. Moore (eds.) Learning about

Sexuality. New York: The Population Council, 1996.
Bang R.A., A.T. Bang, M. Baitule, Y. Choudhary, S. Sarmikaddam, O. Tale. “High Prevalence of
Gynecological Diseases in Rural Indian Women.” Lancet 8629 (January 14,1989): 85-88.
Baroda Citizens Council, Child in Need Institute, SEWA-Rural, and Streehitikarini. “Prevalence of
Clinically Detectable Gynecological Morbidity in India: Results of Four Community Based Studies.

Unpublished paper, 1995.

Bhatia, J.C. and J. Cleland. “Self-reported Symptoms of Gynecological Morbidity and Their Treatment
in South India.” Studies in Family Planning 26, 4 (1995): 203-216.
Bhatia J.C., J. Cleland, L. Bhagavan, and N.S.N. Rao. “Prevalence of Gynecological Morbidity among
Women in South India.” Indian Institute of Management- Bangalore, unpublished paper, 1995.
Bhattacharya, S., “Gynecological Morbidities among Women in West Bengal.” Child in Need Institute
unpublished report, 1994.
Brabin L., J. Kemp, O.K. Obunge, J. Ikimalo, N. Dollimore, N.N. Odu, C.A. Hart, and N.D. Briggs.
“Reproductive Tract Infections and Abortion among Adolescent Girls in Rural Nigeria”. Lancet 8945
(February 4, 1995): 300-304.

Bulut, A., N. Yolsay, V. Filippi, and W. Graham. “ In Search of Truth: Comparing Alternative Sources
of Information on Reproductive Tract Infection.” Reproductive Health Matters 6 (1995): 31 -39.

Dixon-Mueller, R. and J. Wasserheit. The Culture of Silence: Reproductive Tract Infections among
Women in the Third World. International Women’s Health Coalition, 1991.

Germain, A., K.K. Holmes, P. Piot, and J.N. Wasserheit. Reproductive Tract Infections: Global Impact
and Priorities for Women's Reproductive Health. New York: Plenum Press, 1992.
Gittelsohn, J., M.E. Bentley, P.J. Pelto, M.Nag, S. Pachauri, A.D. Harrison, and L.T. Landman (eds.)
Listening to Women Talk about Their Health. New Delhi: Har-Anand Publications, 1994.

Graham, W., C.C.A. Ronsmans, V. G.A. Filippi, O.M.R. Campbell, E.A. Goodburn, T. F. C. Marshall, C.
Shulman, and J.L. Davies. Asking Questions about Women s Reproductive Health in Community-based
Surveys: Guidelines on Scope and Content. London School of Hygiene and Tropical Medicine: Maternal
and Child Epidemiology Unit Publication No. 6, April, 1995.
Jeffcoate, N. Principles of Gynaecology, 9th edition, Butterworth Scientific, London, 1988.

26

Kanani, S.. K. Latha, and M.Shah. “Application of Qualitative Meth'odofogies to Investigate Perceptions
of Women and Health Practitioners Regarding Women’s Health Disorders in Baroda Slums. Pp. 116130 in J. Gittelsohn, et al. (eds.) Listening to Women Talk about Their Health. New Delhi: Har-Anand
Publications, 1994.
Khattab, H.A.S. The Silent Endurance: Social Conditions of Women s Reproductive Health in Rural

Egypt. Cairo, UNICEF, 1992.

Khattab, H., H. Zurayk, N. Younis, and O. Kamal. “Involving Women in a Reproductive Morbidity
Study in Egypt.” Pp. 238-250 in S. Zeidenstein and K. Moore (eds.) Learning about Sexuality. New

York: The Population Council, 1996.
Kumtakar, S. and P. Roy, Council for Social Development. Unpublished results, 1996.

Latha. K., S. Kanani, N. Maitra, and R.V. Bhat. “A Quantitative and Qualitative Assessment of
Morbidity among Urban Disadvantaged Women in Baroda, India.” Unpublished report, 1994.

Measham. A.R. and R.A. Heaver. India's Family Welfare Program: Moving to a Reproductive and Child
Health Approach. Washington, D.C.: The World Bank, 1996.
Mulgaonkar, V. B., I. G. Parikh, V.R. Taskar, N.D. Dharap, and V.P. Pradhan.“Perceptions of Bombay
Slum Women Regarding Refusal to Participate in a Gynaecological Health Programme.’ Pp. 145-167 in
J. Gittelsohn, et al. (eds.) Listening to Women Talk about Their Health. New Delhi: Har-Anand
Publications, 1994.
Oomman, N.M. Poverty and Pathology: Comparing Rural Rajasthani Women’s Ethnomedical Models
with Biomedical Models of Reproductive Behavior. Ph.D. Thesis. Johns Hopkins University, 1996.

Parikh, I., V. Taskar, N. Dharap, and V. Mulgaonkar. “Gynaecological Morbidity among Women in a
Bombay Slum.” Streehitakarini unpublished paper, 1996.
SEWA-Rural Research Team. “Gynecological Aspects of Women’s Health: A Cross-Sectional Study.”
SEWA-Rural unpublished research report, no date.

Streehitakarini. “Gynecological Diseases and Perceptions about them in a Bombay Slum Area.”
Unpublished report, 1995.

Wasserheit J.N., J.R. Harris, J. Chakraborty, B.A. Kay, K. Mason. “Reproductive Tract Infections in a
Family Planning Population in Rural Bangladesh.” Studies in Family Planning 20y 2 (1989): 69-80.

Women’s Work and Child Development Research Network. The Impact of Women’s Work on Child and
Maternal Health. Final report, 1995.
Younis N.. H. Khattab, H. Zurayk, et al. "A Community Study of Gynaecological and Related
Morbidities in Rural Egypt.” Studies in Family Planning 24, 3 (1993): 175-86.

T1

. Footnote^
1 The Ford Foundation has also supported two other community-based prevalence studies of
gynecological morbidity during this period. A study by SARTHI, an NGO based in rural Gujarat, was
discontinued due to both logistical problems related to the clinical and laboratory components and
inadequate community participation. The other study, carried out by the Gujarat Institute of
Development Research, is presently in the field.

2 Other important studies of gynecological morbidity are from Egypt (Younis, et al. 1993), Bangladesh
(Wasserheit, et al. 1989), and Nigeria (Brabin, et al. 1995).
3 Measham and Heaver (1996).

4 The studies which included laboratory components are rural Maharashtra, Bombay, West Bengal,
Karnataka, and Rajasthan.

5 Although the original household enumeration survey was undertaken in 1984, the Bombay study did
not actually begin until 1988. Due to out-migration or demolition of the originally enumerated dwelling
units , the sampling frame for the study was reduced from the original number of 1500 households to a
revised number of 1048 households. See Mulgaonkar, et al. (1994) for a discussion.
6 This included 68 women who were widowed and 28 women who were separated from their spouses.

7 The West Bengal Study also collected data on gynecological morbidity from 150 urban slum women in
Calcutta. Only the results of the rural study are considered in the present paper.
8 Although this information was not reported separately from the laboratory and self-reported morbidity
data in the rural Maharashtra study, it is evident that levels of clinically diagnosed morbidity in this study
also exceeded 70 percent of all women.
9

*

The rural Maharashtra study, for example, reported that only 8 percent of women in their study had
ever undergone a gynecological examination (Bang et al., 1989).
10 Descriptions from the Giza study in Egypt makes it clear that the intensive support and active
intervention by project researchers to ensure that women who participated in their study received
appropriate treatment for identified conditions was a key factor behind the extremely high participation
rate achieved by this study (see Khattab, et al. 1994 and 1996).
11 Although not well-documented, there was clearly variation among studies in terms of the how this
information was incorporated into the survey and clinical study components. At least one study (rural
Gujarat) undertook in-depth exploration after completion of the survey phase.

12 For a description of this research initiative, see Women’s Work and Child Development Research
Network (1995).
13 Council for Social Development (1996).

28

levels of reported conditions such as prolapse or dysuria.

figure when interviewed at home by an interviewer was 65 percent. See Bulut et al. (1^3)
ised largely of self-selected clinic attenders, who would be
16 Since this study population was comprise'
more readily, an even larger gap between unprompted and prompted
expected to volunteer information i
responses might be expected amon g the general population of women.

17 Bhattacharya, p. 29.

18 This was only possible in studies which fielded their surveys of self-reported1 morbidity among all
who returned to interview a
sampled women prior to and separately from the clinical exammation, or
representative sample of women who refused the clinical component.
19 The Karnataka analysis was based upon data from a larger cross-sectional survey. See Bhatia and
Cleland (1995).

20 Similar findings were reported other studies from Bangladesh and Egypt. See Wasserheit, et al. (

)

and Younis, et al. (1993).

SiSSsESSS?
general, where such services are largely unavailable.

22 See, for example, Graham et al. (1995).

and validity of such information.
24 The Giza study in Egypt represents a notable and important example.

29

Figure 1.’

Location of Seven Community-based Studies of Gynecological
Morbidity in India

/

$

0 Rajasthan
Studv

r.tst Bengal Study

airay' Mahar
• ? Bombay St

°Kar

rtudy

a Study

M:

ntra

Table 1. Description of Community-based Gynecological Morbidity Studies in India
Location
and Year of
Study

NGO

Sample

Affiliation

Data Collection
Procedure

Study
Enrollment
Period

Study Population

Ever married and
unmarried women
aged 13 years and
older
Ever married and
unmarried women
aged 13-45 years

Rural
Maharashtra
(1989)

SEARCH

100% sample
in 2 rural
villages

Concurrent interview
and examination at
village level clinics

5 month period, 3
days per week

Rural West
Bengal
(199^0-91)

Child in Need
Institute

Concurrent interview
and examination at
central medical facility

Bombay
(1988-89)

Streehitikarini

Quota sample
of 500 women
in 8 rural
villages
10% random
sample of
project
households in
one urban
slum area

18 month period,
two clinics per
week
2-3 repeat visits
12 month
duration; open
enrollment for
clinical
examination over
entire period

Baroda
(1990-91)'

Baroda
Citizens
Council

50% random
sample from
two urban
slum areas

Karnataka
(1992)

No affiliation

Rural Gujarat
(1988-89)

SEWA Rural

Rural Rajasthan
(1994)

URMUL Trust

Subsample of
a random
sample of 48
villages and
one town
100% sample
in 10 rural
villages
100% sample
of two rural
villages

Interviews conducted at
home and concurrently
with clinical
examination;
examination conducted
at NGO clinic or near
home
Interviews conducted at
home and concurrently
with clinical
examination;
examination conducted
at NGO clinic
Interview conducted
monthly at home;
examination conducted
at project office clinic
Concurrent interview
and examination at
village health fairs
Interview conducted at
home; examination
conducted at village
health clinics

Ever married
women aged 15
years and older

Sample for
Morbidity
Clinical
Sun'cy
Examination
(% participating)
654 women 650 women and
(includes 92
unmarried girls
unmarried girls)
(59%)
(59%)
500 women
500 women and
(includes 237
unmarried girls
unmarried girls)
(44%)
(44%)________
756 women
756 women
(72%)
(72%)

12 month
duration; within
2-3 days of
interview

Ever married
women aged 18-45
years

840 women
(100%)

548 women
(65%)

12 month
duration; day
following last
interview

Ever married
women less than 35
years of age with a
child < 6 mos.

440 women
(100%)

385 women
(87%)

2-4 days during
village health fair

Ever married
women aged 15
years and older
Ever married
women aged 15
years and older

1103 women
(49%)

324 women
(29%)

274 women
(100%)

51 women
(19%)

2 month
duration; 1-2
days in third
month

Tabic 2. .Prevalence of Self-Reported and Clinical Gynecological Morbidity Among Community-based Studies in India

Morbidity

Rural
Maharashtra

Rural West
Bengal

Bombay

Baroda

Karnataka

Rural
Gujarat

Rural
Rajasthan*

% of Women with
SELF-REPORTED:

Menstrual problems

60

33

41

58

65

59

48

Excessive discharge

13

50

31

22

22

57

78

Lower abdominal pain

13

17

21

9

16

NA

10

Lower backache

30

5

39

24

NA

30

16

Dyspareunia

7

2

NR

NR

1

NR

48

55**

65

74

65

NA

84

100

Vaginitis

62

4

15

11

13

10

30

Cervicitis

48

14

40

13

24

8

19

Cervical erosion

46

2

21

5

10

20

30

Pelvic inflammatory disease

24

1

16

8

11

8

36

Prolapse

0.5

17

18

5

NA

27

One or more conditions

NA

43

74

26

43

77

One or more conditions

CLINICALLY DIAGNOSED:

NR = Not recorded
NA = Could not be ascertained from available data

• * Data from community and clinic-based samples combined
** Does not include lower backache and lower abdominal pain

70

Figure 2: Women’s Self-Reported Gynaecological Problems According to Round of Data Collection: Karnataka
and Haryana
40
37.9

I

35
33.2

30

28 9

25
226

20
17.3

14 5

15

13 1 -------

17.3 17.3
_________ 16.5

14 2

13

fi

10 7

10

89
7 8 ___

6.5

I

2

3

5

6

7

Round

Karnataka

Sources: Bhatia, et al., unpublished data
Kumtakar and Roy, unpublished data

8

9

10

TT

12

s I


6.2

5

0

II

T

t

Round
Haryana

5

6

Table 3. Women’s Unprompted and Prompted Self-Reported Gynaecological Problems:
Rural Rajasthan, 1994

/

Unprompted

Both Unprompted and
Prompted

%

%

Menstrual Problems

27.9

47.9

Prolapse

10.0

19.6

Problems during Sexual
Intercourse

1.3

47.7

Vaginal Discharge

68.0

78.8

Infertility

9.8

14.9

Source: Oomman, 1996, p. 198

I

Table 4. Percentage of Women Reporting Gynecological Problems Among Women who
Underwent and did not Undergo Clinical Examinations

I

Study

Women who underwent
a gynecological
examination

Women who did not
undergo a gynecological
examination

% Reporting

Rural Gujarat
Vaginal Discharge

55.9

22.5

Menstrual Problems

84.9

70.4

(324)

(779)

Vaginal Discharge

22.3

15.1

Menstrual Problems

58.0

56.2

(548)

(292)

Vaginal Discharge

13.5

8.2

Menstrual Problems
- scanty/irregular
- profiise/dysmenorrhea

19.5
20.0

31.3
17.9

(N)

(650)

(105)

(N)

Baroda

(N)

Rural Maharashtra

Source: Baroda Citizens Council, et al. (1995) and R.Bang et al. (1989)

Table 5. Clinical Definitions of Morbidity Employed in Community-based Studies of Gynecological Morbidity

Study

Composition of

Vaginitis

Cervical Erosion

Cervicitis

Clinician Team

Rural
Maharashtra

Rural West
Bengal

Prolapse

Disease

1 female
gynecologist

Inflammation of vaginal

All erosions

wall and lab evidence

Cervical
inflammation with

Palpable and tender
tissues around uterus

pus discharge and lab

and cervical motion

evidence of infection

tenderness and lab

Based on traction and

straining

evidence of infection
I female
gynecologist

Inflammation of vaginal
wall with or without
discharge and lab

All erosions

evidence

Bombay

Pelvic
Inflammatory

2 female
gynecologists

Inflammation of vaginal
wall with discharge and
itching/pain

All erosions

Cervical
inflammation with

Tenderness or fixity
of tissues around

pus discharge and lab
evidence of infection

uterus and lab
evidence of infection

Cervical
inflammation with

Lower abdominal
pain and tenderness

pus discharge or

or thickening of

cases with inflammed

tissues around uterus

Based on straining

Based on traction and
straining

scarred cervix

Baroda

3 female
gynecologists

Inflammation of vaginal
wall

All erosions except

Cervical

small, asymptomatic
erosions

inflammation with

Tenderness or fixity
of tissues around the

pus discharge or

uterus

cases with in flam med

Based on traction and

straining; excludes
mild cystocele and
Grade I descent

scarred cervix

Karnataka

1 female

Inflammation of vaginal

All erosions except

gynecologist

wall with discharge

small, asymptomatic
erosions

Cervical
inflammation with

pus discharge

Uterine tenderness
and tenderness of
tissues around uterus

Based on straining

and clinical cervicitis
Rural Gujarat

Rural Rajasthan

2 female
gynecologists

! n flam mat ion of vagina I
wail with abnormal
discharge

All erosions

1 female

Inflammation of vaginal

All erosions

gynecologist

wall with abnormal

discharge

pus discharge

NA = Could not be ascertained from available data

Cervical
inflammation with
pus discharge or all
cases with scarred
cervix

Tenderness or fixity
of tissues around
uterus

Cervical

Abdominal pain and

Based on traction and

inflammation with

uterine tenderness

straining

NA

io h

io

X#,

REPORT OF THE PROJECT

PREVALENCE OF

MATERNAL MORBIDITY

IN A

t

SOUTH INDIAN
W:...

,

:W:;

:



COMMUNITY

-



"kni r

i

pl V

■fai

Dr. D.K. SRINIVASA
Dr. K.A. NARAYAN
Dr. ASHA OUMACHIGUI
Dr. GAUTAM ROY

f

11

^1
2

FORD FOUNDATION STUDY
Department of Community Medicine
Jawaharlal Institute of Postgraduate
Medical Education and Research
Pondicherry - INDIA

i .. < f /'

i

I
PREVALENCE OF MATERNAL MORBIDITY IN A
SOUTH INDIAN COMMUNITY

X

Dr. D.K. SRINIVASA
Dr. K.A. NARAYAN
Dr. ASHA OUMACHIGUI
Dr. GAUTAM ROY

INDIA
MARCH, 1997

Jawaharlal Institute of Postgraduate Medical Education and Research
Pondicherry - 605 006.

Address for communication :
Dr. K.A. Narayan
Professor
Department of Preventive and Social Medicine,
Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER),
Pondicherry - 605 006, INDIA.
Phone : 72380 - 89 Ext. 269 Fax: (413) 72067
E-Mail: director@jipmer.ren.nic.in

/

Cover design:

Dr. K.A. Narayan
S. Sundarambal

Computer Illustration:

C. Ramasamy

Computer Composing:

C. Ramasamy
S. Paramesvary

Printed by :
All India Press,
Kennady Nagar,
Pondicherry - 605 001.
INDIA. Phone : 34979

r
Acknowledgements
x This study was the outcome of the initiative of Dr. Saroj Pachauri, Programme officer, Ford
Foundation and currently Regional Director, The Population Council. She mooted the idea of a multi
centric study on the important area of maternal morbidity, guided us through the initial stages of
protocol formation and assisted in the funding of the project. We would like to express our deep
appreciation for her effort.

We are indebted to Ford Foundation without who’s assistance this study would not have been
feasible. The present Programme Officer Dr. Michael Koenig and Grants Administration Officer
Ms. Neera Sood, helped us during the various phases of the project. Our thanks to them.
We would like to place on record our gratitude to Dr. D.S. Dubey, Director JIPMER, The
Ministry of Health and Family Welfare, Government of India and The Indian Council of Medical
Research for according permission to carry out the project.

The extensive discussions held with Dr. Judith Fortney and Dr. Jason Smith of Family Health
International, Research Triangle Park, NC, USA, and the Principal Investigators of the other country
studies, Dr. Halida Hanum Akhter of BIRPERHT, Bangladesh, Dr. Anna of Indonesia and of
Egyptian Fertility Care Society helped in formulating the study design, developing the questionnaire
and completing the analysis and reporting. We are extremely grateful to all of them.
Our thanks to Dr. J.M. Daisy, Dr. V.J. Sadhana, Mr. Krishna Rao, Mr. P. Rajendiran for
their assistance in conducting the clinical camps and to Dr. M. Danabalan, Dr. Ajit Sahai and other
staff of the Department of P & S M for their co-operation.
Mr. G.K. Padmanabhan, Mr. C. Ramasamy, Mrs. S. Paramesvary, Ms. S. Sundarambal,
Ms.S. Veerammal, the other project staff and field investigators and Mrs. Vijayalakshmi
Vishwanathan of MODE Research need special mention for the effort they have put in for the project.
Lastly this project would not have been feasible without the co-operation of the respondents.
We are ever indebted to them.

■A

■.

r><

WH
riJ

*2

IL

y.



1

f

■■"

w

xT*

:'r ■

*t

^'jff..
.<

La
____

P’ * "'i

O1
r.s
|3

JK' M

F \3

A Os

,.s

Stone Carving of Delivery being conducted in squatting position
- Villianur temple, Pondicherry

CONTENTS
Page No.
1.

Summary

2.

Introduction

i - iv

1
1

Conceptual framework
Objectives

3.

4.

2

Material and Methods

3

Study Design

3

Study Population and Sample Size

3

Sampling Procedure

4

Questionnaire Development

4

Recruitment and Training of Interviewers and Supervisor

5

Field work for Data Collection

6

Clinical Examination

6

Quality Control

7

Data Analysis

7

Results

8

Study Population

8

Selected Characteristics of Study Population

8

Characteristics of Index Pregnancy

10

Prevalence of Maternal Morbidity

11



Number of morbidities per woman ! pregnancy

11



Antepartum morbidity

11



Intrapartum morbidity

12



Postpartum morbidity

14



Longterm morbidities

15



Prevalence of morbidities

15



Morbidities following abortions

16



Age, Parity, Residence and Socio-economic status and
morbidities

17

Prevalence of morbidities and intrapartum care

17

5.

Discussion

20

6.

Intervention

24

7.

References

26

8.

List of Acronyms

28
Contd..

9.

Tables

10.

Annexures
i-iii
iv
V

vi

Study questionnaires
List of project personnel
Training schedule
List of selected areas

V

Summary

Introduction

Pregnancy and childbirth place a woman at a higher risk of morbidity and mortality. Though
a fair degree of success has been achieved in reducing maternal deaths and improving maternal care,
several women continue to suffer from morbidities in the antepartum, intrapartum and postpartum
periods due to reasons directly or indirectly attributable to the pregnancy. These may be life
threatening or serious or may incapacitate a woman for varying lengths of time, sometimes leaving
long term sequelae. A woman s care seeking behaviour for these morbidities are not only dependent
on her own perception of the morbidity but on other factors like perception of her family members
about the condition, the availability, accessibility and acceptability of health services. There has been
just one community based study in India which found that for every maternal death 16 women
suffered from illness during pregnancy, childbirth or within six week of delivery (Datta et al 1980).
In order to obtain better estimates and to find out the various factors influencing the health seeking
behaviour of women during pregnancy and childbirth, this study was undertaken. It was a part of
a multicentre study carried out in four countries - Bangladesh, Egypt, Indonesia and India. It was
funded by The Ford Foundation, New Delhi and coordinated by Family Health International (FHI),
North Carolina, USA. Though the four country studies had a common overall approach and a core
set of research questions, each study had its own unique identity.
The objectives of the study was to estimate the prevalence of specific maternal morbidities
during pregnancy, parturition and puerperium, study the perceptions of maternal morbidities during
the three periods and determine the correlation between these and medically recognised risks and
morbidities, determine social, cultural, economic and institutional reasons why women do not seek
timely health care during pregnancy and childbirth and to suggest interventions based on the results
of the study to reduce the prevalence of specific morbidities.

Methods
The study was conducted during 1992-95 by the Department of Community Medicine of
Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Pondicherry in the
Pondicherry region of the Union Territory of Pondicherry and the adjoining of then South Arcot
district in Tamil Nadu.

It was planned as a population based cross-sectional questionnaire survey of an estimated 4000
women who had been pregnant in the previous two years irrespective of the outcome of their preg­
nancy. The survey was conducted in two stages from February to May 1993 and February to March
1994 in the 45 villages and three towns of South Arcot District of Tamil Nadu and Pondicherry. All
the pregnancies occurring from 14 January, 1991 to 13 January, 1993 in 13,500 households were
included as index pregnancies for the study. Eligible women were identified and interviewed by
trained lady investigators by door to door canvassing. Depth interviews were conducted in a subset
of the respondents who had suffered either serious, chronic or life threatening morbidities to capture
perceptions of seriousness, the health seeking behaviour, the decision making process, means of trans­
port, the cost and the time spent in reaching a health facility.
Clinical examination by a gynaecologist was done for a subgroup of women with selected
morbidities to corroborate the responses obtained by the questionnaire.

i

Results

A total of 13,235 households were visited, of which 61 percent were from rural areas and
39% from urbanJ areas- In these households, 3686 eligible women were identified and 3339 interviews
completed. About 14% of the women had more than one pregnancy and some though a small
number, had even four pregnancies, hence the total number of pregnancies studied were 3844.

Majority of the women (76.2%) were between 20 and 30 years of age. The mean number of
pregnancies was 2.6 (+ 1.5). Primiparous women formed 25.6% of the study population. Women
in the urban areas and those from Pondicherry rural areas had lower number of pregnancies than
women from rural Tamil Nadu. The same difference was seen for the total number of live births.
<?

About half of the women (48.9%), had attended school. Literacy was higher for the urban
women in general and for Pondicherry area in particular.

Majority (81 %) of the families belonged to the low or middle socio-economic status. Urban
women from Pondicherry were economically better off than those from other areas. About half of
the women (49.7%) were working for wages outside their homes. More often women from rural
areas, especially those from Tamil Nadu, worked outside their homes for wages.

Ninety eight percent of the women had antepartum check-up from a government health centre
with each women receiving upto 3.2 visits on an average. The urban women in both Pondicherry and
Tamil Nadu and the rural women of Pondicherry had higher utilisation of services than women of
Tamil Nadu rural area. The first visit for antepartum care in as many as 44% women was in the
first trimester itself.
About eighty eight percent of 3844 pregnancy resulted in live birth, two percent in still birth
and 6.8% in abortion. The proportion of pregnancies medically terminated was 3.5%. Ninety
percent of deliveries were normal, 5.5% were by Caesarean section, and for the rest forceps had to
be applied or vacuum extraction was done. Nearly 60% of deliveries were institutional.
One or more morbidity was reported in 56% of the pregnancies, about 39% during the
antepartum period, 18.7% during intrapartum and 24.6% during postpartum period. In 25% of
pregnancies, there was more than one morbidity.

Potentially life threatening complications were reported in 4.7% of the women during the
three phases of pregnancy and serious morbidities were reported in 40.7% of the pregnancies during
all the 3 phases combined.

The commonest life threatening morbidity was, antepartum bleeding, malaria and
fits/convulsions. Severe vomiting, severe edema of hands and feet, fever of more than 3 days duration
and pregnancy induced hypertension were the serious conditions reported frequently. The other pre­
existing conditions which would potentially increase the risk, and therefore considered as serious,
were heart disease, pulmonary tuberculosis, jaundice and diabetes mellitus.
Seventeen percent of women reported morbidity during intrapartum period, 2% were
potentially life threatening, about 9% were serious morbidities.

ii

5

Excessive bleeding, loss of consciousness and fits / convulsions were the life threatening
conditions. The three common serious morbidities reported were, early rupture of membranes,
abnormal position of foetus and labour more than 18 hours.

Life threatening morbidities, were slightly higher (23 per 1000) among home deliveries
compared to hospital deliveries. Fewer serious morbidities were, however, reported from home
deliveries compared to hospital deliveries. A disturbing finding was that a large (9%) of women
delivering in hospitals had genital tears. This reflects on the quality of care.

Overall, in 23% of pregnancies, women reported illnesses during the postpartum period.
Potentially life threatening conditions like excessive bleeding, fainting or unconsciousness, and fits
were reported in 3% of pregnancies and serious morbidities like fever of more than three days, pelvic
pain in 9% of pregnancies. Morbidity due to infections were common during this period. Infection
of perineal tear, urinary infection, infection of the incision of puerperal sterilisation and infection of
breast were the common morbidities.
Government sources (health centre or hospital) and private sources were utilised equally for
maternal morbidities. Primary Health centres were hardly used for deliveries (2.5%).

Inability to pay for treatment and pressure of work at home, especially care of the young child
were the frequent reasons given for not seeking care.

Any condition attributable to pregnancy but which persisted beyond the postpartum period
was considered as long term morbidity. Uterine prolapse was the most frequent condition reported.
Passage or dribbling of urine while coughing, sneezing or straining (incontinence of urine) in the
initial period of postpartum was reported by some women but the symptoms did not persist and
returned to normalcy with passage of time.
A

In 42% of the pregnancies, there was more than one morbidity during the three stages of
pregnancy. Multiple morbidities were reported more during the antepartum period (25%), than in
intrapartum (13%) and postpartum (11%) periods.
Ratio of morbidity to mortality was higher than expected. For every maternal death there were
478'morbidities. There were 328 serious or life threatening morbidities per maternal death.

Discussion
Several issues emerge from our study which are of concern to planners, administrate
professionals and opinion makers. The prevalence of maternal morbidity was far higher than
expected. For instance during the three phases nearly 60% of women experienced at least one
morbidity and in 25% of pregnancies there was more than one morbidity. The morbidities were
highest during antepartum period (39%), lower during the postpartum period (25%) and least during
intrapartum period (19%). Fortunately, potentially life threatening complications were reported only
in 5% of the women during the three phases of pregnancy but serious morbidities were reported in
as many as 41% of the pregnancies. Further, the number of morbidities in index pregnancy per
maternal death was 450 during pregnancy, childbirth, or puerperium regardless of severity. Two
hundred and thirty eight women experienced a morbidity for every maternal death, Here it may be
recalled that the most often quoted community study by Datta et al had reported 16 maternal
morbidities per maternal death, and many estimates were based on this. That maternal morbidities are
iii

death (Datta et al, 1985) and an estimated maternal mortality rate of 3 per 1000 live births for Tamil
Nadu and Pondicherry (Health Information India, 1991). The crude birth rate according to the 1991
census for Union Territory of Pondicherry was 21.6 per 1000 population and 23 per 1000 for Tamil
Nadu. Based on these rates we estimated that there would be about 4000 pregnancies in about
100,000 population over a period of two years.
Sampling Procedure

The procedure adopted for selection of the sample is given in Table I.
Questionnaire development

Three questionnaires were prepared. They were:
1. Area Profile Recorder Questionnaire
2. Household Eligibility Questionnaire
3. Mothers Morbidity Questionnaire

The questionnaires are given in Annexure (i - iii).
For the development of the questionnaires
two meetings were held in New Delhi and Ban­
galore sponsored by the Ford Foundation. They
were attended by the principal investigators of the
four projects, consultants from FHI, and a medi­
cal anthropologist. In the first meeting a core set
of questions were formulated for the preparation
of questionnaires to be used for interviewing
respondents. In the second meeting, the draft
questionnaire prepared in collaboration with FHI
was reviewed and revised. For our study, the FHI
contracted the assistance of a social research
organization ‘MODE’, based at Madras, to further
develop and pretest all the four questionnaires.
The steps in the development of the questionnaires
are described in (Fig. 2).

MODE briefed on the requirements
of questionnaire

A draft flow chart prepared without inputs from
the field
"Free listing" and focus group discussion con­
ducted to prepare list of local terminology and
to design personal and sensitive questions.

English version of questionnaire developed.
Translation - retranslation to Tamil

Field testing.

The Ford Foundation organised a series of
workshops on qualitative research methods and
they were of great use in the development of the
questionnaire.

Finalisation and printing of
questionnaire.

A lot of preliminary work was done using
the qualitative methods. First was ‘free listing’,
i.e., building up of vocabulary - learning and understanding of terms and expressions used locally for pregnancy related conditions. This was done by group discussion with elderly women and tradi­
tional birth attendants. Gynaecologists well versed in the Tamil language were used as ‘key
informants . Their views regarding obstetric morbidities, their opinion about what constitutes serious
and life threatening conditions, the usual conditions for which women come to the hospital and also
the expressions / symptoms* with which women approach them were collected. Focus group
discussions were held with traditional birth attendants to find out their perceptions of pregnancy care,
illness considered serious by them and remedial measures given to the women.
Fig 2: Process of questionnaire development

4

o-.

Group discussions were also held with elderly women and women having one to two children
to learn about the usual practices followed during pregnancy, childbirth and in the postpartum period,
their views on pregnancy related illness, the ones that were perceived as serious and the remedies
used.

(

About 15 women who had actually suffered from some antepartum morbidity were identified
from JIPMER hospital records and depth interviews were conducted with them. Many of the
interviews were held in more than one sitting. The women gave vivid descriptions of the problems
they had during their pregnancy. Based on these interviews, scenarios and flow charts were created
tracing the path/route taken by them to reach a hospital for the morbidity and finally for delivery.
The above experience provided an insight and rich material which was used in framing the questions.
The pretested questionnaires were translated and retranslated from English to Tamil and viceversa thrice and were finalised.


Recruitment and Training of Interviewers and Supervisors

A team of lady field investigators from rural background, with schooling up to tenth class
and fluent in Tamil were recruited. They were used for conducting the household survey, identifying
and interviewing the eligible women. They were supervised by three female and one male field
supervisors(Annexure-iv). The entire team was given a rigorous training for two weeks (Annexure-v).
The purpose of the training was on understanding the scope of the study and development of
interviewing skills. Role play, small group discussions, lecture cum discussions and mock interviews
were the methods employed during
the training. Emphasis was laid on
how to initiate an interview, inter­
viewing decorum, building rapport
(Orary)
and displaying empathy. In the core
session of the training the interviewers
were divided into small groups and
each interviewer was made to con­
duct "mock interviews’* using the
various questionnaires till they were
fluent with the questions and the
coding. Some of the interviewers and
supervisors were also trained in quali­
tative methods.

To ensure that the inter­
viewers themselves did not have any
misconception which would affect
data collection on pregnancy, child­
birth and puerperium they were intro­
duced to medical aspects such as the
anatomy of female reproductive or­
gans, physiology of menstruation,
Ce-1
pregnancy and the morbidities that
l» •'5
could occur. The interviewers were
(InU*
asked to draw ‘body maps’ of preg­
nant women before and after training.
This improved their understanding of
Figure 3: Body map drawn by an interviewer
the anatomical parts such as uterus
placenta, umbilical cord, position of
uterus, birth passage, and the pregnancy related illness (Fig. 3). The interviewers felt that this
exercise was "extremely useful" and " gave meaning to the questions we are about to ask".
5

Field work for Data Collection

The survey was conducted in two stages from February to May 1993 and February to March
1994 in the 45 villages and three towns of South Arcot District of Tamil Nadu and Pondicherry ,,
(Annexure-vi). All the pregnancies occurring from 14 January, 1991 to 13 January, 1993 in 13,500
households were included as index pregnancies for the study.
Before starting the data collection work, in every village and urban wards one or two
meetings were organised to explain the purpose of the study and to obtain cooperation from the
community. Village social maps were prepared by collective effort by a group of local people. They
were sketches of houses and other land marks such as temple, water tank etc. drawn on mud, floor
or similar surface. The families living in the houses were identified by using grains, leafs, match
sticks or any locally available material. Different coloured grains, different sizes of pebbles were
used to indicate adult men, women, pregnant women and children. There was a lot of interaction
amongst the people and investigators while drawing these maps. This participatory technique enabled
the field investigators to develop a good rapport with the people and these ‘Social maps’ supplemented
as well as hastened the door to door canvassing for identifying eligible women.

Through the household survey, village social maps and verification of records maintained at
the primary health centres and anganwadis, the number of eligible women identified for inclusion
in the study were 3686. Of these 3500 women (95%) could be interviewed. Mothers Morbidity
Questionnaire was the main instrument used for interviewing women. It contained different sections
for recording identification features, socioeconomic and personal attributes, past and present
pregnancy history with their outcomes, any illnesses or sequel and treatment seeking behaviour. It
had many open-ended questions.

The field investigators were encouraged not only to record conversation verbatim where
necessary, but also to prepare flow charts of ‘important’ events such as the perceptions, decision
making process involved, expenditure incurred on transport, doctors’ fee and medicines.
It took about one and a half to two hours to complete a questionnaire. Sometimes, more than
one sitting was required. The interviews were in Tamil after obtaining informed consent and assuring
confidentiality. It was held in the homes of respondents at a time convenient to them, maintaining
privacy. Often interviews had to be conducted late in the evenings after the women returned from
work. The people in rural areas in general and women in particular were cooperative. This is
reflected in the low dropout rate of only 5 percent.
Depth interviews were done later by the supervisors or some of the investigators specially
trained for the purpose, for 220 of the women who had suffered a serious morbidity.

In the second stage of the survey, the 847 women pregnant during the first survey were
revisited and interviewed again so that morbidities, if any, experienced by them during delivery or
postpartum period are also included to complete the data.

Clinical Examination
Doubts are expressed about women’s accounts, problem recognition and recall. We arranged
a gynaecologist to do clinical examination of women with reported morbidities to validate the
women’s responses. The examination was arranged in a few individual villages as well as in five
‘medical camps’ in May and June 1993. The response was better in the camps, for, according to
some women, the camps provided ‘more privacy’; and ‘would not draw attention of others’. Out of
268 women called, 136 underwent examination. There was good correlation between clinical
examination and the conditions reported by women. Women needing further investigation or treatment
were referred to JIPMER hospital.
6

*

Quality Control
Each interview schedule was reviewed by the supervisors to ensure that the data was complete
and accurate. Spot checks and reinterview of 10 percent of the index women were carried out by the
supervisors and the authors. Weekly meetings were held with the investigators for trouble shooting.

Data Analysis

After data collection all the questionnaires were scrutinized for discrepancies in data. Coding
errors detected at this stage were corrected after revisits to the respondents. All answers to openended questions, marginal notes and depth interviews were translated to English. Data was entered
into computers using a validation software developed on Foxbase.
Dummy tables were developed in collaboration with the FHI. The analysis was done with the
help of SPSS software. Prevalence of morbidity was calculated for the three phases of pregnancy
combined and as well as separately. Rates for life threatening and serious morbidities were also
calculated. The total pregnancies (3844) were used as denominator for calculating the antepartum and
combined morbidity rates, whereas the total deliveries (3449) were used for calculating the
intrapartum and postpartum morbidity rates. A content analysis of the notes and depth interviews was
done with ‘text - search’ software.

■ 4

A village social map being made

2

7

Results
Study Population

A total of 13,235 households were visited, of which 61 percent were from rural areas and
39% from urban areas. In these households, 3686 eligible women were interviewed (Table 2). Of
these, 3500 women were available for interview, the remaining 5%, either were unavailable or
refused to be interviewed. Another 6% could not be contacted during phase two of the study.
Leaving out incomplete questionnaires and other discrepancies, completed questionnaires were
available for 3339 women (Table 3). About 14% of the women had more than one pregnancy and
some though a small number, had even four pregnancies (Table 4). Thus, the total number of
pregnancies studied (hereinafter called the ’’index pregnancy”) were 3844 in 3339 women.

Pregnancy passes through three stages - antepartum, intrapartum and postpartum periods.
Accordingly, we have presented the results. Further, we have also presented the findings according
to areas of residence viz: rural Tamil Nadu, Tamil Nadu urban, Pondicherry rural and Pondicherry
urban. Longterm sequelae and morbidities arising out of abortions, have been reported separately.
Selected Characteristics of Study Population

Age Distribution (Table 5)
Majority of the women (76.2%) were between 20 and 30 years of age. The proportion of
teenage pregnancies was 8.7%. There were more teenage pregnancies in the rural areas. The mean
age of the women was 24.7 years (±4.6 years).
The mean age was lower for women from Pondicherry. It was 24.2 years for both rural and
urban Pondicherry as compared to 25.4 years for urban Tamil Nadu and 24.8 years for rural Tamil
Nadu. The differences in the mean ages of women according to the areas of residence was significant
Parity (Table 5)

The mean number of pregnancies was 2.6 (±1.5). In fact, three-fourths of the women
studied had less than 3 pregnancies. Primiparous women formed 25.6% of the study population.
Only 4.5% had more than 6 previous pregnancies. The mean number of pregnancies differed with
the area of residence. Women in the urban areas and those from Pondicherry rural areas had lower
number of pregnancies than women from rural Tamil Nadu. The same difference was seen for the
total number of live births. These differences were statistically significant (p<0.05).
Education of Respondents (Table 6)
About half of the women (48.9%), had attended school. Nearly one fifth had completed
schooling at least for 3 years and another fifth between 4-12 years. A small proportion of women
(1.5%) had more than 12 years of schooling. Very few women (0.3%) had postgraduate education.
Literacy was higher for the urban women in general and for Pondicherry area in particular.

Education of Husbands (Table 6)
Educational status of the spouses was imuch better. Approximately, an equal proportion of
men (22%) had education up-to primary, secondary and high school level.'

Economic Status (Table 6)

Majority (81 %) of the families belonged to the low or middle socio-economic status. Urban
women from Pondicherry were economically better off than those from other areas. About half of
8

the women (49.7%) were working for wages outside their homes. More often women from rural
areas, especially those from Tamil Nadu, worked outside their homes for wages.

Usual Source of Care (Table 7)

We wanted to find out, the usual sources from where women sought health care for illness,
immunisation and gynaecological conditions. This would provide clues to their awareness of health
care facilities available, the distance from their place of residence to the facilities and the mode of
transport generally used to reach the facility. From the responses, it may be seen that more than half
of the women identified either a government facility (27%) or a private facility (27%) as the usual
source of care. These were within half an hour distance according to 58% women, whereas, 36%
said it took more than an hour to reach the facility. About 45% of the women used bicycle,
motorcycle or bus to reach the health facility, 44% walked the distance. In Tamil Nadu, more often
private health facility was identified as the usual source of care (39.6%), especially in the urban areas.
In the urban area of Pondicherry, 41.5% of women mentioned government health facility as the usual
source of care. The travel time to a health facility was lesser in urban areas. The rural areas of
Pondicherry have a good network of health centres. Hence more than 60% of the women said that
it took less than half-hour to reach a health facility. In the rural area of Tamil Nadu, None (or) 0.1 %
of the women mentioned that they went to a traditional or other practitioner for their problems. No
one mentioned TEA as an usual source of care.
Antepartum Care Utilised by Respondents (Table 8)

The utilisation of antepartum service was high. Ninety eight percent of the women had
antepartum check-up. The urban women in both Pondicherry and Tamil Nadu and the rural women
of Pondicherry had higher utilisation of services than women of Tamil Nadu rural area. The first visit
for antepartum care in as many as 44% women was in the first trimester itself. Urban women went
for check-up earlier than their rural counterparts.
Mean Number of Visits (Table 8)

The mean number of antepartum visits ranged from as high as 8 in urban Pondicherry to
about 4 in rural Tamil Nadu.
Place of First Visit (Table 8)
Most often women visited a government health facility, ei±er a primary health centre
(47.3%), or a government hospital (11.3%) for the first antepartum check-up. There was, however,
some variation between Tamil Nadu and Pondicherry in the source of antepartum care. In Tamil
Nadu urban area, private sources such as general practitioners or nursing homes were used (34.7%).
In the case of rural women of Tamil Nadu, the first check-up was at home in 31% of women,
usually by a female health worker from the nearby PHC. In Pondicherry urban area, 77% of the
women went to one of the health centres for the first check-up. This is due to the fact that in the areas
sampled, urban health centres are functioning.

Twenty four percent of the women who had first antepartum check-up at home, subsequently,
went to a health centre or a hospital for further care.
■5

Reason for First Visit (Table 8)

The first antepartum visit was either for confirmation of pregnancy (32% ) or a routine check­
up (31 %). More urban than rural women had their first antepartum contact to confirm pregnancy. In
30% of women, the first antepartum contact was for tetanus toxoid injection. A mere 8% of women
said that they went to a health facility only when they had a specific morbidity and not otherwise.

9

It is encouraging that the utilisation of antepartum services was as high as 98% and on an
average 96% women had received tetanus toxoid. The coverage was maximum for urban women of
Pondicherry (98.2%) and was lowest for the women in urban Tamil Nadu (93.1%).
Reasons for Not Seeking Antepartum Care (Table 9)

Only very few women, (2%) did not receive antepartum care. Often, more than one reason
was given for not seeking antepartum care. The commonest reason was that they did not consider it
necessary to seek care as they were feeling well and perceived no problem. On analysis of their
pregnancy course, it was found only eleven of these women had some antepartum morbidity, mostly
vomiting. Shonage of money or too much work at home were the other reasons frequently given for
not going for a check-up. Only eight women were not aware of the services.

Characteristics of Index Pregnancy
Outcome of Index Pregnancy (Table 10)

About eighty eight percent of 3844 pregnancy resulted in live birth, two percent in still birth
and 6.8% in abortion. The proportion of pregnancies medically terminated was 3.5%. The number
of medical terminations were higher in the urban areas.

Type of Delivery (Table 10)

Ninety percent of deliveries were normal, 5.5% were by Caesarean section, and for the rest
forceps had to be applied or vacuum extraction was done. Caesarean section was highest for women
from Pondicherry urban area (9.9%) and lowest (2.9%) for rural women of Tamil Nadu.
Place of Delivery (Table 10)
Nearly 60% of deliveries were institutional, attended by skilled persons like doctor or nurse
and about 40% were home deliveries. As could be expected, more home deliveries took place in the
rural areas; it was particularly higher in rural Tamil Nadu (64%) than in rural Pondicherry (28%).
It is interesting that, even in rural areas of Pondicherry, women preferred to go to hospital because
of easy accessibility of health services and availability of transport. A matter of concern was the very
poor utilisation of primary health centres (PHC) for deliveries. In Tamil Nadu, only 2% deliveries
took place in PHC’s compared to 7% in Pondicherry. Some 20 women delivered on their way to
hospital.

The home deliveries were attended mainly by trained birth attendants in Pondicherry rural
areas, while in Tamil Nadu rural areas, trained birth attendants conducted 33% and untrained
conducted 23% of the deliveries.
Actual Place of Delivery by Intended Place of Delivery (Table 11)

Government hospitals, their own homes or private hospitals were the preferred places of
delivery in that order. In the rural areas, that PHC’s were not preferred for delivery is shown by
the fact that only about 3% of total women mentioned them as the intended place for delivery. Only
1 % in Tamil Nadu and 10% in Pondicherry had planned to have their deliveries in a PHC.

The figures in the diagonal of the table shows the percentage of women who actually
delivered at the place they had intended to. It ranged from 50% for health centres to 82% for those
wanting a home delivery. The numbers below the diagonal indicate women who had to go for a
higher level of care than they had intended to, possibly, because of development of complications or
referral. The numbers above the diagonal indicate, the women who were unable to reach the desired
place of delivery. For example some women said that due to non-availability of transport, or
10

economic reasons, or insufficient time due to rapid progress of labour, they had to have their delivery
at home.

Prevalence of Maternal Morbidities
Morbidities can occur in any of the three stages of pregnancies, viz, antepartum, intrapartum
and postpartum. We have described the findings separately. The reported symptoms or symptom
complexes were classified into potentially life threatening, serious and other morbidities, after
extended discussion with consultants from FHI, investigators from the other three country projects
and local obstetricians (See box below).

Classification of potentially life threatening and serious morbidities of the index
pregnancy
Postpartum
Intrapartum
Antepartum
Haemorrhage
Fits/convulsions
Malaria

Edema (hands & face)
Hypertension
Fever > 3 days
Severe vomiting
Jaundice
Pulmonary TB
Rheumatic heart disease
Diabetes mellitus

Life-threatening
Haemorrhage
Fits/convulsions
Ruptured uterus
Malaria
Serious
Labour > 18 hours
Vaginal or cervical tear
Caesarian Section

Haemorrhage
Malaria

Fever >3 days
Foul discharge
Shock/loss of
consciousness

Number of Morbidities (Table 12)
One or more morbidity was reported in 56% of the pregnancies. A further breakdown
showed that about 39% morbidities were reported during the antepartum period, 18.7% during
intrapartum and 24.6% during postpartum period. In 25% of pregnancies, there was more than one
morbidity. Multiple morbidities were reported more often during antepartum period (9% of
pregnancies) than intrapartum (4.7%) and postpartum (1.5%) periods.

Potentially life threatening complications were reported in 4.7% of the women during the
three phases of pregnancy. In the antepartum period, life threatening morbidity occurred in 1.3% of
the pregnancies, lower than that during the intrapartum (2%) and postpartum periods (2%). So far
as serious morbidities were concerned, they were reported in 40.7% of the pregnancies during all the
3 phases combined, whereas, they were reported in 24%, 13% and 16% of pregnancies during the
ante, intra and postpartum periods respectively.
Antepartum morbidities (Table 13, 14)
The commonest life threatening morbidity was, antepartum bleeding (6 per 1000 pregnancies),
followed by malaria (4 per 1000 pregnancies) and fits and convulsions (2.6/1000 pregnancies). A
quarter of women had serious morbidities. Severe vomiting (181/1000), severe edema of hands and
feet (51/1000), fever of more than 3 days duration (39.3/1000) and pregnancy induced hypertension
(10.7/1000) were the conditions reported frequently. The other pre existing conditions which would

11

potentially increase the risk, and therefore considered as serious, were heart disease (3.6/1000),
pulmonary tuberculosis (2.1/1000), jaundice (1.3/1000) and diabetes mellitus (1.3/1000).
In addition to the above, women were able to report several more conditions. We classified
them as infective conditions and others (Table 14). Vaginal discharge and diarrhoea among infective
conditions and giddiness, irritative urinary symptoms, blurred vision and anaemia were the other
conditions commonly reported.

Perceptions of Women and Treatment Sought (Table 13, 14)

There are a number of non-physiological factors which influence the decision to seek medical
aid. Among them is how a person perceives a problem is an important one (Zola et al., 1966).
Recognition of a problem, its perceived causes and feared outcome rather than a sophisticated medical
diagnosis which generally determine how a disease will be managed by a patient. Thus demand for
an utilisation of services will depend upon subjective perceptions. We asked women to indicate
whether they considered the reported morbidity as serious or not, and whether they sought care for
those conditions. The three potentially life threatening conditions (antepartum bleeding, malaria and
fits) had varied response. All women considered fits as a serious condition and most women felt
malaria also as serious, all except one woman, sought care for these conditions. Only 78%
considered antepartum haemorrhage as serious, despite that, over 87% sought care. One of the
reasons for not considering bleeding as a serious condition was because they felt bleeding was not
‘heavy’ or ‘it stopped without recourse to any treatment’. In a study on beliefs regarding delivery
and postpartum maternal morbidity in rural Bangladesh, ideas of how much bleeding was ‘too much’
were vague and few women had an appreciation of the danger of the condition. Biomedically,
bleeding at anytime during pregnancy is considered serious. It may be necessary to educate women
not to take this condition lightly.
Jaundice, diabetes mellitus, heart disease and pulmonary tuberculosis were considered as
serious by over three fourths of women and treatment was sought by nearly all of them. Pregnancy
induced hypertension was considered serious only by 71% of women, however, treatment for the
condition was taken by 93% of women. It is not clear why about 29% of women did not consider this
condition as serious. Was it because pregnancy induced hypertension is often asymptomatic (Stewart
et al., 1966) or was it that they were not informed about it when blood pressure was measured?
While, for the former reason, we need to reinforce this during antepartum education, for the latter
this matter has to be brought to the notice of doctors and health functionaries and urge them to inform
their clients about the blood pressure status. The brighter aspect is that women seemed to be aware
of the condition ‘high blood pressure’, and called it as ” boiling of blood ” (Raktha Kodippu) but
could not ascribe any one particular symptom.
For the conditions categorised as ‘others’, the care seeking behaviour varied, ranging from
two thirds to all the women seeking care. The main reason for seeking care depended on the
discomfort caused by the symptoms.

Place of Treatment (Table 15)
For the treatment of antepartum morbidities, private hospitals or clinics were more popular
than government sources. Among the government sources, health centres were preferred more than
the government hospitals except for the treatment of fits/convulsions, pregnancy induced hypertension,
heart disease and pulmonary tuberculosis.
Intrapartum Morbidity (Table 16)
Intrapartum period is the shortest phase compared to the ante and postpartum periods. This
phase is unpredictable too, for any time during labour complications may develop. Seventeen percent
12

of women reported morbidity during this phase, 2% were potentially life threatening, about 9% were
serious morbidities and the rest were classified as others.

Among the life threatening conditions, excessive bleeding was the commonest (10/1000
pregnancies), followed by loss of consciousness (9 per 1000) and fits / convulsions (3 per 1000). The
three common serious morbidities reported were, early rupture of membranes (34/1000), abnormal
position of foetus (28/1000) and labour more than 18 hours (14 per 1000). Caesarean section was
also grouped in this category and accounted for nearly 4% of the serious morbidities. In the category
of other morbidities which were not so serious, genital tear was the commonest (64 per 1000).
Place of Delivery (Table 16)
We looked at the place of delivery with the belief that life threatening conditions are more
likely to be reported from women delivering at home. Infact, life threatening morbidities were slightly
higher (23 per 1000) among home deliveries compared to hospital deliveries (20 per 1000). Fewer
serious morbidities were, however, reported from home deliveries (20 per 1000) compared to hospital
deliveries (132 per 1000). It may be because, women with conditions like abnormal foetal position,
twins or other high risk conditions went or were referred to hospitals for delivery. What is somewhat
disturbing is that a large number of women delivering in hospitals had genital tears (87 per 1000)
compared to those delivering at home (30 per 1000). This reflects the quality of care. It requires
improvement. Retained placenta (4 per 1000), inversion of uterus and cord round the neck of foetus
were reported by two women each delivering at home.
Various interventions were reported by women delivering at hospitals as well at homes. In
some instances, particularly for serious morbidities, more than one intervention was reported, for
instance, drugs and episiotomy or drugs and application of forceps. The specific interventions were
further analysed.

Interventions in Hospital Deliveries (Table 17)

(a) Caesarean Section: About 6% of women underwent Caesarean section. The common
indications were abnormal position of foetus, early rupture of membranes, labour greater than 18
hours, and excessive bleeding. It was highest for women from urban areas of Pondicherry (10%) and
lowest for women belonging to rural Tamil Nadu (3%).

(b) Forceps/Vacuum Extraction: Delivery was conducted in 4.2% of deliveries by application
of forceps or vacuum extraction of foetus when labour was prolonged for more than 18 hours (19%
of pregnancies with morbidities) or for abnormal position of foetus (15.3%) or when there was early
rupture of membranes (13%). Forceps application was done for delivery of twins (12%) and when
there was meconium staining of amniotic fluid indicating prolonged labour or foetal distress.
(c) Episiotomy: Episiotomy was reported frequently among women with early rupture of
membranes (25%), or when there was abnormal position of foetus (12%) and in case of prolonged
labour (19%).

(d) Repair of Tear. The genital tears were repaired in the hospitals.

(e) Drugs: Women recalled drugs being given during labour. Wherever prescriptions or
discharge summary sheets were available names of the drugs or the fact that drugs were given were
noted down by interviewers. From these it is surmised that the drugs given were mainly for local
anaesthesia before repairing the tears (115/166), oxytocics and antibiotics.

13

Number of morbidities per woman (all morbidities) (Table 27)

In 42% of the pregnancies, there was more than one morbidity during the three stages of
pregnancy. Multiple morbidities were reported more during the antepartum period (25%), than in
intrapartum (13%) and postpartum (11%) periods.
Long Term Morbidities (Table 28)

In about eight percent of pregnancies long term morbidity was reported. One percent of
women had more than one condition.

Ratios of morbidity to mortality (Table 29)
For every maternal death there were 478 morbidities. There were 328 serious or life
threatening morbidities per maternal death. Both these figures are clearjy in excess of the previous
estimates of 16.5 morbidities per maternal death found by Datta et al (1980). In their study the
sample size was only 258 women and was restricted to only serious and life threatening morbidities.

Morbidities following Abortions (Table 30)

Ten percent of the 3844 pregnancies ended in abortion. Of these, 66% were spontaneous and
the rest were induced. Table shows the complications that occurred and the treatment seeking
behaviour.
Ninety three percent of women who had spontaneous abortion and 68% of women who had
induced abortion reported some morbidity following abortion. Among the women who had
spontaneous abortion, the morbidities reported were excessive bleeding 81 %, pain abdomen 79% and
fever 29%.

Among those who had induced abortion, pain abdomen was ireported* \
\half of the
by nearly
women, excessive bleeding by 42% and fever by 17%. In general, the proportion of morbidities
reported among women with induced abortion was lower than those for spontaneous abortion.
Treatment seeking behaviour (Table 30)

Nearly two thirds of women in both the groups sought treatment for complications arising out
of abortion. The most preferred place for both the groups was private hospitals or clinics. Only
about a quarter of women went to government hospitals for treatment. In slightly less than a third
of women hospitalisation was necessary, the rest were treated as out patients. The average stay in
the hospital was three days.
Reasons for last induced abortion (Table 31)

We asked the 134 women, the reasons for inducing abortion. Nearly half of them said that
they did not want any more children, about 21% said that they could not afford to have more
children, and 13% felt pregnancy would interfere with their employment. Only 8% of women gave
health reasons for induction of abortion.

Although most of these women were not desirous of the pregnancy, only 3 women (2.3%)
were using contraceptives and pregnancy occurred in spite of that.

Method of Induction (Table 31)
The common method of induction was by dilation and curettage (70%), followed by injection
of hormones (19%) and pills (7%). Majority of women 89% had gone to a doctor for induction.
Only 1.5% of women had abortion induced by a traditional birth attendant.

16

Age, Parity, Residence and Socio-economic status and morbidities

What role do factors like age, parity, residence have with the risk of developing morbidity?
To answer this question, we examined the life threatening and serious morbidities for the above
factors.

Socio-economic status (Table 32)

Risk of morbidity in relation to socio-economic status showed different trends. There was
no relationship of socio economic status with antepartum and overall morbidity rates. In the
intrapartum period an increasing trend was observed with increasing socio economic status. This is
probably due to the higher rates of caesarean section among the higher income groups. In the
postpartum period the rates for morbidities decreased with increasing socio economic status.

Age (Table 33)
Women below the age of 19 years and those over 35 years were at a higher risk of developing
life threatening or serious complication during the antepartum and postpartum periods. For instance,
26.4% and 16.8% of women below 19 years reported life threatening and serious morbidity both in
antepartum and postpartum periods respectively. Whereas, in women above 35 years these figures
were 23.8% in antepartum and 13.4% in post partum periods.

Parity (Table 34)
Primiparous women were at the highest risk of developing life threatening or serious
morbidity. For example, nearly half of the women reported a morbidity compared to 39% women
of para two and above. This risk was highest during the intrapartum period, 17% as compared to
11 % of women of 2-4 parity and 11 % women of para 5 and above.
Residence (Table 35)

The four areas of residence showed significant differences in morbidity rates for life
threatening and serious morbidities. Women in Tamil Nadu area in general irrespective of urban or
rural residence had higher rate of morbidity than the women from Pondicherry. During the
antepartum period more rural women from Tamil Nadu (43%) had morbidity than their urban
counterparts (34%). In Pondicherry area, the proportion reporting morbidity was nearly similar
among urban and rural women (37% and 35% respectively). In the intrapartum period, urban women
from both Tamil Nadu and Pondicherry (24% and 23% respectively) reporting higher morbidity than
rural women of Tamil Nadu (16%) and Pondicherry (18%). During the postpartum period more rural
women from Tamil Nadu reported life threatening and serious morbidities (23%) than urban women
(15%). Both urban and rural women from Pondicherry area reported lower morbidities than Tamil
Nadu women, the least belonging to urban Pondicherry (11 %).
It may be recalled that more rural women delivered at home attended by traditional birth
attendants and the infections were also reported higher among the rural home deliveries. This
explains the higher postpartum morbidity among rural women. The higher intrapartum morbidity may
be attributable to higher rates of caesarean sections among the urban women.
Prevalence of morbidities and intrapartum care
A comparison between our study and the studies conducted in Bangladesh, Egypt and
Indonesia on the incidence of selected post partum complications and the skilled intra partum care
revealed that when a higher proportion of women received intrapartum care from medically trained
persons they were less likely to develop certain complications. In other words, occurrence of
conditions like prolonged labour, postpartum bleeding and fits or convulsions can be influenced by
the quality of intrapartum care (See Table A)
17

Table A. Association between the frequency of selected postpartum conditions and the
proportion of women receiving skilled intrapartum care
Country

Percent
delivered
in hospital

Percent
delivered by
doctor or
midwife

Number
with
complica­
tion

Number
with no
complica­
tion

Odds
Ratio

95%
confidence
limits

PROLONGED LABOUR (n = 3023)

India

59

62.2

47

3402

1.00

Egypt

12.2

47.5

156

4154

2.77

1.9 -4.0

Indonesia

2.3

7.7

96

1795

3.94

2.6-5.4

Bangladesh

1.0

11.1

419

5973

5.17

3.7 - 7.2

X2 for trend

117.1

126.1

< 0.000

< 0.000

P-value

POSTPARTUM BLEEDING (n=3009)

India

59

61.2

63

3386

1.00

Egypt

12.2

47.5

153

4157

2.01

1.4-2.6

Indonesia

2.3

7.7

215

1676

7.02

5.1 -9.6

Bangladesh

1.0

11.1

1096

5296

11.32

8.5 - 15.1

X2 for trend

450.6

727.5

< 0.000

< 0.000

P-value

FITS OR CONVULSIONS (n=3058)

India

59

61.2

13

3436

1.00

Egypt

12.2

47.5

6

4304

0.71

0.2 -2.5

Indonesia

2.3

7.7

37

1854

10.17

4.1 -26.8

Bangladesh

1.0

11.1

224

6168

18.51

8.0 - 46.1

X2 for trend

99.7

186.8

< 0.000

< 0.000

P-value

Source: Adapted form Fortney and Smith, 1996.

There is only one exception to this order: in the case offits and convulsions, Egyptian women have a
lower rate than Indian women, but the hypothesized order is otherwise maintained.

18

The table shows the association between the frequency of three conditions : prolonged labour,
postpartum haemorrhage and fits or convulsions among women with skilled intrapartum care, test for
trend and odds ratio. The test for trend was significant in each case. The odds ratio indicated that
compared with the women in our study (60% had hospital deliveries), women in each of the other
three countries had a higher risk of developing the three complications mentioned above. For instance,
Egyptian women had nearly three (2.77) times the risk of a labour longer than 18 hours, Indonesian
women nearly four (3.94) times, and Bangladeshi women more than (5.17) times the risk. Percentage
of hospital deliveries were 12%, 2% and 1% in these countries respectively.

In case of excessive bleeding and fits and convulsions too the trend was similar and the risk
of complication was higher.

A respondent being interviewed

19

Discussion

Retrospective community based, self reporting studies are said to either underestimate or
overestimate the morbidities. The underestimation is said to be because of (a) recall bias especially when
the interval between episode and interview is long, (b) different morbidities are remembered with different
accuracy, (c) overlooking of mild or asymptomatic conditions and (d) many conditions like white
discharge, backache, nausea, vomiting may be considered as woman’s lot or perceived as normal and not
reported. Over estimation is said to occur because of reporting of so called trivial conditions. Another
problem is validating self reported information.

How valid are self reported information? The accuracy of self reports of the occurrence of disease
and disease symptoms has been questioned. Murray and Chen (1992) referred to studies that showed a
large discrepancy between self reported morbidity and clinical examination. Other studies present a more
optimistic view of the closeness of self reported to observed disease (Colditz et al, 1987; Bush et al. 1989;
Halabi et al. 1992; Midthjell et al. 1992). A recent study from the Philippines observed that mothers were
able to retrospectively report recent illnesses with sufficient accuracy for interview based diagnosis (Kalter
et al 1991). In another study, comparing women’s report with medical diagnosis of reproductive
morbidity conditions in rural Egypt, it was found that women, reports were of great value in providing
information on their perspectives and revealed the magnitude of the feeling of ill-health in the community.
The authors concluded that such information was critical for health services on the relevance of symptoms
and the meanings women attach to them (Zurayk, H. et al. 1995).
We were aware of these issues. We therefore, limited the recall period to two years. This
decision was taken after conducting many depth interviews and group discussions of women of different
parity, background (urban, rural) and birth intervals. In addition, even conditions such as nausea,
vomiting or dizziness when reported were recorded. The questionnaire used had many open ended
questions and respondents were encouraged to talk freely. All conditions mentioned by them were noted
without any judgement. To give broadest possible picture of maternal morbidity in developing countries,
Liskin (1992) recommended that community surveys should involve multiple countries and geographic
regions. This survey was conducted not only in two states (Pondicherry and Tamil Nadu) and within
these states in urban and rural areas in a proportion of 35:65 respectively, but also in Bangladesh, Egypt
and Indonesia using a common frame of questions.
Finally, a subset of women reporting morbidity in our study were clinically examined by a
gynaecologist. There was agreement between reported symptoms / conditions like white discharge,
prolapse uterus, dyspareunia and the gynaecologist’s observations. In case of only incontinence of urine
and piles there was less association. Some women categorised as having piles had prolapse of uterus.

What are the policy and programme issues that emerge from the study? The first is that the
prevalence of morbidity was far higher than anticipated. Nearly 60% of women experienced at least one
morbidity. Some women experienced more than one. About 1.8% of women reported five or more *
morbidities. Not all of course were, serious or life threatening. A life threatening morbidity was reported
by 5% of women during either pregnancy, labour or puerperium. But quite a large number of women
(41%) reported experiencing serious morbidities. Our finding is not an isolated one, for even in
Bangladesh and Egypt, the prevalence rates were also high. The percentage of women with any morbidity
during pregnancy and puerperium was 79.9% among Bangladeshi women and 66.9% among Egyptian
women, and nearer home in the neighbouring Karnataka state (Bhatia and Cleland, 1996), it was 41%.
Further, the number of morbidities in index pregnancy per maternal death was 450 regardless of timing
or severity. And looking from another angle, i.e., maternal mortality ratio to number of women with
morbidities, 238 women experienced a morbidity for every maternal death. This magnitude of the
maternal morbidity needs recognition by programme planners and health professionals so as to plan future
strategies for reducing maternal morbidity and mortality in particular and women’s health in general.

20

*

Should conditions like nausea, vomiting, backache be considered as morbidity? Would they not
over estimate the morbidity? Medical professionals may consider these conditions to be trivial or minor,
and may not attach importance. Perceptions of what constitutes morbidity vary. Liskin (1992) observed
that the so-called minor complaints of pregnancy and child birth are only rarely addressed even though
these conditions may significantly impair women’s well being and affect their ability to work. If a
condition perceived by a woman prevents the performance of normal activity or warrants her to seek
treatment, it is certainly a departure from normality and from sufferer’s perspective, is rightly regarded
as an illness (Bhatia and Cleland, 1996). We agree with these views. It may be recalled that 65% of those
experiencing giddiness, 75% having severe nausea / vomiting and many other conditions during
antepartum period varying from body pain to burning during micturition and body pain to diarrhoea
during postpartum period sought consultation and treatment. This not only affected their ability to work,
for 44% women from rural Tamil Nadu worked outside home for wages, but also caused economic
hardship as they had to spend money for transportation, consultation and for purchase of medicines. Bio
medical perspectives do not include the social and psychological dimensions of ill-health. (Helman, 1984).

. !

Conventionally, morbidities are classified as obstetric (direct) or non-obstetric (indirect). However,
both obstetric and non-obstetric morbidities range from mild to potentially life threatening. Such a
classification is important for planning intervention programmes (Graham and Campbell, 1990). The three
periods of pregnancy have different time frames and the frequency of morbidity in different periods may
vary. During the antepartum period, life threatening complication were few (1.2%) and serious
morbidities were 24% during the antepartum period. It was slightly higher among women below 19 years
of age, primiparous women and those belonging to Tamil Nadu urban and rural areas compared to other
age groups, parity and Pondicherry area. The perceptions of women for the three life threatening
conditions varied. For instance, fits and malaria was considered as serious and treatment was sought by
all of them. However, only 78% of women with antepartum bleeding considered it as serious, although,
87% sought care. Pregnancy induced hypertension was considered serious only by 70% of women despite
the fact that 93% of them took treatment.
During the intrapartum period, life threatening and serious morbidities were 2% and 9%
respectively. Life threatening morbidities were slightly more in home deliveries (2.3% of deliveries) than
hospital deliveries (2% of deliveries). Whereas, serious morbidities were reported more among women
delivering at hospitals than among women delivering at home. This was on expected lines because of about
60 percent of deliveries were in hospitals and women with high risk conditions were either referred or
went on their own to hospitals for delivery. In fact, in Pondicherry region more than 90% of urban
women and nearly 80% of rural women delivered at hospitals.

In spite of high proportion of hospital deliveries the quality of care during labour was wanting,
for 8% of women had genital tears. Among home deliveries life threatening or serious conditions were
not promptly referred by traditional birth attendants. Postpartum morbidities due to infections were
frequent among home deliveries, again reflecting the poor hygienic conditions at the time of delivery.
Rural women did not choose to have deliveries in primary health centres and in fact, only 3%
deliveries were conducted there. In theory women could go to primary health centres for labour but they
did not do so. Perhaps lack of facilities for even normal deliveries and staff not being residential in
majority of PHCs might be the reasons for this situation. This matter needs serious attention.

$

So far as health care seeking behaviour is concerned, the utilisation of antepartum services for
pregnancy care and hospitals for deliveries were high. For instance, it was 98% for the former and 60%
for the latter. The levels of consultation or treatment for antepartum care and obstetric problems were
found to be very high. Similarly hospital deliveries were also high (overall 60%). A number of women
went to private hospitals or to private medical practitioners and they were willing to pay for their services
rather than use government facilities, particularly primary health centres, for delivery. But our depth
21

interviews gave many clues to the difficulties women had to face. Apart from her own perception of the
seriousness of her morbidity, the perceptions of other family members particularly of the husband and
other elders mattered. If she was willing or needed to go to a health care facility, she was dependent on
some one to accompany her. Another major difficulty was in respect of transportation - availability and
cost. Many times more than one visit had to be made (See Fig. 3&4).
Mrs. R., 24 years

A.

MOTHER-IN-LMV't
HOUSE
6 WEEKS PREeNAHT
MOT RAISED URINE
FOR ONE WW
DESIRES TO
SO TO DOCTOR

HUSBAND
ACCOMPANIES

RRIWE HOSPITAL
TINDIMKNAM
OITS RELIEF

____ o

fi

-e-

HUSBAND I

PERUNDANQAL
CLINIC

■e-

ACCOMPANIES

16 at*. \
TRAVEL
'
MONEY SPENT
Re. 60
/

/

MOTHER'S HOUSE

Mth week
MOTHER'S HOUSE
PONDICHERRY
DIZZINESS
URINARY PROBLEM
A FEVER

o

SISTER
■ ACCOMPANIES
A

MATERNITY HOSPITAL
PONDICHERRY

CAESARIAN SECTION
STAYED • DAYS
DUE TO PUS
IN THE WOUND

SSHi week

I

/Ut keer\
/ TRWEL \

SSth *««k

MONEY
\ SPENT /
XReJOZ

Jk

TOTAL EXPENDITURE
Rs. 2000/-

BROTHER
ACCOMPANIES

Figure 3. Flow chart indicating multiple visits, accompanying persons and modes of transport.

s

«etk wk or PMONANOY
RMNYS WMIL1 BATMIMe
M MOTWCR-HM.JW •IVEB,a “*
NORM TROT MORT
O NO RBUIP.
ABDOMINAL RAIN
NO FURTMRR ACTION
aOTMie-HHlW AWUTt
BON TO RRAON NOMR

BLtKMNR OONTmUEB
SON MA0M8S MOM! 0 FN

o

Mrs. V, ir jrra. Oe4l»e« «•
VllllaMir. MerrleO AeeI ereaeeaey erter ■ ■•■thi
HeakaM - 0»vL Bee Orlver

BOM
oroeoee ro
BUT* NRN
ro pocroN

4.

RIACHie
PRIWTI OLINIO
PONOIOHIRRY
t br»- LFTiR

OYCUit

FITOHIB
kUTORIKeHI

AOVIOID TO
so TO
MATERNITY
HOBFIIAL.
FONDIOHaRRY

BUrON TO C
00
MHIMT
TO OOVtfiMi
UArtHMITT
HOWU.

Tiirme ooni
KDYteeO AOMI6MOM
riOM
IMONOONGY OrORFTI

peauMae blood
WV6SAN0 DONWie
•cooo

UNMRWVNT OP6IUVION

MTSNOIVI TRIXTMINT
FOR • <Mft KT I0U
• «W AT ora01 AL YARD
FOR TRUrURNT - 10 CMYO

Figure 4. Flow chart of decision making process

22

Seventy percent of women who had a morbidity had to spend about Rs. 100/- for transport and
purchase of medicines. Given the poor socio-economic status, many a time money had to be raised at short
notice by borrowing at high rates of interest ranging from 3% to 10% per month by pledging their
valuables (See Fig. 5).

Mrs. A.,20 Years, Illiterate, from a Pondicherry village where PHC is situated. First
Pregnancy, had swelling | for which she was treated in the PHC.

a

" I started passing fluid 10 days prior to delivery date, went to health centre, they referred
to hospital, I went at 11 pm. I was operated in the hospital. ... baby had fits, kept in the
glass room. They gave 3 injections daily to baby. We bought the injections and gave. I
pledged my jewels ear stud-Rs. 450, Nose stud - Rs.200 and silk saree - Rs.250 through
my brother. Interest was © 3 paise/Rupee per month. Spent about Rs. 1000/- "
Note:

Non availability of medicines in hospital, need to raise money at short notice. High rate of interest

Fig. 5: Excerpts of a depth interview

Lack of suitable mode of transport, in the absence of public transport, dependence on other
persons, especially men, the need for a companion to accompany her to the health facility for moral
support, non availability of medicines in government hospitals, cost of care - these were only some of her
numerous problems. When analysed thus it becomes more than merely a ’transportation* problem.

3

23

Interventions

The following interventions are suggested with the hope that they would reduce maternal
morbidities as well as improve women’ health.
Pregnancy services
Safe pregnancy services should be designed to ensure timely detection, management, and
referral of complications during pregnancy, labour and delivery. Because of their impact on health
and child as well as the mother, safe pregnancy services are highly cost effective. According to a
World Bank report, providing prenatal, delivery and postpartum services costs less than $ 2000 per
death averted, or between $ 30 and $ 110 DALY saved (World Bank, 1993b). Priority should be
given to improving hygienic practices, providing iron and folate supplementation, and strengthening
linkages and referral services for obstetric complications.
Prenatal Care

Regular prenatal care is needed to help detect and manage pregnancy related complications
and to educate women about danger signs, potential complications (eg. pregnancy induced
hypertension) and where to seek care.
Because most pregnancy related complications cannot be anticipated, all women need access
to appropriate care should complications develop.
Safe Delivery
Delivery care should include safe management of routine deliveries, safe-birth practices by
traditional birth attendants, communication and transport to ensure timely referral and management
of emergency complications and essential obstetric services at first referral level.

Health centres should be able to ensure hygienic normal delivery particularly by female health
workers. Sepsis at delivery can largely be prevented by minimising vaginal examinations and ensuring
clean delivery practices.
Mostly life threatening complications occur during labour and delivery, and because most of
these cannot be predicted, every woman needs access to emergency obstetric case.

In cases of haemorrhage, prolonged or obstructed labour and other obstetrical emergencies, *
the most important element in a woman’s treatment may be transportation. When distance is a factor,
first aid at the community or health centre level may be necessary to save a woman’s life by
stabilising her condition until she reaches a hospital.
Advance planning for emergencies is therefore key to reducing maternal morbidity and
mortality. Efforts must also be made to improve the existing services particularly the primary health
centres and community health centres.
Postpartum Care

Post partum care should include early detection and management of infection and
haemorrhage. Even among women who have delivered in a hospital, postpartum follow-up is
important because complications may arise after leaving inpatient care which is now a days shortened
to 24 - 48 hours. Educating women, their families, birth attendants and health workers to recognise
early signs of infection and seek care for them, for example, may reduce the suffering.

24

Postpartum care should respond to woman’s needs and preferences to ensure utilisation and
effectiveness.
Education of Women and Family

Public education programmes and counselling to teach women how to recognise the signs of
complications and when and where to seek help are needed. Teaching women and family members
to recognise danger signs of pregnancy and to seek prompt medical attention can greatly reduce the
incidence of maternal morbidity. In Zaria, Nigeria, a radio programme stressing the dangers of a
labour lasting more than 24 hours is credited with a significant decrease in the incidence of obstetric
fistulae (Harrison et. al. 1985).
Increased male involvement and support

In many cultures including India, men make the decisions including health related concerns
such as family size, contraception and use of health care. Very few women would decide by
themselves to seek care in the events of pregnancy. For most, decisions rested with their husbands
or elders. Education programmes and services directed to men are needed for the existing antepartum
services are too female oriented.

Health and other agencies need to make concerted effort to make men aware of women’s
health problems and encourage them to take a more active role in preventing unwanted pregnancy,
family planning, increased spousal communication and support for their partner’s pregnancy care.
Expansion of Essential Services

Increased attention needs to be given to the quality of care. To improve quality of care,
maternal morbidity and mortality audits should be introduced and efforts should be intensified to
coordinate supervision and backup from the hospital to the community level. Services need to be
decentralised, and women will need to be redirected to health centres for routine care and normal
delivery because the referral hospitals have become overwhelmed by the demand.
Misplaced emphasis on number of antepartum visits

Appropriate prenatal care with backup for managing obstetric complications is essential for
maternal health. Many countries including India, have over emphasised the number of prenatal visits,
rather than quality of care provided. Encouraging frequent visits strains the resources of both the
pregnant woman (who incurs travel and time costs) and the health system. Properly conducted good
quality care should be provided through as few visits as possible.
Governments can influence the coverage and quality of health service through attention to the
following areas : access to services; delivery strategies; infrastructure; quality of care; number and
distribution of health care providers; and responsibilities of health workers. NGOs can influence
through education and empowerment of women. Assertiveness and empowerment training for women
need to be organised to bring about a positive self image in women and alter their health seeking
behaviour. Another major effort that is required is to enhance political will by educating the leaders,
professionals and administrators to recognise maternal health nay womens’ health as a major public
health concern.

4

25

References
Abdullah, S.A, M.F. Fathalla, A.M. Abdel Aleem, H.T. Salem, and M.Y. Aly. 1985. " Maternal
Mortality in Upper Egypt ". WHO Interregional Meeting on prevention of Maternal Mortality, Geneva,
11-15 November.

Belcher, D.W., A.K. Neumann, F.K. Wurapa, and I.M. Lourie. 1976. " Comparison of morbidity
interviews with a health examination survey in rural Africa ". American Journal of Tropical Medicine
and Hygiene 25, 5:751-758.
Bhatia, J.C. 1985. "Maternal mortality in Anantapur District, India: Preliminary findings of a study".
WHO Interregional Meeting on prevention of Maternal Mortality, Genev^, 11-15 November.

Bhatia, J.C, J. Cleland. 1995. ’’Self reported symptoms of Gynaecological morbidity and their treatment
in South India". Studies in Family Planning 26,4:203-216.
Bhatia J.C, J Cleland. 1996. " Obstetric morbidity in South India: Results from a community survey".
SocSciMed 43:1507-16.
Bush, T.L., S.R. Miller, A.L. Golden, and W.E. Hale. 1989 Nov. " Self-report and medical record
report agreement of selected medical conditions in the elderly ". American Journal of Public Health.
79,11:1554-6.

Colditz, A. Graham, Meir J. Stampfer, Walter C. Willett, B. William, Stason, Bernard Rosner, H.
Charles, Hennekens, and Frank E. Speizer. 1987. " Reproducibility and validity of self-reported
menopausal status in a prospective cohort study ". American Journal of Epidemiology 126,2:319-325.

Datta K.K, R.S. Sharma, P.M.A. Razack, T.K. Ghosh and R.R. Arora. 1980. "Morbidity pattern
amongst rural pregnant women in Al war, Rajasthan-A cohort study ". Health and Population:
Perspectives and Issues 3,4:282-292.

Fortney A. Judith, I. Susanti, S. Gadalla, S. Saleh, P.J. Feldblum, M. Potts. 1988. " Maternal mortality
in Indonesia and Egypt ". International Journal of Obstetrics and Gynaecology. 26:21.
Fortney A. Judith, and Smith B. Jason Eds. 1996. The Base of the Iceberg: " Prevalence and perception
of Maternal Morbidity in Four developing countries. " Family Health International, Research Triangle
Park.

Government of India. Health Information of India (1991), Central Bureau of Health Intelligence,
Directorate General of Health Services, New Delhi.
Graham, W.J. and O.M.R. Campbell. 1990. Measuring maternal health: defining the issues Maternal and
Child Epidemiology Unit Publication. London: London School of Hygiene and Tropical Medicine.
Halabi S, H. Zurayk, R. Awaida, May Darwish and B. Saab. 1992 Jun. " Reliability and validity of self
and proxy reporting of morbidity data: A case study from Beirut, Lebanon." International Journal of
Epidemiology. 21,3:607-12.

Harrison, K.A., A.F. Fleming, N.D. Briggs, and C.E. Rossiter. 1985. Growth during pregnancy in
Nigerian Teenage Primi- gravdae, British Journal of Obstetrics and Gynaecology 5 (Supplement): 32-39.
Helman ,C 1984 p-68, Culture, Health & illness, Wright, Bristol

26

Jacabson Nora. 1993. Pregnancy in Cagayan de Oro, Philipines. A qualitative study in Conjunction with
the safe motherhood survey. " Report submitted to Mother care project". Arlington VA. John Snow Inc.
and Cal vertan MD. Macro International Inc.

Kalter, Henry D., Ronald H. Gray, Robert E. Black, and Socorro A. Gultiano. 1991. "Validation of
the diagnosis of childhood morbidity using maternal health interviews". International Journal of
5 Epidemiology 20,1:193-198.

Khan, A.R., F.A. Jahan, S.F. Begum. 1986. Maternal mortality in rural Bangladesh: The J amal pur
district. Studies in Family Planning. 17:7.

Koenig M.A., V. Faveau, A.I. Chowdhury, J. Chakraborty, M.A. Khan. 1988. Maternal mortality in
Matlab, Bangladesh: 1976-85. Studies in Family Planning. 19:69.
Koblinsky, Marjorie, Oona M.R. Campbell, and Siobhan D. Harlow, 1993. "Mother and more: A
broader perspective on women’s health". In The Health of Women: A Global Perspective. Eds. Marjorie
Koblinsky, J. Timyan and J. Gay. Boulder, CO: Westview Press. Pp. 33-62.

Krueger, D.E. 1957. "Measurement of prevalence of chronic disease by household interviews and clinical
evaluations". American Journal of Public Health 47:953-960.
Liskin LS. 1992 Feb. Maternal morbidity in developing countries: a review and comments. International
Journal of Gynaecology & Obstetrics. 37,2:77-87.
Maine, Deborah. 1991. Safe Motherhood Programs: Options and Issues.
Population and Family Health, Columbia University.

New York:

Centre for

Midthjell K. J. Holmen, A. Bjomdal, and Per G. Lund-Larsen. 1992 Oct. "Is questionnaire information
valid in the study of a chronic disease such as diabetes? The Nord-Trondelag diabetes study." Journal
of Epidemiology & Community Health. 46:537-42.
Murray, Christopher J.L., and Lincoln C. Chen. 1992. "Understanding morbidity change". Population
and Development Review 18,3: 481-503.

Stewart, M. Kathryn and Mario Festin. 1995. "Validation study of women’s reporting and recall of major
obstetric complications treated at the Phillippine General Hospital". International Journal of Gynaecology
and Obstetrics 48, Supplement: S53-S66.
Wasserheit J.N., J.R. Harris, J. Chakraborty, B.A. Kay and K.J. Mason. 1989. "Reproductive tract
infections in a family planning population in rural Bangladesh." Studies In Family Planning. 20,2:69-80.

World Bank. 1993b. World Development Reports 1993: Investing in Health, New York : Oxford
University Press.
World Health Organization. 1991. Essential Obstetric Functions at the First Referral Level. Geneva:
World Health Organization. 1993. Mother-Baby Package: A Road Map of Implementation in Countries.
Geneva: Division of Family Health/WHO

Zurayk, H., H. Khattab, N. Younis, O. Kamal and Mahinaz El-Helw, 1995. Comparing Women’s
Reports with Medical Diagnoses of Reproductive Morbidity Conditions in Rural Egypt. Studies in Family
Planning; 26, 1:14-21.

27

List of Acronyms
CHC
DF
FHI
Govt.
JIPMER
NGOs
PHC
PS
Pvt.
RVF
SD
STD
TBA
URI
VVF
TB

Community Health Centre
Degrees of freedom
Family Health International
Government
Jawaharlal Institute of Postgraduate Medical Education and Research
Non Governmental Organisation
Primary Health Centre
Puerperal Sterilization
Private
Recto Vaginal Fistula
Standard Deviation
Sexually Transmitted Diseases
Trained Birth Attendant
Upper Respiratory Infection
Vesico Vaginal Fistula
Tuberculosis

28

LIST OF TABLES
Table 1

Population sub-group and sampling procedure followed in the study area

Table 2

Households surveyed and the eligible women identified in rural and urban areas in Tamil
Nadu and Pondicherry.

Table 3

Number of women with completed interviews in rural and urban areas of Tamil Nadu and
Pondicherry.

Table 4

Distribution of women with one or more pregnancies during index period by study area.

Table 5

Selected demographic characteristics of eligible respondents by study area and residence.

Table 6

Selected socio-economic characteristics of eligible respondents by study area and
residence.

Table 7

Usual source of health facility availed by study area and residence.

Table 8

Selected antenatal care characteristics of the index pregnancy, by study area and
residence.

Table 9

Reasons for not availing antenatal care.

Table 10

Outcome of index pregnancies, place and person attending delivery by study area and
residence.

Table 11

Intended place of delivery, by actual place of delivery.

Table 12

Morbidity rates in index pregnancy during antenatal, intranatal and postnatal periods by
severity of morbidity.

Table 13

Life threatening or serious antepartum morbidities in index pregnancy and women who
perceived them as serious and who sought care.

Table 14

Other antepartum morbidities in the index pregnancy and women who perceived them as
serious and who sought care.

Table 15

Type of Life threatening or serious antepartum morbidities, by source of care.

Table 16

Type of intrapartum morbidities in the index pregnancy, by place of delivery.

Table 17

Type of intrapartum morbidity, specific intervention for women who delivered at hospital.

Table 18

Type of intrapartum morbidity, specific intervention for women who delivered at home.

Table 19

Life threatening or Serious postpartum morbidities in the index pregnancy.

Table 20

Other postpartum morbidities women who perceived them as serious and women who
sought care.

List of tables

Contd..

Table 21

Postpartum morbidities in the index pregnancy by place of delivery.

Table 22

Type of postpartum morbidity by source of care (Life threatening and serious).

Table 23

Postpartum morbidities by source of care (Others).

Table 24

Reason for not taking treatment for postpartum morbidities.

Table 25

Type of long-term morbidities and women who sought care.

Table 26

Type of long-term morbidities, by source of care.

Table 27

Number of morbidities in the index pregnancy, during ante, infra, postpartum periods
separately and combined by severity of morbidity.

Table 28

Number of long-term morbidities among eligible women.

H Table 29

Ratio of morbidities to maternal mortality

Table 30

Complications and treatment after last spontaneous or induced abortion.

Table 31

Reasons, method and type of practitioner for last induced abortion.

Table 32

Phase of pregnancy and type of morbidities in the index pregnancy by socio economic
status of respondent.

Table 33

Phase of pregnancy and type of morbidities in the index pregnancy by age of respondent.

Table 34

Phase of pregnancy and type of morbidities in the index pregnancy by parity.

Table 35

Life threatening or serious morbidities in different phases of pregnancy by study area and
residence.

_________

Table 1. Population sub-group and sampling procedure followed in the study area
Area

Population
Size

Study Population
selected from

Sampling Design

1. Pondicherry
1.1 Urban area:
a) Urban Wards of
Pondicherry

10000

2 out of 36
Census Wards

Simple
random

b) Urban Health
Centre area
of JIPMER

8000

Whole Population

Not applicable

20000

2 PHCs out of
17 PHCs

Simple random

6600

Whole Population

Not applicable

20000

6 out of 32 wards in
Tindivanam and 3
out of 36 wards in
Villupuram

Simple random

40000

Villages served by 8
sub centres

Three stage simple
random

1.2 Rural area

1.3 Rural Health
Centre area
of JIPMER

2. Tamil Nadu
South Arcot Dt.(TN)

2.1 Urban area Urban wards of
two towns
2.2 Rural area - *
from Vanur
Marakanam
Vikravandi and
Olakkur blocks.

Total

104600

* The sub group 2.2 (Rural area of Tamil Nadu ) was selected by a three stage simple random method. In the
first stage, four community health centres (CHC’s) out of 12 in the area were selected randomly. In the second
Stage, four primary health centres (PHC’s) were selected, one each from the four CHC’s. In the third stage
eight sub centres were selected from those PHC’s at the rate of two per PHC. Each sub centre catered to
about 5000 population. All the village attached to these sampled eight sub centres were included for the study.

Table: 2. Households surveyed and the eligible women identified in rural
and urban areas in Tamil Nadu and Pondicherry

Households visited

Eligible women
identified

No.

(%)

No.

(%)

Rural

8127

61.4

2539

68.9

Urban

5108

38.6

1147

31.1

Total

13235

100.0

3686

100.0

Study area and residence

Table 3. Number of women with completed interviews in Rural and
Urban areas of Tamil Nadu and Pondicherry.

Study area
and residence

No. of
« eligible
women
interviewed
in 1st phase

Eligible women with completed
interviews after 2nd phase ♦
No.

(%) Sample

(%) Overall

TAMIL NADU

Rural

1654

1570

72.6

47.0

Urban

621

592

27.4

17.7

Sub Total

2215

2162

100.0

64.8

PONDICHERRY

*

Rural

785

755

64.2

22.6

Urban

440

422

35.8

12.6

Sub Total

1225

1177

100.0

35.2

Grand Total

3500

3339

100.0

100.0

847 women pregnant during 1st phase of interview were reinterviewed after
delivery in 2nd phase.

Table: 4. Distribution of women with one or more pregnancies
during index period by study area.

Study Area

Number of Pregnancies
Total

1

2

3

4

Tamil Nadu

1890
(87.4)

255
(H.8)

16
(0.7)

1
(0.1)

2162
(100.0)

Pondicherry

977
(83.0)

187
(15.9)

11
(0-9)

2
(0.2)

1177
(100.0)

Total Women

2867
(85.9)

442
(13.2)

27
(0.8)

3
(0.1)

3339
(100.0)

Total Pregnancies

2867
(74.6)

884
(23.0)

81
(2.1)

12
(0.3)

3844
(100.0)

Figures in parenthesis are percentages.

Table 5. Selected demographic icharacteristics
'
of eligible respondents by study area and residence.
Characteristics

Tamil Nadu Rural| Tamil Nadu Urban
No.
(%)
No.
(%)

Pondicherry Rural
No.
(%)

Pondicherry UrbaD
No.

(%)

36
213
122
43
8
0
0
422
24.2

8.5
50.5
28.9
10.2
1.9
0.0
0.0
100.0
±4.1

292
1401
1142
343
132
24
5
3339
24.7

8.7
42.0
34.2
10.3
4.0
0.7
0.1
100.0
±4.6

26.4
57.5
14.4
1.7
100.0
±1.2

140
231
47
4
422
2.2

33.2
54.7
11.1
0.9
100.0
±1.2

854
1733
603
149
3339
2.6

25.6
51.9
18.1
4.5
100.0
±1.5

6.4
31.3
56.0
6.2
0.1
100.0
±1.1
22.6

16
171
215
20
0
422
1.6
422

3.8
40.5
50.9
4.7
0.0
100.0
±1.0
12.6

151
1023
1788
343
34
3339
2,1
3339

4.5
30.6
53.5
10.3
1.0
100.0
±1.2
100.0

148
618
547
161
75
17
4
1570
24.8

9.4
39.4
34.8
10.3
4.8 ’
1.1
0.3
100.0
±4.9

31
237
212
75
31
5
1
592
25.4

5.2
40.0
35.8
12.7
5.2
0.9
0.2
100.0
±4.8

Number of pregnancies'
1
2-3
4-5
> 6
Total
Mean (± SD)

77
333
261
64
18
2
0
755
24.2

10.2
44.1
34.6
8.5
2.4
0.3
0.0
100.0
±4.1

360
772
335
103
1570
2.8

22.9
49.2
21.3
6.6
100.0
±1.6

155
296
112
29
592
2.6

26.2
50.0
18.9
4.9
100.0
±1.5

199
434
109
13
755
2.4

56
439
832
216
27
1570
2,1
1570

3.6
28.0
53.0
13.8
1.7
100.0
±1.3
47.0

31
177
318
60
6
592
1.9
592

5.2
29.9
53.7
10.1
1.0
100.0
±1.3
17.7

48
236
423
47
1
755
1.8
755

Number of women

Table notes: * Significant difference

j

(%)

Age group (yrs)'
15-19
20-24
25-29
30-34
35-39
40-44
Missing
Total
Mean age (+ SD)

Number of live births'
0
1
2-3
4-5
>6
Total
Mean (+ SD)_________

No.

Total

>

Tamil Nadu Rural

Characteristics

No.

Socio-economic status
Low
Medium
High
Total

Employment status
Receives no payment
Receives payment
Total
__________
Place of employment*
Works at home
Works outside home
Works at home & outside
Total

nsiZZ

Tamil Nadu Urban

Pondicherry Rural

Pondicherry Urban

(%)

No.

(%)

No.

1007
315
149
97
0
0
2
1570

64.1
20. i
9.5
6.2
0.0
0.0
0.1
100.0

214
108
127
122
13
8
0
592

36.1
18.2
21.5
20.6
2.2
1.4
0.0
100.0

606
398
320
226
15
2
3
1570

38.6
25.4
20.4
14.4
1.0
0.1
0.2
100.0

108
97
133
196
46
12
0
592

586
883
101
1570

37.3
56.2
6.4
100.0

504
1066
1570
443
1086
41
1570

Education of Respondents
None
I Primary school
Secondary school
High school
Graduate
Post graduate
Don’t know
Total______________
Husband's education
None
Primary school
Secondary school
High school
Graduate
Post graduate
Don’t know
Total

by

Tahte 6. Selected sodo^conomic characteristics of

Table notes: * Based on women who do paid work

Total

(%)

No.

(%)

No.

(%)

378
170
114
89
4
0
0
755

50.1
22.5
15.1
11.8
0.5
0.0
0.0
100.0

104
86
81
115
33
3
0
422

24.6
20.4
19.2
27.3
7.8
0.7
0.0
100.0

1703
679
471
423
50
11
2
3339

51.0
20.3
14.1
12.7
1.5
0.3
0.1
100.0

18.2
16.4
22.5
33.1
7.8
2.0
0.0
100.0

171
169
189
206
19
1
0
755

22.6
22.4
25.0
27.3
2.5
0.1
0.0
100.0

59
56
109
130
55
13
0
422

14.0
13.3
25.8
30.8
13.0
3.1
0.0
100.0

944
720
751
758
135
28
3
3339

28.3
21.6
22.5
22.7
4.0
0.8
0.1
100.0

86
349
157
592

14.5
59.0
26.5
100.0

81
527
147
755

10.7
69.8
19.5
100.0

13
199
210
422

3.1
47.2
49.8
100.0

766
1958
615
3339

22.9
58.6
18.4
100.0

32.1
67.9
100.0

456
136
592

77.0
23.0
100.0

402
353
755

53.2
46.8
100.0

359
63
422

85.1
14.9
100.0

1721
1618
3339

51.5
48.5
100.0

28.2
69.2
2.6
100.0

443
143
6
592

74.9
24.2
1.0
100.0

388
362
5
755

51.4
47.9
0.7
100.0

353
67
2
422

83.6
15.9
0.5
100.0

1627
1658
54
3339

48.8
49.7
1.6
100.0

Table 7, Usual source of health facility used by study area and residence.
Tamil Nadu Rural

Place of service*
No usual place
Government services
Health center
Hospital
Private services
Clinic
Hospital
Pharmacy
Other systems
Other"

Tamil Nadu Urban

Pondicherry Rural

Pondicherry Urban

No.

(%)

No.

(%)

No.

(%)

No.

(%)

769

50.0

270

45.6

277

36.7

171

134
139

8.5
8.9

0
87

0.0
14.7

294
79

38.9
10.5

237
253
4

15.1
16.1
0.3

95
138
2

16.0
23.3
0.3

74
28
2

34

2.2

0

0.0

100.0

592

Total

Total

No.

(%)

40.5

1487

44.5

118
57

28.0
13.5

546
362

16.4
10.9

9.8
3.7
0.3

38
36
2

9.0
8.5
0.5

444
455
10

13.3
13.6
0.3

1

0.1

0

0.0

35

1.0

100.0

755

100.0

422

100.0

3339

100.0

Time to health facility***
< lA hour
A - 1 hour
> 1 hour

294
59
448

36.7
7.4
55.9

256
10
56

79.5
3.1
17.4

294
35
149

61.5
7.3
31.2

235
3
13

93.6
1.2
5.2

1079
107
666

58.3
5.7
36.0

Total

801

100.0

322

100.0

478

100.0

251

100.0

1852

100.0

Usual mode of transport***
Walking only
Non-mechanized means
Mechanized means
Walking and other means

148
13
599
7

19.3
1.7
78.1
0.9

228
45
34
15

70.8
14.0
10.6
4.7

262
30
181
4

54.9
6.3
37.9
0.8

148
58
11
34

59.0
23.1
4.4
13.5

821
146
825
60

44.3
7.9
44.5
3.3

Total

767

100.0

322

100.0

477

100.0

251

100.0

1852

100.0

Table notes: ♦ Based on all women
Includes village doctor, medical assistant, homeopath
*** Based on women who could name a usual source

o

Table 8. Seterted antenatal care characteristics of index pnymno, by rtudy am, and
_______________ Study area and residence
Characteristics___________

Tamil Nadu Rural

No.

(%)

Number of antenatal care visits
0
1-3
4-6
^7
Total
Mean number of visits”

50
748
684
130
1612
3.9

3.1
46.4
42.4
8.1

Gestational month of first visit”
1-3
4-6
>7
Total

Tamil Nadu Urban
No.
(%)

Pondicherry Rural

Pondicherry Urban

Total

No.

(%)

No.

(%)

No.

(%)

1.0
28.5
34.3
36.1

0.0
12.7
19.3
67.9

±3.5

0
58
88
309
455
8.1

±3.6

83
1220
1254
892
3449
5.1

2.4
35.4
36.4
25.9
100.0
±3.2

±2.1

25
192
215
172
604
5.1

±3.2

8
222
267
281
778
5.9

540
849
173
1562

34.6
54.4
11.1
100.0

312
214
53
579

53.9
37.0
9.2
100.0

308
397
65
770

Location offirst antenatal visit”
Govt health center
Govt hospital
Private sources
Home, by health worker
Other
Total

40.0
51.6
8.4
100.0

312
126
17
455

68.6
27.7
3.7
100.0

1472
1586
308
3366

43.7
47.1
9.2
100.0

489
133
453
483
4
1562

31.3
8.5
29.0
30.9
0.3

233
126
201
19
0
579

40.2
21.8
34.7
3.3
0.0

518
91
130
30
1
770

67.3
11.8
16.9
3.9
0.1

77.1
6.6
15.6
0.2
0.4

Reason for first visit”
Had specific problem
For confirmation
For antenatal check-up
Tetanus injection
Total

351
30
711
2
455

1591
380
855
533
7
3366

47.3
11.3
25.4
15.8
0.2
100.0

177
377
394
614
1562

11.3
24.1
25.2
39.3

30
239
151
159
579

5.2
41.3
26.1
27.5

5.8
27.9
42.7
23.5

19
228
157
51
455

4.2
50.1
34.5
11.2«

Subsequent hospital visits for the women
who had first visit at home_____________

45
215
329
181
770

271
1059
1031
1005
3366

8.1
31.5
30.6
29.9
100.0

121

25.1

1

5.3

7

23.3

1

100.0

130

24.4

Tetanus Toxoid Injection*
Received
Not Received
Total

1542
70
1612

95.7
4.3
100.0

562
42
604

93.1
6.9
100.0

760
18
778

97.7
2.3
100.0

447
8
455

98.2
1.8
100.0

3311
138
3449

96.0
4.0
100.0

Table Notes: * Based on all women. **
Based on all women with care.

4.1
31.8
35.6
28.5

Table 9. Reasons for not availing antenatal care*

Number

Percent

Felt fine

50

64.1

Financial problems

37

47.4

Family issues
Husband/family didn ft allow
No one to watch children
Too busy/too much work
Do nor want rhe child
Emotionally disturbed
Fear of injections
No reason
Others

7
13
34
7
4
3
5
3

8.9
16.7
43.6
8.9
5.1
3.8
6.4
3.8

Accessibility problems
Didn ft know about services
Too far away
No transport
Inconvenient hours

8
3
4
3

10.3
3.8
5.1
3.8

Acceptability problems
Poor quality care/bad
reputation

3

3.8

Number of women giving
No reason
One reason
Two reasons
Three reasons

5
13
27
38

6.4
16.7
34.6
48.7

Number of women

83

Reasons**

a

Table notes:

4

* Based on women with no antenatal care.
** Multiple response.

Table 10. Outcome of index pregnancies, place and person attending delivery by study area and residence. (N - 3449)
Study area and residence Tamil Nadu Rural Tamil Nadu Urban Pondicherry Rural Pondicherry Urban

Total

Characteristics

No.

%

No.

%

No.

%

No.

%

No.

%

Outcome of index pregnancy'
Live birth
Still birth
Spontaneous abortion
Induced abortion
Total

1578
34
105
45
1762

89.6

86.0
1.5
7.1
5.4
100.0

758
20
75
22
875

86.6
2.3
8.6
2.5
100.0

445
10
32
30
517

86.1
1.9
6.2
5.8
100.0

3375
74
261
134
3844

87.8

100.0

594
10
49
37
690

100.0

Type of delivery
Normal
Caesarean
Instrumental

1492
46
74

92.6
2.9
4.6

519
52
33

85.9
8.6
5.5

702
45
31

90.2
5.8
4.0

402
45
8

88.4
9.9
1.8

3115
188
146

90.3
5.5
4.2

369
28
181
1023
11
1612

22.9
1.7
11.2
63.5
0.7
100.0

291
0
179
132
2
604

48.2
0.0
29.6
21.9
0.3
100.0

434
56
66
217
5
778

55.8
7.2
8.5
27.9
0.6
100.0

305
1
118
29
2
455

67.0
0.2
25.9
6.4
0.4
100.0

1399
85
544
1401
20
3449

40.6
2.5
15.8
40.6
0.6
100.0

215
420
529
378
70
1612

13.3

210
304
55
24
11
604

34.8
50.3
9.1
4.0
1.8
100.0

200
372
141
50
15
778

25.7
47.8
18.1
6.4
1.9
100.0

159
268
14
8
6
455

34.9
58.9

784
1364
739
460
102
3449

22.7
39.5
21.4
13.3
3.0
100.0

Place of delivery"
Govt, hospital
Govt, health centre
Private hospital
Home

Other
Total

1.9

6.0
2.5

1.9

6.8
3.5

Attendant at delivery"
Physician

Midwife
Untrained TBA

Trained TBA
Other
Total

26.1
32.8
23.4
4.3
100.0

Table notes: * Based on all women.
**
Based on women with live or still births

3.1

1.8
1.3
100.0

Table 11. Intended place of delivery, by actual place of delivery (Overall) N—3449
Actual place of delivery
Intended place of
delivery

Govt, hospital

Govt, health
center

Private
sources

Total

Other

Home

No.

(%)

No.

(%)

No.

(%)

No.

(%)

No.

(%)

No.

(%)

Govt, hospital

1137

76.9

62

4.2

9

0.6

260

17.6

10

0.7

1478

42.8

Private sources

73

12.6

422

5

0.9

77

13.3

0

0.0

577

16.7

Govt, health center

29

29.0

2

2.0

50

50.0

18

18.0

1

1.0

100

2.9

Home

139

11.3

47

3.8

21

1.7

1022

82.7

7

0.6

1236

35.8

Other

21

36.2

11

19.0

0

0.0

24

41.4

2

3.4

58

1.7

Table Notes: * Based on women with live or stillbirths.

Table 12. Morbidity rates in index pregnancy during antepartum, intrapartum
and postpartum periods by severity of morbidity.

Overall morbidities n —3844°

Life threatening
Serious
Total

%

4.7
40.7
55.6

Antepartum morbidity 0=3844°
Life threatening
Serious
Total

1.3

24.3
38.8

Intrapartum morbidity n=3449‘
Life threatening
Serious
Total

2.1
12.8
18.7

Postpartum morbidity n=3449‘
Life threatening
Serious
Total

Table notes: @ Includes abortion.
* Live births and still births.

2.2
16.4
24.6

o

Table 13. Life threatening or Serious antepartum morbidities in index pregnancy and women who perceived them as serious and
who sought care. (N=3844)

Type of morbidity

Women with the morbidity
Number

Rate / 1000

Perceived as serious

Sought care

Number

Percent

Number

Percent

Life threatening morbidities
Antepartum bleeding

23

6.0

18

78.3

20

87.0

Malaria

15

3.9

14

93.3

15

100.0

Fits/convulsions

10

2.6

10

100.0

9

90.0

Serious morbidities

Severe vomiting

696

181.1

333

47.8

524

75.3

Severe edema (hand /feet)

196

51.0

123

62.8

168

85.7

Fever > 3 days

151

39.3

92

60.9

145

96.0

Pregnancy induced hypertension

41

10.7

29

70.7

38

92.7

Heart disease

14

3.6

12

85.7

12

85.7

Pulmonary TB

8

2.1

6

75.0

7

87.5

Jaundice

5

1.3

5

100.0

5

100.0

Diabetes mellitus

5

1.3

4

80.0

5

100.0

„d

T.W. .4 , Oftc
Women with the morbidity
Number
|
Rate/1000

Type of morbidity

White discharge

T

Perceived as serious
Number | Percent

Sought care
Number

|

Percent

Infections
72

18.7

49

68.1

39

54.2

31

8.1

24

77.4

27

87,1

12

3.1

Typhoid

9

75.0

10

83.3

10

2.6

Chicken pox

4

40,0

7

70,0

9

2.3

3

33.3

3

2

33.3

0.5

2

100.0

2

100.0

0.3

1

100.0

1

100.0

Diarrhoea

URI

STD

Filaria

1

Other conditions

Giddiness

668

173.8

Irritative urinary symptoms

331

49.6

431

411

64.5

_______ 106.9

125

30.4

105

267

25,6

_________ 69.5

161

Edema (hand/feet)

60,3

148

55.4

235

_________ 61.1

Anaemia

94

40,0

183

29

77,9

__________7.5

22

75.9

24

6

82.8

_________ 1.6

Epilepsy

4

75.0

3

50,0

4

Ulcer

__________ 1.0

2

50.0

2

50.0

4

_________ 1.0

4

100.0

4

3

100.0

_________ 0.8

2

66.7

2

3

66.7

_________ 0.8

2

66,0

3

100.0

2

_________ 0.5

Skin disease

2

100.0

2

100.0

2

_________ 0.5

Hydatidiform mole

0

0.0

2

100.0

1

_________ 0.3

0

0.0

1

100.0

1

_________ 0.3

0

0.0

0

0.0

1

_________ 0.3

1

100.0

1

100.0

1

_________ 0.3

1

100.0

1

100.0

1

0.3

0

0.0

0

0.0

Blurred vision

Bleeding from other sites "

Swelling/Tumour
IUGR
Growth in vagina

Mental upset

Snake bite

Congenital anomaly
Phys ica Uy handicapped

* Bleeding from nose and mouth

r

Table 15. Type of Life threatening or Serious antepartum morbidities, by source of care. (N=3844)
Source of care

Type of morbidity

Govt, health
center

Govt.
Hospital

No.

No.

%

%

Private
sources

No.

%

Chemist’s
shop

Others

No
treatment

Total

No.

%

No.

%

No.

%

No.

%

Life threatening morbidities
Antepartum bleeding

3

13.0

3

13.0

14

60.9

0

0.0

0

0.0

3

13.0

23

100.0

Malaria

2

13.3

1

6.7

12

80.0

0

0.0

0

0.0

0

0.0

15

100.0

Fits/convulsions

2

20.0

3

30.0

3

30.0

0

0.0

1

10.0

1

10.0

10

100.0

Serious morbidities
Severe vomiting

80

11.5

63

9.1

337

28.4

5

0.7

38

5.5

173

24.9

696

100.0

Severe edema (hands/feet)

31

15.8

11

2.7

67

34.2

3

1.5

18

9.2

28

14.3

196

100.0

Fever > 3 days

19

12.6

10

6.6

67

44.4

14

9.3

1

0.7

40

26.5

151

100.0

Pregnancy induced hypertension

6

14.6

9

22.0

15

36.6

0

0.0

0

0.0

11

26.8

41

100.0

Heart disease

1

7.1

5

35.7

4

28.6

2 ■

14.3

2

14.3

2

14.3

14

100.0

Pulmonary TB

0

0.0

3

37.5

4

50.0

0

0.0

0

0.0

1

12.5

8

100.0

Jaundice

1

20.0

0

0.0

2

40.0

0

0.0

2

40.0

0

0.0

5

100.0

Diabetes mellitus

1

20.0

1

20.0

3

60.0

0

0.0

0

0.0

0

0.0

5

100.0

Note: Govt. = Government

Table 16. Type of intrapartum morbidities in the index pregnancy, by place of delivery (Overall) N~3449'.
Type of morbidity

______ Morbidities
Number
Rate / 1000
3449

Excessive bleeding

34

Loss of consciousness
Fits /convulsions

30

Caesarean Section

Early rupture of membrane
Abnormal position
Labor >18 hours_____
Twins_________

Meconium staining
Placenta previa

10

129
116
98
47
26
8
3

Hospital delivery
Number
Rate/1000 of
2028
hospital delivery

Life threatening morbidities
9.9
23___
8.7
13___

2.9

_5___

Serious morbidities
37,4
129

33.6
28.4

13.6
7.5
2.3
0.9

107
94
40
17
6
3

Retained placenta

6

Other morbidities
63.5
176
1,7
0

Inversion of uterus

2

0.6

0

Cord round neck
Any morbidity

2

0.6

1___

585

169.6

Any tear

219

Table notes: * Based on women with live or stillbirths.

487

11.3
6.4
2.5
63.6
52.8
46.4
19,7
8.4
3.0

11
17
5

7.7
11.9
3.5

_9
7_
9_
2_
0

6,0
2.8
4.0
6.0
1.4
0.0

43
6
2
1
98

30.2
4.2
1.4
100.0
69.0

4_

1.5
86.8
0.0
0.0
0.5
240.1

_______ Home delivery
Number Rate/1000 of home
1421
delivery

Table 17. Type of intrapartum morbidity, specific intervention’ for women who delivered at hospitel N=2028.
Place of delivery

Type of morbidity

______ Hospital Delivery

Total
No.

Drugs

No.

C section

%

No.

%

Forceps/Vaccum
extraction

Episiotomy

No.

%

No.

%

Suturing of
tear
No.

%

No specific
intervention

No.

%

Life threatening morbidities
Excessive bleeding

23

9

26.5

6

17.7

2

5.9

3

8.8

3

Loss of consciousness

8.8

13

3

10.0

0

0.0

6

20.0

2

6.7

Fits /con vulsions

2

6.7

6

1

10.0

0

0.0

0

0.0

2

20.0

3

30.0

Serious morbidities

Early rupture of membrane

107

30

25.9

23

19.8

15

12.9

29

25.0

10

Abnormal position

8.6

94

8

8.2

57

58.2

15

15.3

12

12.2

2

Labor > 18 hours

2.0

40

7

14.9

12

25.5

9

19.2

9

19.2

Twins

3

6.4

16

5

19.2

I

3.9

3

11.5

6

23.1

2

Meconium staining

7.7

6

1

12.5

2

25.0

2

25.0

0

0.0

Placenta previa

1

12.5

3

1

33.3

2

66.7

0

0.0

0

0.0

0

0.0

3

1.4

0

0.0

Other morbidities

Any tear

176

108

49.3

0

0.0

7

3.2

0

0.0

Cord round neck

1

0

0.0

1

50.0

0

0.0

0

0.0

Table notes: * Based on women with live or stillbirths.
Women may have more than one morbidity or intervention.
There was no hysterectomy

166

75.8

Table 18. Type of intrapartum morbidity, specific intervention* for women who delivered at home N-1421
Total

Episiotomy

Drugs

No specific intervention

Type of morbidity

No.

No.

(%)

No.

(%)

No.

(%)

Life threatening morbidities
Lxcessive bleeding

11

1

9.1

1

9.1

9

81.8

Loss of consciousness

17

3

17.6

1

5.9

13

76.5

Fits/con vulsions

5

0

0.0

0

0.0

5

100.0

Serious morbidities
Early rupture of membrane

9

2

22.2

1

11.1

6

66.7

Abnormal position

4

0

0.0

0

0.0

4

100.0

Labor >18 hours

7

1

14.3

0

0.0

6

85.7

Twins

9

1

11.1

1

11.1

7

77.8

0

0.0

0

0.0

2

100.0

Meconium staining

Other morbidities
Any tear

43

4

9.3

0

0.0

39”

90.7

Retained placenta

6

1

16.7

0

0.0

5

83.3

Inversion of uterus

2

1

50.0

0

0.0

1

50.0

Cord round neck

1

0

0.0

0

0.0

1

100.0

Table notes: * Based on women with live or stillbirths.
** Includes 5 women in whom, tear was sutured by the midwife

__ ____________ __

.................................................. 1

:

,

. .

.

...................................................

------- —

i

"......................

Table 19* Life threatening. Serious and Infections among postpartum morbidities in index pregnancy, and according to
women who perceived them as serious and women who sought care (N=3449)

Type of morbidity

Number of
women

Rate per 1000
pregnancies

Perceived serious
Number of
women

Sought care

Percent

Number of
women

Percent

Life threatening morbidities
Excess bleeding

63

18.3

39

61.9

41

65.1

Unconsciousness

30

8.7

17

56.7

19

63.3

Fits

13

3.8

10

76.9

10

76.9

Serious morbidities
Fever >3 days

151

43.8

99

65.6

141

93.4

Pain abdomen

122

35.4

80

65.6

89

73.0

Puerperal sepsis

25

7.2

16

64.0

23

92.0

Foul smelling lochia

16

4.6

9

56.3

9

56.3

Infections

Infected tear

93

27.0

78

83.9

90

96.8

Painful/Burning micturition

90

26.1

51

56.7

52

57.8

Breast swelling

80

23.2

57

71.2

63

78.7

Urinary infection

29

8.4

20

69.0

28

96.6

Infected tear with fever

28

8.1

22

78.6

25

89.3

Mastitis

26

7.5

20

76.9

24

92.3

Infected PS wound

19

5.5

11

57.9

18

94.7

-

Table 20. Other postpartum morbidities women who perceived them as serious and women who sought careen»3449)
Type of morbidity

Body pain

Diarrhoea
White discharge
~URI
Mental illness

Burning sensation
Manguthu valli
Chest pain
Gastritis
Constipation
Head ache
Jaundice
Vomiting
Pain in vagina
Blurred vision
Asthma
Scanty' flow
High blood pressure
Swelling of leg

Number of
women with
morbidity
40
29

Perceived as serious

Sought care

Rate/1000
pregnancies

No.

%

No.

%

11.6

30

75.0

82.5

23

79.3
85.7

33
27

21
H

"oT

H~6

8
8
8
7
4
3
3
3
3
3
3
2
~2

~23~

~2?3~

~6
~6

2.3

T

To"
~oV
"o'

4
~2

1

"oT

"o’

0.3

1

~oT
oT
"oT

jT

50.0
100.0
66.7
66.7
100.0
66.7
100.0
50.0
100.0
0.0
100.0
0.0

i

100.0

1

T

100.0

"7

0.3

i

0.0

100.0
100.0
100.0

"7
"7
"7
"7

’oV
’(jy

"oV
~0?9~
0.9

"oT

T
T
T
T
T
T
T
~2

Bleeding from other sites
Tumour

i
i
i

Heart attack
Prolapse

1
1

Tuberculosis
Typhoid
STD

i

0.3

i

1

"oT

T
T

i

0.3

54.5
75.0

To
T

75.0
37.5
57.1

4
~6

4

T
2

~T
T
~T
2

1

"7
1

IT

93.1
76.2
90.9

50.0
87.5
50.0
85.7
100.0
100.0
66.7
100.0
100.0
100.0
100.0
100.0
50.0
100.0
100.0
0.0
100.0

100.0
100.0

100.0
100.0
0.0

Table 2L Postpartum morbidities in the index pregnancy by place of delivery♦ (N=3449)
Women with morbidity

Type of morbidity

Number

Excess bleeding
Unconsciousness
Fits

Fever >3 days

Pain abdomen____
Puerperal sepsis
Foul smelling lochia

63
30
13
151
122
25
16

Infected tear______

93

Painful/Burning micturition

90

Breast swelling

80

Urinary infection_______

29

Infected tear with fever

28

Mastitis

26

Infected PS* wound

19

Total deliveries
*

PS = puerperal sterilization

3449

Hospital delivery

Home delivery

Percent of Rate/1000
Rate/1000 Number• total with of hospital
morbidity
delivery
Life threatening morbidities
18.3
37
58.7

Number

Percent of Rate/1000
total with of home
morbidity delivery

46.2

18.2
7.9
3.0

26
14
7

41.3
46.7
53.8

18.3
9.9
4.9

Serious morbidities
43.8
86
56,9

42,4

65

43.1

45.7

50.0

30.1

61

50.0

42.9-

52.0

6.4

12

48.0

8.4

37.5

3.0

10

62.5

7.0

84,9
52.2
56.3
31.0
71.4
53.8
_ 68.4

39,0
23.2
22.2
- 4.4
9.9
6.9
6.4

14

15.1
47.8
43.8
69.0
28.6
46.2
31.6

9.9
30.3
24,6
14.1
5.6
8.4
4.2

8.7
3.8

16
6

53.3

35.4
61
7.2
13
4.6 _______ 6
27.0
26.1
23.2
8.4
8.1
7.5
5.5

Infections
79
47

45
9
20

14
13
2028

43
35
20
8

12
6
1421

1

1"

=====^ .



Tabic 22. Type of postpartum morbidity by source of care.

Source of care
Type of morbidity

Govt* health
center

No.

I %

Govt.
hospital

No.

Private
sources

No.

%

%

Chemist’s
shop

No.

No
treatment

Other

Total

%

No.

%

No.

%

No. |

%

Life threatening morbidities

Excess bleeding

12

19.0

10

15.9

14

22.2

0

0.0

5

7.9

22

34.9

63

100.0

Unconsciousness

1

3.3

8

26.7

9

30.0

0

0.0

0

0.0

12

Fits

40.0

30

100.0

1

7.7

4

30.8

5

38.5

0

0.0

0

0.0

3

23.1

13

100.0

9.3

5.3

10

6.6

151

100.0

9.0

33
2
7

27.1

122

100.0

8.0

25

43.8

16

100.0
100.0

4.3
42.2

93

100.0

90

100.0

80

100.0

29

100.0

Fever > 3 days

17

Pelvic pain
Puerperal sepsis

18
2

Foul smelling lochia

3

11.3
14.7

32

Serious morbidities
21.2
70
46.4
14

23

18.8

33

27.1

4

3.3

8
11

8.0

8

32.0

20.0

4

16.0

4

18.8

3

18.8

5
2

12.5

0

0.0

1

16.0
6.3

Other morbidities
Infected tear

17

18.3

32

34.4

37

39.8

2

2.2

1

1.1

4

Painful/Burning micturition

9

10.0

11

12.2

2

2.2

12

6.3

19

10.3

12

30.0
41.4

5

6

10
3

12.5

Urinary infection

6.3
20.7

13.3
23.7

38

5

18
24

20.0

Breast swelling

1

3.4

6

20.7

1

21.3
3.4

Infected tear with fever

6

21.4

8

28.6

10

35.7

1

3.6

0

0.0

3

10.7

Mastitis_____

28

100.0

1

7

26.9

10

38.5

4

15.4

2

7.7

2

7.7

26

Infected PS wound

100.0

4

3.8
21.1

8

42.1

6

31.6

0

0.0

0

0.0

1

5.3

19

100.0

Table notes:

* Govt=govemment
** Others = home treatment and traditional treatment

17

Table 23, Other Postpartum morbidities by source of care.

Source of care
Govt, health
centre

Govt, hospital

No?

(%)

No.

(%)

No.

No.

(%)

No.

(%)

No.

(%)

3

w

No.

Body pain

7.5

9

22.5

16

40,0

4

7

17.5

100.0

17.2
~ 9.5

17

0

6.9

29

100.0

23.8

~6

2?

100.0

9.1

54.6

~0

0

9.1

12.5

7

25.0

12.5

25.0

~4

50.0

0

~77

12.5

1

~2

25,0

P.O

50,0

14,3

7

0

Chest pain

12.5
~ 0.0

~5

71.4

~0

Gastritis

o’

0

75.0

o’

0.0

Constipation
Head ache

4

100.0

0.0

1

33.3

~2

o’

0.0

0.0

1

33.3

Vomiting

o’

"7
~o

100.0

33.3

Jaundice

7
7

0.0

0.0
0.0

1

33.3

66.7

”o

Pain in vagina

2

0.0

33.3

“o

i

o’

333

Blurred vision

0.0

7
7

"7
"7

7
7
o
7
7

100.0

25.0

~7

7

1

T
~7
~7

14.3

0.0

"77
577
sTI
~77
777
777

7T
"7
T
T

100.0

0.0

T

~77
~77
~77

7

17

48.2
-57.I

~o
~o

40

17.2

?
7
T
7
7

10.0

Diarrhoea

y
7
T
T
T
T
T

i

■7

66.7

Scanty flow

0.0

1

0.0

0.0

7

100.0

0.0

”o

0.0

1

100.0

"7
"7
~7

~7
T

0.0

0.0

1

100.0

0.0

0.0

o’

0.0

Heart attack

o’

0.0

Prolapse

7
7
7
7

"7
"7
~7
~7
T
T
~7

~o
"7
~o

Tumour

7
7
7
7
7
7
7
7

0.0

Bleeding from other sites

o'
o'
o'
o'
7

0.0
?50.0

7
7
7
7
7
7
7
7
7
7
7
7

"77
TT777
777
~77
777
777
777
777
777
777
~77

Type of morbidity

White discharge
~URI
Mental illness
Burning sensation

Manguthu valli

Asthma
High blood pressure
Swelling of leg

Tuberculosis

Typhoid
~STD

o’

9.5

77.3
72.5

0.0

100.0

7o.o
0.0
0.0

33.3

100.0

0.0
0.0
0.0
0.0
0.0

Private
sources

Chemist’s
shop

“o

33.3

0.0

Too.o

0
~0

0.0

~0

100.0

~0

100.0

"o

0.0

~0

0.0

"77
~67
"77
~77
”77
"77
”67
~67
"77
333

”77
”77
"77

"77
77
"77
"77
"77
0.0

"77
"77
"77
"67
"67

Other

0

No treatment

■5

7
7
T
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7

50.0

12.5

Total

33.3

100.0

Too.o
100.0

100.0

0.0

T
T
T
T

50.0

~2

100.0

0.0

~2

100.0

"77

T
"7
T
"7
T
"7
T
T
~i~

100.0

0.0

0.0

0.0

0.0
100.0

0.0
0.0
0.0
0.0
0.0

100.0

100.0

7677
100.0
100.0

100.0
100.0

100.0
100.0
100.0

100.0
100.0
100.0

Table 24: Reason for not taking treatment for postpartum morbidities.
Costly

Poor
care

Too far /
No
transport

Too much
work at
home

Family
Problem

Don’t
know the
place

Home
treatment

Did not
feel the
need

Others

Total

Foul smelling discharge

3

0

I

1

0

2

0

1

3

6

Fever

3 days

4

0

2

3

0

0

0

5

3

8

Pain Abdomen

11

1

1

9

2

0

1

13

13

30

Pus in stirches

8

0

2

3

3

0

0

8

11

20

Puerperal sepsis

0

0

1

0

0

0

0

0

1

2

Infected tear with fever

1

1

0

0

0

0

0

0

1

3

Infected tear

0

0

0

0

0

0

0

0

4

4

Morbidity

Table 25. Type of long-term morbidities* and women who sought care.
(N=3844)

Type of morbidity

Women with specific
morbidity

Sought care

Number

Rate/1000

Number

Percent

Uterine prolapse®

117

30.4

23

19.7

Incontinence of urine**

114

29.7

18

15.8

Dyspareunia

61

15.9

11

18.0

Haemorrhoids11

48

12.5

24

45.8

Rectovaginal fistulae (Passage of
gas/Passage of stool)

2

0.5

1

50.0

Table Notes:
* Based on all women.
** No women had vesico-vaginal fistulae in our study
7 women had forceps and 2 women had prolonged labour.
Symptom was not persisting in 41 women.

@ 1 woman had prolonged labour and 7 had forceps.
If Symptom was not persisting in 11 women.

5

Table 26. Type of long-term morbidities, by source of care. (N=3844)

Source of care

Type of morbidity

Govt, health
center

Govt,
hospital

Private
doctor/clinic

Private
hospital

Chemist’s
shop

Other**

No care

Total

No.

%

No.

%

No.

%

No.

%

No.

%

No.

%

No.

%

No.

%

Uterine prolapse

2

1.7

7

6.0

6

5.1

7

6.0

0

0.0

1

0.9

94

80.3

117

100

Incontinence of urine

4

3.5

1

0.9

7

6.1

2

1.8

2

1.8

2

1.8

96

84.2

114

100

Dyspareunia

1

1.6

0

0.0

5

8.2

3

4.9

0

0.0

2

3.3

50

81.9

61

100

Haemorrhoids

3

6.3

4

8.3

8

16.7

5

10.4

0

0.0

4

8.3

24

50.0

48

100

Rectovaginal fistulae (Passage
of gas/Passage of stool)

0

0.0

1

50.0

0

0.0

0

0.0

0

0.0

0

0.0

1

50.0

2

100

Table Notes: ** Includes traditional practitioners, TBAs, village doctors, homeopaths and medical assistants.

>

»

Table 27. Number of morbidities in the index pregnancy, during ante, intra,
postpartum periods separately and combined by severity of morbidity.

f

Severity of morbidity
Number of
morbidities

Life
threatening

No.

(%)

Serious

No.

(%)

Other
morbidities

Ail morbidities

No.

(%)

No.

(%)

Antepartum morbidities N=3844
0

3795

98.7

2911

75.7

2353

61.2

2353

61.2

1

49

1.3

759

19.7

1145

29.8

549

14.3

2

0

0.0

164

4.3

309

8.0

583

15.2

3

0

0.0

10

0.3

37

1.0

359

9.3

Average

.01

3

.8

.5

Intrapartum morbidities N=3449

0

3376

97.9

3007

87.2

2805

81.3

2805

81.3

1

71

2.1

275

8.0

462

13.4

202

5.9

2

2

0.1

137

4.0

151

4.4

209

6.1

2> 3

0

0.0

30

0.8

31

0.8

233

6.7

r

Average

.02

.3

.2

.5

Postpartum morbidities N=3449

0

3374

97.8

3088

89.5

2836

82.2

2835

82.2

1

74

2.1

340

9.9

557

16.1

242

7.0

2

1

0.0

19

0.6

50

1.4

270

7.8

2* 3

0

0.0

2

0.1

6

0.2

102

3.0

Average

.02

r

.1

.2

.3

Ante*, intra- and postpartum morbidities of the index pregnancy N=3844

5A

0

3662

95.3

2280

59.3

1708

44.4

1667

43.4

1

166

4.3

957

24.9

1178

30.6

565

14.7

2

14

0.4

436

11.3

611

15.9

649

16.9

3

2

0.1

121

3.1

231

6.0

335

8.7

4

0

0.0

39

1.0

82

2.1

286

7.4

5

0

0.0

11

0.3

34

1.0

342

8.8

Average

.05

.6

.9

1.6

Table 28. Number of long-term morbidities
among eligible women* (N = 3844).

Number of
morbidities

Women
Number

Percent

0

3065

91.8

1

235

7.0

2

36

1.0

> 3

3

0.1

Table notes:
*
Morbidity could have been caused by other than the
index pregnancy.
Long-term morbidity includes RVF, incontinence,
uterine prolapse, dyspareunia, haemorrhoids.

Table 29. Ratio of morbidities* to maternal mortality**.

Indicator

Life threatening

Life threatening or
Serious

All morbidities

No.

Ratio

No.

Ratio

No.

Ratio

Number of antepartum morbidities per
maternal death

49

6.1

1117

139.6

1880

235.0

Number of intrapartum morbidities per
maternal death

75

9.4

864

108.0

864

108.0

Number of postpartum morbidities per
maternal death

76

9.5

643

80.4

858

107.3

Number of ante-, intra- and postpartum
morbidities per maternal death

200

25.0

2624

328.0

3602

450.3

2

0.25

225

28.1

2626

328.3

3827

478.4

Number of long-term sequelae per maternal
death

Number of morbidities (all types) per
maternal death
Table Notes:

200

25.0

* Based on all morbidities.
** Maternal mortality ratio used in calculations: 2.3 per 1000 live births. (Source: Danida Final Report)

Table 30, Complications and treatment after last spontaneous or induced abortion

Characteristics

Spontaneous abortion

Induced abortion

Number

Percent®

Number

Percent

Symptoms reported after most
recent miscarriage / abortion"
None
Bleeding
Fever
Other (pain abdomen)
Total

16
212
76
205
245

6.1
81.2
29.1
78.5
93.9*

42
56
23
68
92

31.3
41.8
17.2
50.7
68.7*

Sought treatment"
Yes
No
Total

179
82
261

68.6
31.4 .
100.0

88
46
134

65.7
34.3
100.0

Place of treatment"*
Govt hospital and health center
Private hospital / doctor / clinic
No treatment
Total

66

102
77
245

27.0
41.7
31.4
100.0

21
42
28
92

22.8
45.0
31.5
100.0

Admitted to hospital*"
No
Yes
Total
Mean days in hospital (±SD)

177
68
245
3.1

72.2
28.8
100.0
± 2.1

70
22
92
3.4

76.1
23.9
100.0

± 2.7

Table Notes:
Based on all women, but does not necessarily refer to the index pregnancy.
** Based on women who reported a sponUneous or induced abortion.
Based on women who reported having a complication.
@ Percentage of total women.
# 93.9 % of women had some complication. The total % cannot be 100 as this
is a multiple response question.

Table 31. Reasons, method and type of practitioner for last induced
abortion* (N=134)

Characteristics

Number

Percent®

66
28
17
11
4
1

49.2
20.9
12.6
8.2
2.9
0.8

Yes, modern contraceptives
Total

131
3
134

97.7
2.3
100.0

Method of induction
Dilation & curettage
Injection
Pilis
Herbs
Massage
Other
Total

94
26
9
1
2
2
134

70.1
19.4
6.7
0.7
1.5
1.5
100.0

Type of practitioner
Doctor
Midwife or nurse
Pharmacist
TBA/traditional practitioner
Relati ve/husband/friend
Other
Total

119
4
0
2
4
5
134

88.8
3.0
0.0
1.5
2.9
3.7
100.0

Reason for induction*
No more children desired
Cannot afford to have more children
Pregnancy interferes with work
Health reasons
Family reasons
Previous child too small
Trying to prevent conception?
No

Table notes: * Based on women who reported an induced abortion.
@ Percentage of total women.
** Multiple response question.

Table 32. Phase of pregnancy and type of morbidities in the index pregnancy by income of respondent.
Phase of
pregnancy

Antepartum
morbidity
N=3844
Intrapartum
morbidity
N=3449

Postpartum
morbidity
N=3449
Total
morbidity
N=3884

Income Group (Rs.)

Type of morbidity

< 500

501 - 1000

Chi square test

> 1000

No.

%

No.

%

No.

%

Value

DE

Sig.

Life threatening & Serious

436

23.8

334

26.0

163

22.5

3.6

2

0.16

All other morbidities

611

36.9

519

40.4

295

40.6

5.1

2

0.08

Life threatening & Serious

192

11.6

141

12.3

109

16.8

11.7

2

0.02

All other morbidities

280

16.9

205

18.0

159

24.5

18.4

2

0.01

Life threatening & Serious

204

12.3

114

10.0

43

6.6

16.5

2

0.01

All other morbidities

335

20.2

199

17.4

80

12.3

19.9

2

0.01

Life threatening & Serious

746

40.7

529

41.2

289

39.8

0.4

2

0.84

All other morbidities

1019

55.6

737

57.4

421

58.0

1.6

2

0.44

Table 33, Phase of pregnancy and type of morbidities in the index pregnancy by age of respondent,

Age Group (in years)
Phase of
Pregnancy

Antepartum
morbidity
Intrapartum
morbidity
Postpartum
morbidity
Total
morbidity

Type of morbidity

< 19

Chi square test

19 - 35

> 35

No.

%

No.

%

No.

%

Value

DF

Sig.

Life Threatening & Serious

39

26.4

875

24.2

19

23.8

2.3

3

0.52

All other morbidities

63

41.9

1401

38.8

27

33.8

2.7

3

0.44

Life Threatening & Serious

7

5.6

423

13.0

12

17.9

8.4

3

0.04

All other morbidities

21

16.8

610

18.8

13

19.4

1.7

3

0.64

Life Threatening & Serious

21

16.8

331

10.2

9

13.4

6.9

3

0.07

All other morbidities

36

28.8

566

17.4

12

17.9

11.9

3

0.01

Life Threatening & Serious

65

43.9

1464

40.6

35

43.8

5.1

3

0.17

All other morbidities

102

68.2

2032

56.3

43

53.8

12.5

3

0.01





.

Phase of
Pregnancy

Antepartum
morbidity

Intrapartum
morbidity'

Postpartum
morbidity
Total
morbidity

Table 34. Number of morbidities in the index pregnancy by parity
By parity
Type of morbidity

1

Chi square test

2-4

No.

%

No.

%

No.

%

Value

DF

Sig.

Life threatening & Serious

320

28.0

503

22.1

110

25.6

14.5

2

0.00

All other morbidities

514

44.9

811

35.7

166

38.7

27.2

2

0.00

Life threatening & Serious

171

16.5

229

11.3

42

11.0

18.3

2

0.00

All other morbidities

264

25.5

332

16.3

48

12.6

48.7

2

0.00

Life threatening & Serious

108

10.4

211

10.4

42

11.0

0.1

2

0.93

AU other morbidities

190

18.4

357

17.6

67

17.6

0.3

2

0.85

Life threatening & Serious

520

45.5

871

38.4

173

40.3

15.9

2

0.01

All other morbidities

737

64.4

1210

53.3

230

53.6

40.3

2

0.01

Table 35. Number of pregnancy of Life threatening and serious morbidities by study area and residence (N = 3449)

No.

%

No.

%

No.

%

No.

%

Chisquare
value

Tamil Nadu Rural Tamil Nadu Urban

Pondicherry Rural

Total

Pondicherry Urban

, Signifi
DF
cance

Antepartum

1491

761

43.2

235

34.1

304

34.7

191

36.9

T1J

3

o.Oo

Intrapartum

644

263

16.3

142

23.5

136

17.5

103

22.6

20.6

3

0.00

Postpartum

614

376

23.3

90

14.9

100

12.9

48

10.5

66.5

3

0.00

Total

2177

1042

59.1

386

55.9

472

53.9

277

53.6

9.2

3

0.00

J

Annexure - i

PROJECT JIPMER

MATERNAL MORBIDITY
STUDY

AREA PROFILE
RECORDER - RURAL

SL
NO

Survey area/center

Village No

Village

_ Taluk

Compiled by

Sign

Date

Scrutinized by

Sign

Date

RECORD AREA PROFILE OF THE VILLAGE

SECTION 1 : PEOPLE

la.

Major communities

lb.

Major occupations :

Cultivation
(Specify crop)
Fishing
Household Industry
(Specify industry) ..
Unskilled worker ..
Skilled worker.......
Petty traders

1
2

3
4
5
6

Shop owners
Businessmen/industrialists ..
Self employed professionals
Clerical/Salesmen
Supervisory level
Officers/Executive
- Junior
- Middle/Senior

7
8
9
10
11
12
13

SECTION 2 : INFRASTRUCTURAL FACILITIES

2.

Ascertain the dis ance of the following facilities from the village or starting point/ locality
If the d.stance is less than 1 Km , code as zero. If distance greater than 1 Km record exact
number of kms and the name of the nearest place


Q. No.

Facility

Nearest place where available
Place

2a

Post office___________

2b

Bus stop__________

2c

Railway station__________ ________

2d

Noon meal centre

2e

Primary School Upto Class (V)

2f

Secondary School Upto Calss (VHI)

2g

High School Upto Class (X)

2h

General provision store

Distance (Km)

6

SECTION 3 : HEALTH CARE FACILITIES

3.

Ascertain the distance of the following facilities from the village or starting point/locality.
If the distance is less than 1 Km, code as zero, if distance greater than 1 Km. record exact
number of Kms and the name of the nearest place.

Q. No

Nearest place where avail able

Health care facility

Place

3a

Primary Health Centre

3h

Sub centre

3c

Government Hospital

3d

Government Dispensory

3e

Private Hospital

3f

Private Practitioners - General

3g

Doctors who see pregnant women______

3h

Pharmacy/Chemist___________________

3i

Community Nutrition Centre/Anganwadi

21

Dais

4a.

Approximate population of the village

4b.

Coverage by VHN

Yes

1

No

2

4c.

Coverage by MPHW
- Male

Yes

1

No

2

Yes

1

No

2

Female
‘ 4d.

6

Distance (Km)

Number of dais in the village

:

Trained
Untrained

5a.

Nearest town

5c.

Distance in (Km) of the village from town

6a.

District headquarters

6b.

Distance in (Km) of village from district headquarters

5b. Population of town

2

Annexure - ii

PROJECT JIPMER

MATERNAL MORBIDITY STUDY

HOUSEHOLD ELIGIBILITY QUAIRE

SERIAL NUMBER

Survey Area/Centre

Sample type

: Urban

IF URBAN

: SP NO

Ward NO

Town

IF RURAL

: Village No

Village

Block

1

Rural

2

Interviewer

Sign :

Date:

Supervisor

Sign :

Date:

Check Details

: ACC 1

BC 2 Spot 3 By

Scrutiny Details

Field

: Yes 1

No 2

By

- Analysis

: Yes 1

No 2

By

Name of respondent
Name of head of household
Household No
Street/Hab Station

Pin

Interview Housewife
Vanakkam. My name is ________________ . I am from a Social Research Agency. From time
to time, we conduct surveys on mother and child health, employment opportunities, nutrition, health,
education and so on.

We are currently doing a study on mother and child health in various parts of the country. As per
research ethics, your name and the identity of your household will not be revealed to anybody.

So I would be grateful, if you could spend a few minutes of your time to answer a few questions.
1

la.

Could you tell me how many people live in your household. Please exclude guests and
servants when giving the answer.

lb.

Could you tell me how many married women live in your house ? By married women, I
mean all women who are married and who are living in your house. It does not matter
whether they are widowed or divorced now.

2.

I would like to know a few more details about each of these married women in your
household. Let us start with the youngest married women in your household.

ASK 2a TO 3c FOR ALL MARRIED WOMEN IN THE HOUSEHOLD.
RECORD Q.2a TO Q.3c IN TABLE BELOW

2a.

What is her name ?

2b.

Has (NAME) become pregnant in the last two years. She may still be pregnant or may have
delivered a baby or she may have had an abortion. The abortion may have been spontaneous
or induced. It does not matter.
Yes

1

2

No

IF NO CODED GO TO NEXT YOUNGEST MARRIED WOMAN.

2c.

ASK ONLY IF YES CODED IN Q.NO 2b. What was the outcome of (NAME)’s latest
conception. Is (NAME) currently pregnant or has she delivered a baby or had an abortion ?
PROBE TO FIND OUT OUTCOME OF LATEST CONCEPTION IN LAST 2 YEARS.

Conceived in last 2 years and delivered a baby

GO TO Q.3a

Conceived in last 2 years and had an abortion

GO TO Q.3b

Currently pregnant

3 ->

GO TO Q.3c

3a.

How old is the child today ?
GO TO Q.2a FOR NEXT YOUNGEST MARRIED WOMAN.

3b.

How many months back did (NAME) have the abortion ? RECORD NUMBER OF
MONTHS BACK WHEN (NAME) HAD ABORTION. GO TO Q.2a FOR NEXT
YOUNGEST MARRIED WOMAN.

3c.

How many months pregnant is (NAME) currently ?
GO TO Q.2a FOR NEXT YOUNGEST MARRIED WOMAN.

2

Name of
woman

SI
No

Out come

Age of
child

Mon. back
when had
abortion

Months
pregnant
currently

Q.2c

Q.3a

Q.3b

Q.3c

Conceived in
last 2 years 1

Yes 1

No 2

1

1

2

1

2

3

2

1

2

1

2

3

3

1

2

1

2

3

4

1

2

1

2

3

5

1

2

1

2

3

Q.2a

ENSURE THAT THE TOTAL NUMBER OF MARRIED WOMEN TALLIES WITH
THE WITH THE NUMBER IN Q.lb.
4.

ELIGIBILITY VERIFICATION RESPONDENT.
Respondent type

Eligibility criteria

1.

Conceived in last 2 years and delivered a
baby

If child aged 1 1/2 years or less

2.

Conceived in last 2 years and had an
abortion

If abortion was in the last 1 year or less

3.

Currently pregnant

All currently pregnant eligible.
TABLE

SI No

5.

Name of woman

Respondent Type

DQ.NO

1

1

2

3

2

1

2

3

3

1

2

3

4

1

2

3

5

1

2

3

Has any married women in your family died in the last 2 years ? IF NO STATED CODE
ZERO AND GO TO DETAILED QUAIRE IF THERE ARE ELIGIBLE
RESPONDENTS IN THE HOUSE. IF YES ASK : How many of them were menstruating
at the time of death. IF NONE STATED CODE ZERO AND GO TO DETAILED
QUAIRE IF THERE ARE ELIGIBLE RESPONDENTS IN THE HOUSE.

3

r

6.

ASK ONLY IF ZERO NOT CODED IN Q.5. I would like to talk to you in greater detail
about the married women who died in the last 2 years and who were still menstruating at the
time of death. Let us start with the youngest married woman who died in the last 2 years.

6a.

Could you tell me her name ?

6b.

What was (NAME)’s age at the time of death ?

6c.

What was the cause of death ? PROBE. ANYTHING ELSE ? CAN YOU EXPLAIN
FURTHER ? RECORD VERBATIM.
Age at the time of death

I. Name
Cause of the death :

Age at the time of death

2. Name
Cause of the death :

CHECK THAT IS THE NUMBER OF WOMEN FOR WHOM Q.6 a,b,c IS ASKED
TALLIES WITH THE NUMBER IN Q.5.

GO TO DETAILED QUAIRE (MOTHER’S STUDY)

4

Annexure - iii

PROJECT JIPMER

MATERNAL MORBIDITY STUDY MOTHER’S STUDY DQ No.

Name of town/village :
Block

Sample type

Urban 1

IF URBAN

PSU SI.No

Rural

2

Interviewer

Sign

Date

Supervisor

Sign

Date

Check Details
Scrutiny Details

ACC 1

BC 2

Spot 3

- Field

Yes 1

No 2

By

- Analysis

Yes 1

No 2

By

By

Name of respondent
Name of head of household
Household No/Identification
Street/Habitation
Pin

Respondent
type
(REFER :
PREGNANCY
HISTORY)

Currently not pregnant and has a child
born after....
Currently not pregnant and had an abortion
in the last 12 months or less
Currently pregnant and gave birth to a child
after....
Currently pregnant and did not give birth
to a child after

1

2
3

4

INTERVIEW OUTCOME :
Completed interview
1
Not completed interview because -

. House locked
. Woman away from home

2
3

Refusal by woman or family
Others (SPECIFY)


1

4
5

INTERVIEW ELIGIBLE WOMAN

Vanakkam. My name is
.___________________ . I am from a Social Research
Agency. From time to time, we conduct studies on mother and child health, nutrition etc. We are
currently conducting a study on child birth and pregnancy. As per research ethics, we will not reveal
your name or that of your family to anybody. I would be grateful if you could answer a few
questions.

SECTION 1 : PREGNANCY HISTORY

ASK ALL RESPONDENTS
Now, I would like to talk to you about all your children

101

How many children do you have who are alive today ?

IF NONE ALIVE TODAY CODE ZERO AND SKIP TO Q.lll. ELSE CONTINUE.
I would like to talk to you about each of these children in a little more detail. Let us start
with the youngest child.
Table 1.1
CHILDREN ALIVE TODAY

102

103

#1

#2

1
2

1
2

1
2

1
2

1
2

1
2

1
2

1
2

1
2

Was it a single or a multiple
birth ?
Single
Multiple

#3

#4

ASK ONLY IF MULTIPLE (2 CODED)
IN Q.102.
Are both the children alive
today or is one child alive and
the other dead ?
One child alive, other dead
Both alive

104

What is the name of the
child/child who is alive today ?

105

Is (NAME) a boy or a girl ?
Boy
Girl

1
2

1
2

2

1
2

106

What kind of delivery did you have when (NAME) was born ?
Was it a normal delivery or cesarean ? IF NORMAL ASK :
Did the baby come out by itself
or did they use forceps ?
Normal
Normal - forceps

107

1

1

1

Caesarean

2
3

2
3

2
3

Where did you have the delivery
when (NAME) was born ?
Public Health Centre /
Sub - centre
Private Hospital
Government Hospital
Private doctor/clinic
Home
Home of birth attendant
Others (specify)

1
2
3
4
5
6
8

1
2
3
4
5
6
8

1

1

2
3
4
5
6
8

2
3
4
5
6
8

1
2
3

Q

108

How many months pregnant were
you when (NAME) was born ?

109

What was your age when (NAME)
was born ?

110

When was (NAME) born ?

GO TO Q.102 FOR NEXT YOUNGEST CHILD ALIVE TODAY.

DOUBLE CHECK THAT THE NUMBER OF CHILDREN IN TABLE 1.1 ABOVE
TALLIES WITH Q.NO. 101.
Ill

vrc
gLVen birth t0 3 ChiId Wh° was dead even at 1116 time of birth or died later on
’ft u
°W many ch,Idren have y°u £iven
to, who were dead even at the time
of birth or died later on ?

on?
V.iiy. iscois COIN 1INUE.

°R DIED LATER ON’C0DE ZERO AND SKIP TO

™LLDREN WHO WERE BORN DEAD OR DIED LATER WHOSE TWIN IS ALIVF
rATFR ’ SHOULD NOT BE COVERED UNDER ‘CHILDREN BORN DEAD OR DIED

eNv°pn LTr like
t0 y°uU in.Sreater detail abou' each of your children who were dead
even at the time of birth or who died later on.
Let us talk about your last child who was dead at the time of birth or died later on.

3

TABLE 1.2

CHILDREN BORN DEAD OR DIED LATER

112.

113

114

#1

n

#3

#4

Dead at the time of birth
Died later on

1
2

1

2

1
2

1
2

Was it a single or multiple birth ?
Single
Multiple

1
2

1
2

1
2

1
2

1
2
3

1
2
3

1
2
3

1
2
3

1
2
3
4

1
2
3
4
5
6
8

1
2
3
4
5
6
8

1
2
3
4
5
6
8

Was the child dead at the time of birth
or died later on ?

What kind of delivery did you have when
■ this child was born ?
Was it a normal delivery or
caesarean ? IF NORMAL ASK :
Did the baby come out by itself
or did they use forceps ?
Normal

Normal - forceps
Caesarean

115

Where did you have the delivery when this
child was born ?
Public Health Centre / Sub - centre
Private Hospital
Government Hospital
Private doctor/clinic
Home
Home of birth attendant
Others (Specify)

116

How many months pregnant were
you when this child was born ?

117

What was your age when this
child was born ?

118

When was this child born ?

5

6
8

GO TO Q.112 FOR THE NEXT YOUNGEST CHILD BORN DEAD.

DOUBLE CHECK THAT THE NUMBER OF CHILDREN IN TABLE 1.2 TALLIES
WITH Q.lll.
119

Have you ever had any abortions so far ? IF YES ASK : How many abortions have you had
so far ?

4

IF NONE CODE ZERO AND SKIP TO Q. 123 ELSE CONTINUE.

Now, I would like to talk to you in detail about each of your abortions. Let us start with
your latest abortion.

120

1

ABORTION
#2
#3

1

1

1

1

2

2

2

2

#4

Was it a spontaneous or induced abortion
Spontaneous
Induced

121

How many months pregnant were
you at the time of abortion ?

122

How old were you at the time of abortion

GO TO Q.I20 FOR NEXT LAST ABORTION. DOUBLE CHECK THAT NUMBER OF
ABORTIONS IN TABLE ABOVE TALLIES WITH Q.119.

123

Are you currently pregnant ?
Yes
No.

124

1
2

ASK ONLY IF YES (1 CODED IN Q.123) How many months pregnant are you
currently ? RECORD COMPLETED MONTHS OF PREGNANCY.

fj

5

SECTION 2 : AWARENESS AND UTILISATION OF HEALTH SERVICES

ASK ALL RESPONDENTS
201

Now, I would like to talk to you about where you go for treatment if somebody falls ill, for
immunisation and so on.

FOR 0.201 TO Q.203 PROBE FOR LOCATION/NAME OF TOWN AND CODE
APPROPRIATELY IN TABLE 2.1. ENTER NAME OFpLACES
LOCATION/TOWN
IN
MENTI0NED
SPONTANEOUSLY.

In the last 1 year, has anybody in your house fallen ill. IF NO STATED CODE NOT
APPLICABLE (CODE 0) IF YES STATED ASK : Where
all did you go for treatment ?
202

ASK ONLY IF RESPONDENT HAS A CHILD BELOW 3 YEARS (REFER
PREGNANCY HISTORY). Have you got your children immunised ? IF NO STATED
NOT APPLICABLE (CODE 0). IF YES STATED ASK : Where did you get your children
immunised ?

203

Where do women in your house go for treatment of female
diseases ? By female diseases,'i mean white discharge, stomach pain during periods, problems
at menopause etc.,
TABLE 2.1
Female
Immunisation Diseases
Illness
(Q.202)
(Q. 203)
(Q.201)
Source of treatment
.)

1

1

1

Private Hospital

)

2

2

2

Govt. Hospital

)

3

3

3

Private doctor/clinic

)

4

4

4

Public Health Centre

(

Pharmacy

(

.)

5

5

5

Dai

(

)

6

6

6

Traditional
practitioner

(

.)

7

7

7

Others (Specify)

(.

)

8

8

8

Don’t know

(

)

9

9

9

Not applicable/don’t
go anywhere

(

)

0

0

0

ASK Q.204 TO Q.206 IF LOCATIONS 1 TO 5 CODED IN Q.201 TO 203. ELSE
SKIP TO Q.301.
Say, "I would like to know a few more details about the places you had gone for treatment
of illness/immunisation/female diseases."

6

TABLE 2.2
Source of Treatment
Public Health
Centre
204

1

2

Pharmacy

1
2

1
2

1
2

1
2

1
2
3
4
5
6
7
8

1
2
3
4
5
6
7
8

1

1

2
3
4
5
6
7
8

2
3
4
5
6
7
8

How do you usually travel
to the (source) ?
On foot
Auto Rickshaw
Bicycle
Bullock Cart
Bus
Taxi
Cycle Rickshaw
Others (SPECIFY)

206

Private doctor/
clinic

Where is the (Source)
located - in the same village
I town or in another
village?
Same village / town
Another village / town

205

Private hospital Govt. Hospital

1

2
3
4
5
6
7
8

How long does it take using
your usual mode of
transport to reach (source) ?
Minutes
Hours

Don’t know/Don’t
remember

1

1

1

1

1

2

2

2

2

2

9999

9999

IF LESS THAN TWO
HOURS RECORD
MINUTES. ELSE
RECORD HOURS

7

9999

9999

9999

SECTION 3: PREGNANCY IDENTIFICATION AND ANTE NATAL CARE

ASK ALL RESPONDENTS.
ALL QUESTIONS IN THIS SECTION TO BE ASKED OF ALL RESPONDENTS FOR THEIR
LAST PREGNANCY.

IF RESPONDENT TYPE 3, (i.e) CURRENTLY PREGNANT AND GAVE BIRTH TO A
CHILD AFTER, CIRCLE FOR WHICH PREGNANCY THIS
SECTION ASKED :
Current Pregnancy
Last Pregnancy

1
2

I would now like to talk to you in detail about your last/current pregnancy.

301

Once you missed your periods in your last/current pregnancy did you consult anybody to
confirm that you were pregnant. IF YES STATED ASK : Whom did you consult to
confirm that you were pregnant.

1
2
3
4
5
6
0

Doctor
Nurse
MPHW (male)
MPHW (female)
Mother/Mother-in-law/relative
Dai
Did not consult anybody
302

ASK ONLY IF 1 TO 4 CODED IN Q.301. How many months pregnant were you at the
time of confirmation ?


303

During your last/current pregnancy have you taken any tetanus immunisation injections ?
1
2

Yes
No......................................................
IF NO (2 CODED) SKIP TO 306.

304

How many tetanus immunisation injections did you take ?

305

In which months of pregnancy did you take the tetanus immunisation injections and where
did you take the tetanus immunisation injections ? PROBE FOR LOCATION/TOWN
AND CODE APPROPRIATELY.

SI. No

Place

Month of pregnancy

1

1

2

3

4

5

2

1

2

3

4

5

8

CODES FOR PLACE :
1
2
3
4

Public Health centre/Sub centre
Private Hospital
Govt. Hospital
Private doctor/clinic
...........................................
Home/Village.......................................................................................................................

5

NO. OF MONTHS IN Q.305 SHOULD TALLY WITH NO. OF INJECTIONS IN
Q.304.

SKIP TO Q.NO.307.
306

Why did you not take the tetanus immunisation injections? LET RESPONDENT ANSWER
SPONTANEOUSLY AND CIRCLE UPTO 3 ANSWERS IF GIVEN.

Didn’t know about the injections/where to get the
injections
Too far away
Was not open at times when I could go there
Not welcoming /attitude of doctors nurses not good
Service not good/no medicine
Always felt well/did not think it was necessary
No money
Too busy at home
No transportation
No one to watch children at home
Health worker/nurse did not come
Husband/family would not allow
Abortion occurred before taking the injections
Others (Specify)
307

5

6
7
8
9
10
11
12
14
15

During your last/current pregnancy have you swallowed (eaten) any iron, folic acid or vitamin
tablets ? You may have bought these tablets yourself or a doctor or nurse may have given
them free to you.
1
2

Eaten
Not eaten
308

1
2
3
4

During your last/this pregnancy have you ever shown yourself to a doctor or nurse to check
if you and the child were okay ? IF YES STATED ASK : How many times have you shown
yourself to a doctor ?

RECORD EXACT NUMBER. IF 6 OR 6+ RECORD AS 6. 0 = NOT SHOWN. 99
= DON’T KNOW/DOESN’T REMEMBER. IF NOT SHOWN (0 CODED) SKIP TO
Q.316.
309

At which month of pregnancy did you first show yourself to a doctor or nurse to check if
you and the child were okay?

310

Who advised you to show yourself to the doctor or nurse ?

Husband
Mother/M other-In-Law
Relative

1
2
3
4
5

Neighbour

Dai

9

Nurse/doctors
Nobody (I myself decided)

6

7
8

Others (SPECIFY)
311

I

I

I

Where did you show yourself to the doctor or nurse ?
PROBE LOCATION/TOWN AND CODE APPROPRIATELY.
LOCATION/TOWN ().

ENTER NAME OF

CHECK THAT NUMBER OF PLACES CODED IS NOT MORE THAN NUMBER OF
CHECK-UPS.
Public Health Centre/Sub Centre (
Private Hospital
(_
Govt. Hospital
C
Private Doctor/Clinic
c
Home Visit By Health Worker
c
Others (SPECIFY)
c

J
J
J
J
J
J

CHECK IF LOCATIONS RECORDED IN CODES 1 TO 4 IN Q.311 SAME AS IN
CODES 1 TO 4 IN Q.201 TO Q.203. IF ALL LOCATIONS SAME, SKIP TO
SECTION 4. ELSE ASK Q.312 TO Q.314 ONLY FOR THE DIFFERENT
LOCATIONS.
Say, "I would like to know a few details about the place where you had gone for showing
yourself to the doctor/nurse."

10

TABLE 2.2
Source of Treatment
Public Health
Centre/Subcentre
312

1
2

1

2

1
2

1
2

1
2
3
4
5

1
2
3
4
5

How do you usually travel to
the (source) ?
On foot

1
2
3
4
5

Auto RickshawBicycle
Animal-drawn cart
Bus
Taxi
Cycle Rickshaw
Others (SPECIFY)
314

Private
doctor/ clinic

Where is the (Source) located in the same village / town or in
another village?

Same village / town
Another village / town
313

Private hospital Govt. Hospital

1
2
3
4

5

6

6
7

7
8

8

6

6

7
8

7
8

How long does it take using
your usual mode of transport to
reach (source) ?

Minutes
Hours

Don’t know/Don’t remember

1

1

1

1

2

2

2

2

9999

9999

9999

9999

IF LESS THAN TWO
HOURS RECORD MINUTES.
ELSE RECORD HOURS

315.

316.

Some women we spoke to told us of a few check-ups that the doctors/nurses did when they
had shown themselves to the doctor or nurse ? As I read out the check-ups to you, I would
like you to tell me if this was done for you or not ?

Check-ups

Yes

No

a) BP Checked
b) Urine examined
c) Internal examination
d) Stomach measured with ai tape/listened to baby’s heart beat
e) Weight checked
f) Scanned/seen baby in TV

1
1
1
1
1
1

2
2
2
2
2
2

not sh°w yourself to a doctor/nurse during your last/current pregnancy ? LET
RESPONDENT ANSWER SPONTANEOUSLY AND CIRCLE UP TO THREE
ANSWERS.

11

*

Did not know where to go / where it is
Too far away
Was not open at times when I could to there
Not welcoming / attitude of doctors nurses not good
Service not good / no medicine
Always felt well / did not think it was necessary
No money
Too busy at home
No transportation
No one to watch children at home
Husband / family would not allow
Abortion occurred before showing to the doctor
.
Others (Specify)

7

12

1

2
3
4

5
6

7
8
9
10
12
14

15

SECTION 4 - ANTE PARTUM MORBIDITIES
ASK ALL RESPONDENTS.

401.

Did you have any illnesses or problems during your last/current pregnancy ? IF YES : What
were all the problems or illnesses you had during your last/current pregnancy.
CIRCLE UNDER UNAIDED IN TABLE 4.1 THE ILLNESSES MENTIONED.

402.

I have with me a list of problems/illnesses mentioned by various people we have spoken to
as problems/illnesses that they have had during their pregnancy. As I read out each
problem/illness to you, I would like you to tell me whether you have experienced this
problem during your last/current pregnancy ?

ASK FOR ALL PROBLEMS NOT CIRCLED IN Q.401 IN TABLE 4.1 : Did you have
(PROBLEM/ILLNESS) during your last/current pregnancy ?

IF YES STATED, CIRCLE THE PROBLEM UNDER AIDED IN TABLE 4.1.
CHECK : IF NO PROBLEM CIRCLED UNDER Q.401 (UNAIDED) OR Q.402
(AIDED) IN TABLE 4.1. SKIP TO Q.426.
TABLE 4.1

Problem/Illness

Unaided Aided
Q.401 Q.402

Problem/Illness

Unaided Aided
Q.401 Q.402

1 Swelling of hands and feet

1

1

9 Severe vomiting where
treatment had to be taken

1

1

2 Blurred vision

1

1

10 Tuberculosis

1

1

3 Giddiness

1

1

11 Malaria

1

1

4 Fits

1

1

12 Hepatitis/Jaundice

1

1

5 Urinary problem

1

1

13 Heart disease

1

1

6 Varicose veins

1

1

14 Diabetes

1

1

7 Fever more than 3 days

1

1

15 Others (Specify)

1

1

8 High blood pressure

1

1

ASK Q.403 AND Q.404 ONLY IF SWELLING CODED IN Q.401 OR Q.402.
403.

Let us talk about the swelling you had experienced in your last/current pregnancy.
During the course of this pregnancy, in which all parts of the body did you experience
swelling ? CIRCLE APPROPRIATELY.
MULTIPLE CODING PERMISSIBLE.

Feet only
Above feet upto knees
Above calves below waist ....
Palms and back of palms
Above palms below shoulder
Neck
Face

404.

1

2
3
4
5
6
7

Till which month of pregnancy did you have swelling ? PROBE FOR COMPLETED
MONTHS OF PREGNANCY

Mon

13

• ASK Q.405, 406 ONLY IF BLURRED VISION CIRCLED IN Q.401 OR Q.402.

403.

Let us talk about the blurred vision you had during your last/current pregnancy. Did
you have blurred vision only after dark or during day time also.



Only after dark
During day time and after dark

404.

1
2

Till which month of pregnancy did you have blurred vision ? PROBE FOR
COMPLETED MONTHS OF PREGNANCY
Mon

407.

ASK ONLY IF GIDDINESS CIRCLED IN Q.401 OR Q.402. Let us talk about the
giddiness you had during your last/current pregnancy. Till which month of pregnancy did
you have giddiness ?
_______

Mon
PROBE FOR COMPLETED MONTHS OF PREGNANCY.
ASK Q.408 AND Q.409 ONLY IF FITS CIRCLED IN Q.401 OR Q.402.

408.

Let us talk about the fits you had during your last/current pregnancy. Did you get fits
only during this pregnancy or have you also got fits even when you were not pregnant ?
Fits only during this pregnancy ................................................................................
Fits during pregnancy and when not pregnant ......................................................

409.

I
2

Did you have swelling in any part of your body at the time of fits ?

1
2

Yes
No
ASK Q.4I0 TO 412 ONLY IF URINARY PROBLEM CIRCLED IN Q.401 OR Q.402.

410.

Let us talk about the urinary problems you had during your last/current pregnancy ? Did
you have difficulty in passing urine during your last/current pregnancy ?

1
2

Yes
No

411.

Did you have a painful or burning sensation when passing urine ?

1
2

Yes
No
412.

ASK ONLY IF YES ( 1 CODED ) IN Q.411.
Did you also have fever at the time you had painful, burning sensation when passing
urine ?
Yes
No...................................................................................................................

ASK Q.413 TO Q.415 ONLY IF VARICOSE VEINS CIRCLED IN Q.401 OR Q.402.
14

1
2

413.

Let us talk about the varicose veins you had during your last/current pregnancy. In
which parts of your body did you have varicose veins ?

Legs upto knee
Full legs below waist

414.

1
2

Until which month of pregnancy did you have blurred vision ?

PROBE FOR COMPLETED MONTHS OF PREGNANCY.

415.

Did you have varicose veins before your last current pregnancy ?
Yes
No
................................................................

1
2

416.

1N.k9:4?LOR Q-402- In this Pregnancy, how many
times did you get fever that lasted more than 3 days.

417.

ASK ONLY IF VOMITING CIRCLED IN Q.401 OR Q.402.

Until which month of pregnancy did you vomit ?

Mon
PROBE FOR COMPLETED MONTHS OF PREGNANCY.

--------------

agai^orrk^^^

£ACED ,n table 42 ASK Q'418 TO

I

418.

Now I would like to talk to you in detail about ( PROBLEM ). At which month of
pregnancy did you first experience ( PROBLEM ) ?
PROBE TO ASCERTAIN COMPLETED MONTHS OF PREGNANCY AT THAT
TIME.

419.

B^JOzVJronsult any°ne or take any treatment for ( PROBLEM ). IF NO STATED CODE
1 tviSA 1 ItllSiN 1 o.

Allopathy

- Public Health Centre
Private Hospital
Government Hospital
Private doctor/Clinic
Pharmacy..........................................

.

,
oo

Ayurvedic/Homeo/Unani/Sidha
Home remedies - Elders/Dai/Traditional Practitioners
Religious/Superstitious Practices
Not consulted
Don’t remember/Don’t know ......7. ...........7/.^

................... *
....................

1
2
3
4
5

6
7
8
000
999

COw!krE'FELSEOSKirC?OQD<MS 0NE 0F THE ™EATMENTS IN Q.419 -

15

j

420.

In which month did you first visit the (REFER Q.419) ? PROBE TO ASCERTAIN
COMPLETED MONTHS OF PREGNANCY AT THAT TIME.

421.

Who advised you to go to (REFER Q.419).

1
2
3
4
5
6
7
8

Husband
Mother/Mother-in-law
........................
Relative
Neighbour
Dai
......................... .............
Nurse/Doctor
Nobody ( I myself decided )
Others (SPECIFY).

422.

4

When you went to the (REFER Q.419), how did you travel ?

IF MORE THAN ONE VISIT HAS BEEN MADE EITHER TO THE SAME PLACE OR
DIFFERENT PLACES FOR THIS PROBLEM AND DIFFERENT MODES OF
TRAVEL USED EACH TIME, RECORD MODE OF TRAVEL FOR THE FIRST
TIME, TO THE FIRST PLACE.
1
On foot
2
Auto rickshaw
3
Bicycle
4
Bullock cart
5
Bus
6
Taxi
7
Cycle rickshaw ....
8
Others (SPECIFY)

GO TO Q.424.
423.

Why did you not go to a doctor/hospital for treatment of (PROBLEM) ?
RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.

Expensive/Could not afford
Not good/Don’t treat well/Attitude of doctors not good
Too far away/No transportation ....................................
Timing/Other inconveniences like no one at home
Members of family would not allow
....................
Did not know where to go...............................................
Preferred to take treatment from traditional sources ..
Didn’t think it was necessary
Others (Specify)

LET

1
2
3
4
5
6
7
8
9

424.

How much money did you spend on the treatment of (PROBLEM) PROBE FOR TOTAL
COST OF TREATMENT INCLUDING DOCTOR’S/PRACTITIONER’S FEES,
MEDICINE, TRAVEL COST ...

425.

When you had (PROBLEM), at that time, did you think it was serious or not ?

1
2

Serious
Not Serious
GO TO Q.418 FOR NEXT PROBLEM WITH TICK MARK.

DOUBLE CHECK : Q.418 TO Q.425 ASKED FOR ALL PROBLEMS AGAINST
WHICH TICK MARKS HAVE BEEN PLACED IN TABLE 4.2

16

Probleui/Illness

First
exp.lb th
Q.418

Treatment
Q.419

Advised by
Q.421

Reasons
not gone
Q.423

Travel
mode
Q.422

1

1

2

2

3

Serious
Q.425

Money
Q.424

Yes No
1

2

3

1

2

1

1

1

2

2

2

3

3

1

2

1

I

!

2

2

2

3

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

2

1 Swelling of hands
and feet

2 Blurred vision

3 Giddiness

4 Fils

5 Urinary problem

Mouth
doctor first
visited Q.420

o

3

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

'2

3

3

1

2

1

1

1

2

2

2

3

3

r

2

i

i

i

2

2

2

3

3

1

2

6 Varicose veins

7 Fever more than 3
days

8 High blood pressure

9 Severe vomiling
where (realmenl had
to be taken

10 Tuberculosis

17

Problem/Illness

ll Malaria

First
exp. m th
Q.418

Treatment
Q.419

Month
Advised by
doctor first
Q.421
visited Q.420

Travel
mode
Q.422

Reasons not
gone
Q.423

1

1

2

2

3

Money
Q.424

Serious
Q.425

Yes No
1

2

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

2

3

3

1

2

1

1

1

2

2

2

3

3

1

2

f

12 Hepatiiis/Jaundlce

I
13 Heart disease

14 Diabetes

75 Others (Specify)

426.

Apart from the various problems or illnesses that we discussed so far did you experience any
other problem or illness during your last/current pregnancy ?

1
2

Yes
No.
427.

IF YES : (1 CODED) IN Q.426 : Can you tell me what all other problems or illnesses you
experienced during this pregnancy ? PROBE : ANYTHING ELSE ? CAN YOU
EXPLAIN FURTHER.
PROBE FOR TWINS/BABY IN WRONG POSITION.
RECORD THE PROBLEMS IN TABLE 4. 2 AND ASK Q.418 TO Q.425.

CHECK RESPONDENT TYPE.

RESPONDENT TYPE
Currently not pregnant and has a child born after
Currently not pregnant and had an abortion in the last 12 months or less
Currently pregnant and also had a child born after
Currently pregnant and did not have a child born after

18

GO TO
Section 5
Section 7
Section 5

Q.428

428.

Where do vou plan to have your delivery ?
AND CODE APPROPRIATELY.

PROBE : LOCATION/NAME OF TOWN

1
2
3
4
5
6
8
9

Public health centre .
Private hospital
Government hospital
Private doctor/clinic .

Home
Have not decided ....
Others (Specify)
Don’t know
GO TO SECTION 9.



l

19

SECTION 5 - DELIVERY
*

ASK RESPONDENT TYPE 1,3 (i.e) THOSE WHO GAVE BIRTH TO A CHILD
AFTER

501.

Where did you originally plan to have your delivery ?

1
2

Public health centre
Private hospital.............
Government hospital....
Private doctor/clinic.....
Home
.
Home of birth attendant
Others (SPECIFY)

F
502.

3
4
5

6
8

Where did you tell me, you finally had your delivery ?
1

Public health centre
Private hospital
Government hospital ...
Private doctor/clinic
Home
Home of birth attendant
Others (SPECIFY)
503.

2
3
4

5
6

8

Who attended on you at the time of delivery ?
1

No one
Relative
Untrained Dai
Trained Dai
MPHW (Female) .
Doctor
Nurse
Others (SPECIFY)

2
3

4
5
6

7
8

IF PLACE ORIGINALLY PLANNED (Q.501) AND PLACE OF DELIVERY (Q.502)
DIFFERENT, ASK Q.504, 505. ELSE SKIP TO Q.506.

504.

You said you had originally planned to have your delivery at
(REFER 501), but
you finally had it at
(REFER 502). What was the reason for changing
the place of delivery ?
LET RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.

Prolonged labour/Could not bear pain
Dai/Mother said can’t do delivery at home

505.

1
2

Excessive bleeding
Even after sac burst, delivery did not take place
Others (SPECIFY)
.........

3
4

Who was the person most responsible for having the delivery at
(REFER Q.No 502).

9

20

8

506.

Husband
Mother/Mother-in-law
Elder female relatives ,

1
2
3

Neighbours

4

Was any tiling done to induce labour ? IF YES STATED ASK : What was done to induce
labour ? MULTIPLE CODING PERMISSIBLE.

HOME METHODS

Oil massage
Hot water poured below waist
Kashayam
Others (SPECIFY)

11
12
13
14

HOSPITAL METHODS
Injection
Enema
Drips
Others (SPECIFY)
Nothing done

507.

21
22
23
24
00

What type of delivery did you tell me you had in your last delivery. Was it natural or
cesarean ? IF NATURAL STATED ASK : Did the baby come out by itself or did the
doctor or nurse use forceps ?

Normal
.
Normal - forceps
Cesarean
508.

3

Did you have a tear in your vagina when the baby was coming out ? IF YES STATED
ASK: Was the tear natural or did the doctor or nurse make the tear ?
Natural tear,
No tear
.........<
Tear
made

509.

1
2
3

Was anything done to close the tear ? IF YES STATED ASK : What was done to close the
tear ?
Stitched
Ash
Stuffed with cloth ..
Nothing specifically
Don’t Know

510.

1
2
3
0
9

Was the placenta fully removed ?
Yes
No
Don’t know

511.

1
2
> Q.512

.........

1
2
3

How was the umbilical cord cut ?

Neu Blade
Kitchen Knife
Scissors
Sickle/Aruvamanai/chisel
Others (SPECIFY)
Don’t Know
.

1

2
3
4
8
9

21

512.
r

Various women we spoke to told us about various problems they had at the time of delivery.
As I read out the problems, I would like you to tell me if you had the problem in your
last/current pregnancy at the time of delivery and if you think the problem at the time of
delivery is serious or not serious

For each problem in table below ask 512a.

Have you experienced (PROBLEM) in your last pregnancy at the time of delivery ?

512b.

ASK ONLY IF EXPERIENCED : When you had (PROBLEM), at that time did you think
it was serious ?
TABLE 5.1

Problem

Experienced in last
pregnancy
Q.512a

Serious
Q.512b

Yes

No

Yes

No

1

Labour more than 18 hours

1

2

1

2

2

Excessive bleeding (More than 3 sarees
stained)

1

2

I

2

3

Sac hurst and even after 5 hours child
was not born

1

2

1

2

4

Sac burst and the fluid was greenish
coloured

1

2

I

2

5

Fainted during labour

1

2

1

2

6

Fits or convulsions

1

2

1

2

7

Baby was in breech position/not in
normal position

1

2

1

2

8

Placenta was down

1

2

1

2

9

Twins/multiple births

1

2

1

2

22

SECTION 6 - POST PARTUM MORBIDITIES
ASK RESPONDENT TYPE 1, 3 (i.e) THOSE WHO GAVE BIRTH TO A CHILD AFTER
I would now' like to talk to you about the period after your last delivery (ENSURE THAT THE
RESPONDENT UNDERSTANDS THAT WE ARE TALKING ABOUT PERIOD AFTER LAST
DELIVERY).

SHORT TERM SEQUELAE
601.

Did you have any problems or illnesses in the first one or two months after your last
delivery? PROBE: ANYTHING ELSE ? CAN YOU EXPLAIN FURTHER ? CIRCLE
UNDER UNAIDED IN TABLE 6.1, ALL THE PROBLEMS MENTIONED.

602.

I have with me a list of problems or illnesses mentioned by various women we have spoken
to as problems or illnesses ±at had occurred to them in the first one or two months after the
delivery. As I read out each problem I would like you to tell me if you have experienced it
after your last delivery.

ASK FOR ALL PROBLEMS NOT CIRCLED IN Q.601 IN TABLE 6.1 : Did you have
(PROBLEM) in the first one or two months after your last delivery ?
IF YES STATED, CIRCLE THE PROBLEM UNDER AIDED.
CHECK : IF NO PROBLEM CIRCLED UNDER Q.601 (UNAIDED) AND Q.602
(AIDED), IN TABLE 6.1, SKIP TO Q.610.
FOR EACH PROBLEM CIRCLED UNDER Q.60I OR Q.602 ASK Q.603 TO Q.609.
RECORD IN TABLE 6.1.
603.

Now- I would like to talk to you in detail about (PROBLEM). How many days after your
delivery did you first experience (PROBLEM) ?

604.

Did you consult anyone or take any treatment for (PROBLEM). IF NO STATED, CODE
NOT CONSULTED (000). IF YES STATED ASK : Whom all did you consult and what
all treatments did you take ? LET RESPONDENT ANSWER AND CODE UPTO 3
TREATMENTS.
Allopathy

Public Health Centre
Private Hospital
Government Hospital

1
2
3
4
' 5

Private doctor/Clinic

Pharmacy
Ayurvedic/Homeo/Unani/Sidha
Home remedies - Elders/Dai/Traditional Practitioners
Religious/Superstitious Practices
Not consulted
Don’t remember/Don’t know
.....

6
7
8
000
999

cSMK1PC?SqD60A7S 0NE Or ™E TREA™ENTS IN O«» ■

23

*

i

605.
!

Who advised you to go to (REFER Q.604) ?

Husband

.....................

1

Mother/Mother-in-law

2
3
4
5

Relative
Neighbour
....... .............
Dai
Nurse/Doctor
Nobody (I myself decided )
Others (SPECIFY)
606.

6

7
8

When you went to the (REFER Q.604), how did you travel ?

IF MORE THAN ONE VISIT HAS BEEN MADE EITHER TO THE SAME PLACE OR
DIFFERENT PLACES VISITED PROBLEM AND DIFFERENT MODES OF TRAVEL
USED EACH TIME, RECORD MODE OF TRAVEL FOR THE FIRST TIME TO THE
FIRST PLACE.
On foot
Auto rickshaw
Bicycle
Bullock cart
Bus
Taxi
Cycle rickshaw ...
Others (SPECIFY)

1
2
3
4
5
6

7
8

GO TO Q.608.

607.

Why did you not go to a doctor/hospital for treatment of (PROBLEM) ?
RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.
Expensive/Could not afford
Not good/Don’t treat well/Attitude of doctors not good
Too far away/No transportation
Timing/Other inconveniences like no one at home
Members of family would not allow
Did not know where to go
Preferred to take treatment from traditional sources ..
Didn’t think it was necessary
Others (Specify)

I

LET

1
2
3
4
5
6
7
8
9

608.

How much money did you spend on the treatment of (PROBLEM) ? PROBE FOR
TOTAL COST OF TREATMENT INCLUDING DOCTOR’S/ PRACTITIONER’S
FEES, MEDICINE, TRAVEL COST ...

609.

When you had (PROBLEM), at that time did you think it was serious or not ?

1
2

Yes
No

GO TO Q.603 FOR NEXT PROBLEM CIRCLED IN Q.601 OR Q.602.

DOUBLE CHECK : Q.603 TO Q.609 ASKED FOR ALL PROBLEMS CIRCLED IN
Q.601 OR Q.602.


24

TABLE 6.1

Problem/lllne.s

L'naided Aided
Q.601
Q.602

1 Pus formation
in tear

1

3 Loss of
consciousness
for more than
15 minutes

5 Painful, burning
feeling when
urinating

6 Changes in
mental make-up

9 Breast abscess

10 Excess bleeding

11 Others

(Specify)

1

1

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

2

2

3

3

1

1

1

1

3

1

1

8 Discharge that
smells

2

1

1

7 Fits /
convulsions

2

1

1

1

1

1

Reasons
not gone
Q.607

Travel
mode
Q.606

1

1

1

Advised by
Q.605

1

1

1

4 Pain in lower
abdomen

Treatment
Q.6O4

1

1

2 Fever >3 days

Days after
delivery
Q.603

25

Serious
Q.609

Money
Q.608

Yes

No

1

2

1

2

1

2

1

2

1

2

1

2

1

2

1

2

1

2

1

2

1

2

610.

Apart from the problems we discussed now, did you have any other problem during the
1 or 2 months after your last delivery ?
Yes
No.

1
2

611.

IF YES (1 CODED IN Q.610) ASK : Can you tell me what all other problems or illnesses
you had ? PROBE : ANYTHING ELSE ? CAN YOU EXPLAIN FURTHER ?
RECORD THE PROBLEM IN TABLE 6.1 AND ASK Q.603 TO Q.609.

612.

Have you resumed normal work after delivery ?

I

Yes
No.

1
2

IF NO GO TO Q.614.

613.

When did you resume work after delivery. IF LESS THAN 3 MONTHS CODE DAYS,
ELSE CODE MONTHS.

Days

1

Months

2

GO TO Q.615
614.

When do you plan to resume work ? IF LESS THAN 3 MONTHS AFTER DELIVERY
CODE DAYS, ELSE CODE MONTHS.

Days

1

Months

2

615.

I have with me a list of problems mentioned by various women we have spoken to as
problems that had occurred to them after delivery. As I read out each problem I would like
you to tell me if you have experienced any of these problems.

615a.


ASK FOR FIRST PROBLEM IN TABLE 6.2 AND RECORD IN TABLE 6.2 Have you
had (PROBLEM) after delivery ?

i

1
2

Yes
No.

I
i

IF NO GO TO Q.615a FOR NEXT PROBLEM IN TABLE 6.2. IF YES ASK Q.616
TO Q.623.
CHECK : IF LAST DELIVERY WAS FIRST CHILD (REFER PREGNANCY
HISTORY) DO NOT ASK Q.616 AND UNDER 616 CODE AS NOT APPLICABLE=0.

616.
/■

When did you have this (PROBLEM) - after this child or an earlier child ?

1
2
0

This child
Earlier child..
Not applicable

26

617.

Did you consult anyone or take any treatment for (PROBLEM). IF NO STATED CODE
NOT CONSULTED (000) IF YES STATED ASK : Whom all did you consult and what
all treatments did you take ? LET RESPONDENT ANSWER AND CODE UPTO 3
TREATMENTS.
Allopathy

Public Health Centre
Private Hospital
Government Hospital
Private doctor/Clinic
Pharmacy

1

2
3
4

5

Ayurvedic/Homeo/Unani/Sidha
Home remedies - Elders/Dai/Traditiona! Practitioners
Religious/Superstitious Practices
Not consulted
.
Don’t remem her/Don’t know

6

7
8
000
999

CHECK : IF 1 TO 4 CODED IN Q.617 - CONTINUE. ELSE SKIP TO Q.621.
618.

You said you had gone to (REFER Q.617) for treatment. Did they perform an operation
or did they only give you medicine and injections ?

Operation
Medicine & injections only
619.

1
2

Who advised you to go to (REFER Q.617) ?

Husband
Mother/Mother-in-law
Relative

1
2
3
4

Neighbour

Dai
Nurse/Doctor
Nobody ( I myself decided )
.
Others (SPECIFY)
____________
620.

5
6

7
8

When you went to the (REFER Q.617) how did you travel ?
On foot

1
2

Auto rickshaw
Bicycle
Bullock cart
Bus
Taxi
Cycle rickshaw
Others (SPECIFY)

3
4
5
6

7
8

GO TO Q.622.
621.

Why did you not go to a ('doctor/hospital for treatment of (PROBLEM) ? LET
RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.

Expensive/Could not afford
.................
Not good/Don’t treat well/Attitude of doctors not good
Too far away/No transportation
Timing/Other inconveniences like no one at home
Members of family would not allow
Did not know where to go..................... ».<*...?
Preferred to take treatment from traditional sources ..
Didn’t think it was necessary
Others (Specify) ...................................... .....7......;

27

1

2
3
4

5
6
7
8
9

e

622.

How much money did you spend on the treatment of (PROBLEM) ? PROBE FOR
TOTAL COST OF TREATMENT INCLUDING DOCTOR’S/ PRACTITIONER’S
FEES, MEDICINE, TRAVEL COST ...

623.

Do you still have this problem ?
Yes
No.

*

1
2

GO TO Q. 615 a FOR NEXT PROBLEM IN TABLE 6.2.

DOUBLE CHECK THAT Q.615a to Q.623 HAS BEEN ASKED FOR ALL PROBLEMS
IN TABLE 6.2.
TABLE 6.2

Problem

Which child
Type
Experienced
Treatments
This 1
Oper 1
Yes 1 No 2
taken
Earlier 2 NA
Med inj 2
Q.615a
Q.617
0 Q.616
Q.618

3 Passing stools
through the vaginal
canal

1

1

1

1

4 Piles

Reasons Money
not gone
Q.621 Q.622

2

2

2

2

1

1

1

1

2

2

2

2

0

0

0

0

Yes No

2

1

2

2

3

3

1

1

2

1

2

2

3

3

1

1

2

1

2

2

3

3

1

1

2

1

2

2

2

1

2

1

2

1

2

ASK ONLY IF PILES EXPERIENCED IN TABLE 6.2: Did you have piles only during
last pregnancy or before ?
1
2

Have you started having sex with your husband after your last delivery ?
Yes
No ,

8

1

3

Only during last pregnancy
Before

625.

Problem
still
Q.623

1

3

624.

Travel
mode
Q.620

1

1 Feeling of heaviness
in the abdomen or
feeling of uterus
coming down

2 Experienced problem
- passing of urine such
as passing urine all
the time or when
coughing, sneezing

Advised
by
Q.619

1

2 —> Section 8

28

626.

How old was your child when you started having sex with your
husband ?
IF LESS THAN 3 MONTHS CODE DAYS. ELSE CODE MONTHS.
Days

1

Months

2

Not applicable .. 000
627.

Some women we have spoken to told us that they had pain when having sex after their
delivery.
As I read out what they said, could you tell me which statement best describes your case.
True
Sex was painful only the first few
times after delivery

1

It is still painful sometimes
It is still painful always ....

2
3

|-> Sections
628

Say : ’’You said it is still painful always when having sex with your husband. I would like
to talk to you in detail about this problem”.
ENSURE THAT THE RESPONDENT UNDERSTANDS THAT THE NEXT FEW
QUESTIONS (Q.628 TO Q.634) PERTAINS TO THE PROBLEM OF EXPERIENCING
PAIN WHEN HAVING SEX WITH HER HUSBAND.

CHECK : IF LAST DELIVERY WAS FIRST CHILD (REFER PREGNANCY HISTORY)
DO NOT ASK Q.628. UNDER Q.628 CODE AS NOT APPLICABLE = 0.
628.

When did you have this problem - after this child or an earlier child ?

1
2
0

This child ....
Earlier child .
Not applicable

629.

Did you consult anyone or take any treatment for this problem. IF NO STATED CODE
NOT CONSULTED (000) IF YES STATED ASK : Whom all did you consult and what
all treatments did you take ? LET RESPONDENT ANSWER AND CODE UPTO 3
TREATMENTS.

Allopathy

Public Health Centre
Private Hospital
Government Hospital
Private doctor/Clinic
Pharmacy

1
2
3
4

Ayurvedic/Homeo/Unani/Sidha
Home remedies - Elders/Dai/Traditional Practitioners
Religious/Superstitious Practices
Not consulted
.
Don’t remem her/Don’t know

6
7
8
000
999

5

CHECK : IF 1 TO 4 CODED IN Q.628 - CONTINUE. ELSE SKIP TO Q.633.

29

4

630.

You said you had gone to (REFER 628) for treatment ? Did they perform an operation or
did they only give you medicine and injections ?
1
2

Operation
Medicine & injections only

631.

Who advised you to go to (REFER Q.629) ?

1
2

Husband
Mother/Mother-in-law

.

3
4

Relative
Neighbour
Dai
Nurse/Doctor
Nobody (I myself decided )
Others (SPECIFY)

632.

5
6
7
8

When you went to the (REFER Q.629), how did you travel ?
IF MORE THAN ONE VISIT HAS BEEN MADE EITHER TO THE SAME PLACE OR
DIFFERENT PLACES VISITED PROBLEM AND DIFFERENT MODES OF TRAVEL
USED EACH TIME, RECORD MODE OF TRAVEL FOR THE FIRST TIME TO THE
FIRST PLACE.

1
2
3
4

On foot
Auto rickshaw
Bicycle
Bullock cart
Bus
Taxi
Cycle rickshaw ....
Others (SPECIFY)

5
6

7
8

GO TO Q.634.

633.

Why did you not go to> a doctor/hospital for treatment of this problem ?
RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.
Ex pensive/Could not afford
Not good/Don’t treat well/Attitude of doctors not good
Too far away/No transportation
Timing/Other inconveniences like no one at home
Members of family would not allow
Did not know where to go
Preferred to take treatment from traditional sources ..
Didn’t think it was necessary
Others (Specify)

634.

LET

1
2
3
4
5
6
7
8
9

How much money did you spend on the treatment of this problem ? PROBE FOR TOTAL
COST OF TREATMENT INCLUDING DOCTOR’S/ PRACTITIONER ’S FEES,
MEDICINE, TRAVEL COST ...
GO TO SECTION 8.

*

30

SECTION 7 - ABORTION
ASK RESPONDENT (i.e) THOSE WHO ARE CURRENTLY NOT PREGNANT AND HAVE
HAD AN ABORTION IN THE LAST 12 MONTHS OR LESS
I would like to talk to you about your last pregnancy which you had said ended in an abortion.

You said that the abortion was
. (induced/spontaneous) and that you had the
abortion in
month of pregnancy. (REFER PREGNANCY HISTORY)

701.

INDUCED ABORTION

> 701

SPONTANEOUS ABORTION

> 705

What was the method used to end your pregnancy ? PROBE FOR METHOD.
MULTIPLE CODING PERMISSIBLE.
Curettage
.
Injection.................................
Pills
Traditional/herbal Medicine.
Massaged womb
.
Inserted stick.........................
Others

702.

11
12
13
21
22
23
88

Who was the person who did the abortion (EXPLAIN : Person who performed the abortion
or gave you the pills or injection )

Doctor
.
Dai....................................................
Chemist / Pharmacist
Nurse
Friend
Mother / Mother-in-law / Relative
Person who specialises in this
Don’t remember

703.

Why did you have the abortion

1
2
3
4
5
6
7
9

Let respondent answer and circle upto 3 answers.

Doctors asked me to have an abortion
Did not want any more children..
Pregnancy not when expected ...
Would interfere with employment
Husband not good
Was using-------------------------to avoid becoming
Thought it will affect the child
Family income not enough ....
Others (specify)

31

1
2 “l
3 I
4 I
5 |—>705
6 I
7 I
8 I
9 J

704.

Why did the doctor ask you to have an abortion ? PROBE.

705.

When you had the abortion did you have any of the following ?
Severe abdominal pain

PC

1
2

Yes
No

706.

Severe bleeding, more than what you usually get during periods.

1
2

Yes
No

707.

Fever
1
2

Yes
No

708.

Did you receive any medical attention after the abortion ? By medical attention, I mean did
you see or consult a doctor or nurse ?

1
2

Yes
No

IF NO STATED (2 CODED) GO TO 711.

709.

Where did you go for medical treatment ?
PLACE AND CODE APPROPRIATELY.

PROBE FOR LOCATION / NAME OF

1

Public Health Centre ...
Private Hospital............
Government Hospital ..
Private doctor/Clinic ...
Traditional Practitioner
Dai
Pharmacy ......................

710.

2
3
4
56 | - 711
8 -

Did you have to stay overnight at the hospital/clinic. IF YES STATED ASK : How many
nights did you have to stay at the hospital ?

GO TO SECTION 8.
32

711.

Why did you not go to a hospital/doctor ?

LET RESPONDENT ANSWER AND CIRCLE UPTO 3 ANSWERS.
Did not know where to go /where it is....................................................
Too far away...............................................................................................
Was not open at times when I could go there.........................................
Not welcoming /attitude of doctors nurses not good .............................
Service not good/no medicine.................................................................. .
Did not think it was necessary..................................................................
No money....................................................................................................
Too busy at home.......................................................................................
No transportation.......................................................................................
No one to watch children at home...........................................................
Husband/family would not allow.............................................................
Other (specify)........................................................................................... .

33

1
2
3
4
5
6
7
8
9
10
12
13

SECTION 8 - FUTURE FERTILITY INTENTIONS
ASK RESPONDENT TYPES 1, 2 (i.e) THOSE WHO ARE CURRENTLY NOT PREGNANT
801.

Have you or your husband undergone any family planning operation to avoid having any
more children ?

1
2

Yes
No

IF YES STATED (1 CODED) GO TO SECTION 9.
802.

Have you attained menopause ?

1
2

Yes
No

IF YES STATED (1 CODED) GO TO SECTION 9.

803.

Would you like to have another child ?
1
2

Yes
No

34

SECTION 9
ASK ALL RESPONDENTS
Now I would like to ask you a few questions about you and your household ?
901.

What is your religion ?

Hindu
.
Christian
.
Muslim
.
Others (Specify)

902.

1
2
3
8

What community do you belong to ?

FC
BC
SC
ST.

1
2
3
4

903.

What caste do you belong to ?

904.

What is the source of drinking water for members of your household ?
Piped water.
Piped into Residence/Yard/Plot....
Public tap
Well water
Well in residence/Yard/Plot
Public well
Handpump
Handpump in residence/Yard/PloL
Public handpump
Surface water
River/stream
Pond/lake
Tanks Truck
....................
Bottled water...................................
Others (Specify)
____________

905.

11
12 *

21
22
31
32

41
42
51
61
81

What kind of toilet facility does your household have ?
Flush toilet
In home ...
Public

11
12

Non flush
In home ..
Public ....

21
22

Open

31

35

906.

Does your household have
a) Electricity
Yes
No

I

1
2

b) Radio

Yes
No

1
2

c) Television

J

Yes
No

907.

1
2

Could you describe the main material of the floor of your home ?
Natural floor

*

908.

Earth/Sand ..
Dung
.

11
12

Rudimentary floor
Woodplanks
Palm/bamboo
.

21
22

Finished floor
Polished wood
Cement/Red Oxide ..
Mosaic/Ceramic Tiles
Carpet
.
Others (Specify)

31
32
33
34
81

Does any member of your household own :
a) A Bicycle
Yes
No

1
2

b) A Motorcycle
Yes
No

1
2

c) A Car
Yes
No

1
2

36

909.

How many rooms in your household are used for sleeping ?

910.

Is your husband alive ?
Yes

1

No

2

IF HUSBAND ALIVE (1 CODED IN Q.910) ASK Q.911 TO Q.914 SEPARATELY FOR
RESPONDENT AND HUSBAND.
IF HUSBAND NOT ALIVE (2 CODED IN Q.910 ASK ONLY FOR RESPONDENT).

911.

Respondent

Yes
No

912.

1
2->913

914.

1
2->913

What is the highest level of school/college you/your husband attended ?

Primary (upto class V)
Secondary (class upto VIII)
Higher (class VIII to XII)
Graduate
Post graduate
913.

Husband

Have you/has your husband attended school ?

l-> 913
23 | 914

4 I
5-

Can you/your husband read and understand a letter or
newspaper easily, with difficulty or not at all ?
Easily
With difficulty
Not at all

l-> 913
23 I 914

4 I
5-

1
2
0 3

1
2
3

1
2
3
4
5
6
7
8
9
10
11
12

1
2
3
4
5
6
7
8
9
10
11
12

What is your/your husband’s occupation ?
Cultivates own land
Fishing
Agricultural labourer
Unskilled worker
Skilled worker
Business
Service
Student
Housewife
Household Industry
(Specify )___________
Others (Specify)_____

915. IF 1 CODED IN Q.914 ASK : How much land do you own?
(acres)______________

37

I

916..
’•

IF 2 CODED IN Q.914 ASK : Do you own a boat ? IF YES STATED :
Do you own a country boat or motor boat ?

1
2
3

Country boat
Motor boat...
Not owned ...
917.

ASK FOR EACH ITEM IN TABLE BELOW
0

Do you own

!
; (ITEM) IF YES STATED ASK : How many
(ITEM) do you own ? (999 = Don’t know ; 000 = Not owned)

918.

No. owned

Item

Q. No.

917 a

Farm Cattle

917 b

Milk Cattle

917 c

Sheep

917 d

Poultry

IF WIFE’S MAIN OCCUPATION IS MENTIONED AS HOUSEWIFE OR WORKING
IN OWN LAND, THEN ASK :

You said you work in your own fields/are a housewife. Apart from this, do you do any
type of work for which you get paid in money or any other form ? IF YES STATED ASK:
Where do you do this work ?
1
3
2
0

At home
Outside home
Both
Not worked ...
919.

During a typical day, what type of work do you do ? Please describe the work in detail.
Would you describe it as light, moderate or heavy.
WRITE DESCRIPTIONS OF TASK ASSOCIATED WITH AND RECORD WHETHER

IT IS LIGHT
MODERATE
HEAVY

1
............................. 1
3
Tasks

a

Light

Moderate

Heavy

1

2

3

1

2

3

1

2

3

1

2

3

1

2

3

1

2

3

THANK AND TERMINATE

38

Annexure - iv

*

List of Project Personnel

Name

Designation

1.
2.
3.
4.

Dr. D.K. Srinivasa
Dr. K.A. Narayan
Dr. Asha Oumachigui
Dr. Gautam Roy

Principal Investigator
Co-Investigator

5.
6.
7.

Ramasamy C.
Maury Marie Emmanuel
Prakash G.

Programming Assistant
Accountant / Data Entry Operator

8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

Sister. John
Ms. Vijalakshmi
Rajasekaran R.
Prakash
Usha A.
Paramesvary S.
Jayanthi N.
Chithra S.
Lakshmi D.
Subramany
Ravi R.
Karunanidhi R.

20.
Vijayalakshmi
21.
Raghavan
22.
R. Narayanan
23.
Girija K.S.
24.
Lalitha K.
25.
Karpagam G.
26.
Chithra S.
27.
N armath a A.
28.
Devy T.
29.
Daisy Jayakumari V.
30.
Amutha G.
31.
Vijayarani R.
32.
Jemiangline H.
33.
Vardhini M.D.
34.
Chennammal C.
35.
Vijaya Sheela J.
36.
Sakthi G.
37.
Gnanambal V.
38.
Bhuvaneswari A.
39.
Rajathi A.
40.
Bhuneswari B.
41.
Vijayalakshmi T.
42.
Geetha P.K.N.R.
43.
Latha C.
44.
Leela B.
45.
Hemalatha R.
46.
Ruth K.
47.
Vijayambale R.
San thy N.
48.
49.
Tamilselvi K.

II
II

II

Field Supervisor
II

Data Entry Operator / Scrutinizer
II

II
•I

Scrutinizer
•I

Driver / Data Entry Operator
Driver
H

Director - MODE Research

Field Organiser - MODE Research
Field Supervisor - ”
Team Leader
II

•I

Field Investigator
II

II

II

•I

•I

II

II

II

•I

II

H

If

II

II

II

•I



4

Annexure - v

Maternal Morbidity Project
Training Programme for Interviewers

Date
09-02-’93

I
10-02-’93

I

Topic

Time

Resource Person

2 - 3 P.M.

What is social research ?
Why it is done ?

Mr. R. Narayanan

3 - 4 P.M.

Introduction to MODE

Mr. T.R. Selvakumar

10-11 A.M.

• Recap of previous day
• Various types of
research instruments

Mrs. K.S. Girija

11 -01 P.M.

Field terminology

Mr. Selvakumar
Mr. Narayanan

2 - 4 P.M.

How to approach
respondents / interviewing
technique

Mr. Narayanan
Mrs. K.S. Girija
Mr. Selvakumar

4-5.30 P.M.

Revision

«

11-02-’93

Technical training

Dr. D.K. Srinivasa
Dr. K.A. Narayan

12-02-’93

10-5.30 P.M.

• House hold eligible
Questionnaire

Mr. R. Narayanan
Ms. K.S. Girija

13- O2-’93 &
14- 02-’93

10-5.30 P.M.

• Mother’s Study

Ms. Vijayalakshmi
Mrs. V. Mohana

15- 02-*93 &
16- 02-’93

10-5.30 P.M.

Mock calls

Mrs. K. Lalitha
Mrs. R. Karpagam
Mr. Tilak Choudhury
Mr. Ashok Nair
Dr. U.V. Somayajulu

17-02-’93

10-5.30 P.M.

Supervised field work

Mr. R. Narayanan
Mrs. V. Mohana

18-02-’93

10-5.30 P.M.

Supervised field work

Mr. R. Narayanan
Mrs. V. Mohana

19-02-*93

10-5.30 P.M.

Supervised field work

Mr. R. Narayanan
Mrs. V. Mohana

20-02-’93

10-5.30 P.M.

Review

Annexure - vi
List of Selected Areas

SI. No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48

Name of the Village

Area type

Nedi
Nedimozhianur
Palapattu
Sendur
Pathirapuliyur
Mariamangalam
Kattalai
Ravanapuram
Perumukkal
T. Nallalam
Vellakulam
Chokkathangal
Alagiapakkam
Vadakullapakam
Keelarunkonam
Nalmukkal
Thumbur
Kaspakaranai
Bambathripettai
Thangal
Poongunam
Ashokaburi
Ulagalampoondi
Kottiyampoondi
Kundalapuliyur
Purankarai
Panchai am
Keelkaranai
Evaloor
Melpettai
Ongur
Pathiri
Thanankuppam
Kattupunchai
Ramanathapuram
Thondamanatham
Thuthipet
Pillaiyarkuppam
Sellipet
Vinayagampet
Sorapet
Vambupet
Madagadipet
KaJ i th eerth al kuppam
Andiyarpalayam
Villupuram Wards - 6,18 & 30
Tindivanam Wards - 5,8,9,20,26 & 32
Pondicherry - Town, North & South

Rural
H
H

ii

H
ii

H

•i

H
H
H

ti
H

it
ti
H

H
ii

it
fl
•I
ii

H

If
ft

H
H
H
H

tf
H
H

H
H

■I
H

■I

•I
ff

Urban
ii

ii

i

<

Photographs of study events

jr *

r

ST- '





E'

e..,

A village social map being drawn

Focus Group Discussion with
Traditional Birth Attendants

Lr

<

]

II-!


S

I

kF
* WK

% V 'J

4

KWi«
■I^^K < • ’^wK«
e



1

' --I ' J

WK

E.. i

fK
1

' it 4>1

Field Investigators Training

Field Investigators Training

Field Investigators Training - Mock Interview

Interview' being conducted

-1.^

I PI

ri

I

Supervisory Visit

Verification of data

$

b :?-

B';

.T

Meeting of Principal Investigators

Scrutiny of Questionnaires

; •

Project Computer Centre

Equip-IOCU Workshop

VIQNI 'NVHlSVrVd
■ LOO £L£ - HDdlVan
IVlIdSOH VNVNVZ

u\ijai\i3iNin3dns
3D3T1OO “1VOIC13IA1 TN'H
3IAII/\IVHDOHd lAiniHVd ISOd

Hoioiwa

->Q o *-lf

ADOTOO3VNAD ’S 8318131530 dO ld3Q
QV3H QNV y0SS3d0Hd HOIN3S

/Q t

3sarj3d vavnia da
A3

96 ’S S6 '1z661.
98 ’S 178 '886 L

xanis dDMOTiod
xanis ■muni

VIQNI ‘NVHlSVrVM
‘andivan hviidsoH vnvnvz ±v
S3SV0 OOP JO Ad ms V
: AinVJLMOIAl 1VNM31WIA1
NO 33N3mdNI dl3H± ONV
aoidsd NouvsnviidsoH - ana
3Hl Nl ONIlVHSdO SUOlOVd

...|

I

I

f

• Vl

I

I

The committee appointed by F.I.G.O. (International Federation of
Gynecologists and Obstetricians, 1969) has defined Maternal Mortality (MM)
as "the death of any women dying of any cause while pregnant or within
42 days of termination of the pregnancy irrespective of the duration and site
of pregnancy". Analysis of maternal mortality at an institution can be divided

!

into two parts, firstly the part concerning various factors operating in the
>

pre - hospitalisation period and secondly the part which deals with the nature

of hospital care given to the patient. The present paper analyses the first
part.
In a developing country socio economic, health and other aspects of
maternal mortality are more important than the directly responsible obstetric

or medical causes. In most of the cases this component is so strong and
evident that without taking this into consideration, it is fruitless to attempt
lowering of the maternal mortality. In each maternal death the role played by
illiteracy, lack of awareness, irresponsible attitude of the case, her family
and community, poor socio-economic status, non availability of proper health
care and obstetric services is obvious. Hence an attempt has been
i

made to analyse the various factors operating Prior to hospital admission in
100 cases of maternal death.
As a followup, the same study was carried out to assess whether a
decade of change and development has brought about any change in the
factors observed to influence the maternal mortality in the earlier study.

MATERIAL AND METHOD

In the past, the study was carried out on 100 consecutive cases of
maternal death in all the 4 units at Zanana Hospital, Udaipur, Rajasthan from
1.1.83 to 31.7.85. The present study was carried out from 1.3.94 to 30.6.96
in the same setting.
Those cases in which all the desired details could not be obtained due
to practical reasons were excluded. In both the studies equal number
(100 each) of MM cases were prospectively and retrospectively analysed.
The data so obtained in the two studies were compared and were also
analysed in the light of data of 81 and 91 censuses.

2

SOME OF THE COMPARATIVE CENSUS DATA
OF THE YEAR 1981 & 1991
OF

UDAIPUR DISTRICT RELATED TO THE STUDY.

i

Population

Urban
Rural
Inhabited Villages
Female Literacy
F/M Ratio
Villages connected

1

4

by pucca road
Villages connected
by Bus/Rail route
Villages having electricity
Villages having some
educational facility

Villages having Post and
Telecommunication
ST/SC Population
Villages having
medical facilities

1991

1981
2,356,959
15%
85%
3117
10.76%
977F/1000M

2,889,301
17.10%
82.90%
3179
19.00%

19.83%

28.31%

21%

30%

37.62%
56.43%

46.13%

72.54%

12.62%
42.54%

20.79%
45.19%

14.47%

20.45%
+ 22.59%

Growth rate 1981-1991 of the

965F/1000M

SC's 78.18% & ST's 97% live in rural area.

Rajsamand become a seperate district in 1991-92 and hence in Census
1991 it was assessed with Udaipur district.

In each case alongwith medical history following additional information

was recorded as it was done in the past.
Caste
Postal Address and Residence
Educational status of the couple
Per capita income
Interval between admission and death

3

Nutritional status and anaemia of the case

Nature of aid received prior to hospital admission
Demographic distribution
Distance covered to reach the hospital
Nature and expenditure of transport
Preventive aspects in Major causes of Maternal Mortality
OBSERVATIONS AND COMMENTS

MATERNAL AGE
Maternal age and parity-wise highest number of cases were in age group
21-30 years and Gravida 2 to 4 respectively as it was in the past. Higher
number of cases in age group below 18 years indicate that the practice of

early marriage and childbirth is still very much prevalent and is likely to stay.
Table la

AGE GROUP OF MM CASES
4

Age in Yrs
< 18
19-20
21 - 30
31 - 40
41 >

Past No. of Cases

7

Present No. of cases

12
21
45
22
0

20
50
21
2

The genesis of child marriage is very much deep rooted in the traditional
rural families of rajasthan. Birth of a daughter is mostly unwelcome because
she is considered to be burden. She is born to be given away to other family.
The objective of child marriage is to unburden one's family responsibilities.
The sooner it is done the better it is. It's perceived advantages are that
during childhood there is no raising voice against it by the individual, it's a
mark of respect for elder's decision and wish, family will not face problem of
not finding a suitable match later on etc. In the rural area family prestige is
likely to be affected and eyebrows are raised if daughter is allowed to grow

up without tying a nuptial knot. Latest RAPSWAS data also show 68%
marriages took place before the legal age of marriage of 18 years. Even

4

today several children irrespective of age in the families are made to tie the

nuptial knot on a particular auspicious day (Akhateej) in one ceremony. This

is considered practical convenient and economically viable because the
marriage expenditure of individuals comes very low. Subsequently the children
are allowed to grow up in respective family. The daughter is sent to her
husband's house once she attains puberty or nearing menarche.

The next but the most important duty of the daughter-in-low is to prove
her fertility as soon as she can. Therefore sex life starts very early i.e. in the

initial phase of sexual development. On close questioning we have found in
small number of cases it started even before menarche. Physiologically, the
initial 2-3 years menstrual cycles are anovulatory and this is "Mother Nature's
Device" to safeguard a proportion of teenage daughter-in-law from the risk
of premature early motherhood. But the question is how many are fortunate
to have temporary protection.

4

Twelve percent MM in the present series of teenage group ( 18 years )
as compared to 7% MM in the past should be viewed more seriously. It
indicates exploitation of forcing motherhood on girls of 1 5, 1 6 and 1 7 years
is on increase. What is tragic is under the banner of social custom or family
tradition this exploitation is going on.
Every villager today knows about the legal age of marriage but still he

does not want to abide by it. He knows how poor implementation of rules is.
He also knows poor law and order situation and by greasing palms he can
get away with it. We therefore feel that child marriage is likely to perpetuate.
Higher number of MM in this group is indicative of that.
PARITY
We have not observed any change in parity wise distribution of cases.

Table I b

PARITY OF IVIM CASES
Gravida

G1
G2 - 4
G5 >

Past No. of Cases

36
43

Present No. of cases

29
47
24

21

5

Nearly one fourth of MM cases were in age group 30 + and gravida 5 +
indicating that not much has changed over a decade in relation to grand

multiparity and reproduction at advanced age. Large family size is still an
accepted norm. This has been also revealed by Census 1991. TFR of Rajasthan
State is higher (3.63) than the national average (3.39) and for rural Rajasthan
is was found to be still higher i.e. 4.81. Some of the reasons behind it are
more hands are desired to manage the land and cattle, desire for more number
(2-3) of male children, high IMR i.e. poor neonatal of infant survival in rural
area etc.

EDUCATIONAL STATUS
In relation to literacy and educational status of MM cases a gloomier
picture has emerged. Illiterate status was found in 88 as compared to 82 in
the past. Secondary and above level of education was found in only 4 as

compared to 9 cases in the past. This indicates that MM cases of present
study were found to be educationally more backward as compared to those
in the past.
4

Table II a
EDUCATIONAL STATUS OF COUPLES

Husband
Level

Past No.
of Cases

Illiterate
Literate
Primary

52
10
10
13
6
9

Secondary
Hr. Secondary
Degree

Wife

Present No.
of Cases

Past No.
of Cases

Present No.
of Cases

88
6
2

10
4

82
7
2
6
1

3

2

44
21

18

1
2
1

Availability of educational facilities in 72% villages in Udaipur district
(Census 1991) and literacy campaign started in the state in the year 92-93

6

are yet to show their impact on our poor isolated majority in the rural area
particularly among the backward caste population. They are still not impressed
or convinced of benefits of education. Rather at present they are impressed
by the improvement in market economy even at village level. With opening
up of market more jobs are created. If you can earn desired wages easily
where is the necessity of sending your children for education or to educate
oneself ? Education as a prerequisite for bread earning or jobs is losing its
impression amongst our rural poor.

Another change is brought by the electronic media. With TV network
and other electronic viewing equipment reaching villages it has made a place
in their daily life. The entertainment programmes keep them and their thoughts
preoccupied. Messages regarding literacy and education are yet to be taken
up by the rural folk and show their impact on female literacy which is only
19% for entire state of Rajasthan. A rather depressing picture is observed in
female literacy especially in rural areas of Udaipur District where it is only
2.75% in Kota Tehsil (Census 1991).

t

We have found that sending their daughter to school is still not a practice
or proposition. A daughter is supposed to take care of her younger brothers

and sisters, cattle and help the mother. Majority of primary schools in remote
rural areas show very poor attendance of female children. But then it is also

a fact that a number of schools in remote rural areas are existing only as

structures and in records.

The change in educational status we have found in our follow up study
certainly does not indicate that it has decreased in population at large because

the type and SE profile of population in which MM occurred has changed.
And this will be highlighted subsequently also.

CASTE

In follow up study we have found that the number of MM cases belonging
to SC/ST have considerably increased (45 vs 74). The number of MM cases
belonging to higher castes has shown a sharp decline (53 vs 22).

7

Table II b
CASTES OF MM CASES
Castes

S.T.

Past No. of Cases

Present No of cases

31

48

14

26

2

4

53

22

SC and other

backward classes

Muslims & Other

minority religions

Higher castes

( X2=13.91 df = 3 p

.001 highly significant )

Several factors have influenced this change. It's a known fact that a
lion's share of benefits of technological and economic development has been

cornered by urban population. It has benefitted higher casts, middle and
richer classes and urbanites more than backward castes, poorer and rural

people. This is a world wide phenomenon but no where is it as glaring and
gross as it is in India and Rajasthan in particular.
This has markedly improved SE status of urbanites who constitute only
17.10% of state's population (Census 1991) and this has added to their
awareness and spending. This has occurred in relation to health and pregnancy

care also. Naturally MM cases from urban area (Udaipur city) have dropped
down from 15 in the past to 7 present (Table VI).

The SC/ST constituted 45.19% of total population of Udaipur district
(Census 1991). In our present follow up study we have found MM cases

belonging to SC/ST constituted 74%. We have also found that there is marked
increase in the number of MM cases belonging to SC/ST needs serious thinking
and attention.

For centuries together SC/ST population has remained isolated from

mainstream and neglected. Majority of such population particularly tribal live
in remote, undeveloped, isolated and poorly accessible areas (Census 1991).
Their philosophy of life has always been "Live for Today''. Such population

has remained poorest of the poor even today and got least benefitted by the
process of development. Under tribal development

programmes nothings

much has changed except little rise in employment avenues, primary schools

8

and development of roads and transport facilities in previously inaccessible
tribal belt. It is the last change which has led to more number of serious or
emergency including obstetric cases reaching the hospital. The last factor
i.e. easier access, may explain the higher number (45 vs 74) of serious
cases belonging to SC/ST and residing in remote and previously inaccessible
areas reaching the hospital in the present study.

The physical hardship and exploitation of tribal women is beyond ones

imagination. Alongwith poverty and unregulated reproduction she is expected
to compromise with all adversities of life. Bringing water, fodder and fuelwood
from remote places alongwith earning daily wages by physical and manual

work even in advanced stage of pregnancy can still be witnessed when we
travel in those areas. There is no concept of antenatal care nor facilities are
developed at most of tribal area. Therefore number of tribal mothers dying
of neglect, severe anemia and malnourishment and poor ante and intranatal
care are likely to be much more. Such MM has always remained unreported
and unregistered because she died at home. The MMR of tribal area must be
much more. The positive aspect of the whole situation is that number of
MM of tribal areas or belonging to SC/ST are occuring in the hospital now

and are being reported and registered.

The MM cases belonging to SC/ST coming from tribal or non-tribal rural
area tell us about the miseries of human life and motherhood in those areas

which have worsened by resurgence of Malaria and spread of STD's.

HIGHER CASTE

Table II c
HIGHER CASTE MM CASES

9

Number of Cases

Caste

Past

Present

Rajput
Vaish
Brahmin
Jat
Others

18
12
11
7
5

14

TOTAL

53

22

9

3
2
2
1

The number of MM
cases belonging to higher castes has gone down to
nearly 1/3rd of what it was i
m the past. On fractionation we found that
number of MM belonging to Rajput cases did not reduce
special mention.
reauce and this needs

rses beionBin9 to

consX'sr. w

study it has gone up t0 /0. ( 43’0 X7

bMM but in ,he ,jrescm

XX- - numb„ o, MM hXXX gb=XT'
the slaws^; women '"this"™'6'

~ a„d
marriages etc stiH prev^o

be'°n9in9 "> R^< caste unfolds

x:z™ r *
acceptance of widow

f

PER CAPITA INCOME
devJXnt0XZa’isT' 7

caiam as

tremend°US '—^-l

7“ X109™' O‘ ™ra'

POO,. And in

ZeT9

priority of economic

benefits.

Poverty means that <
one does not have sufficient rreserves to spend
even during emergency. This
j encourages the business of
money lending

10

and further exploitation of the poor. This poverty still worsens when there
is uncontrolled price rise even of essential

commodities and unwanted

unregulated additions to the family size. Under such circumstances it is the
pregnant mother who suffers the most on many counts and this explains
why the majority of cases of MM both in the past and at present were poor.

Table III
FINANCIAL STATUS OF MM CASES

Group
A

Assessment Criteria

No. of Cases

Agriculture Labourer Class Very Poor

46

Beggars, Cooli, Baggage Lifters or Carriers
No steady income
Earn meagre wages only few days a week
and few months a year
Possess no Land, House & Assets
Lives in temporary house or hutment

B

Labourer Class Poor

22

Works in Factories, Private Establishments,
Shops etc. and does Odd jobs

Earn wages daily Rs.30-40/- per day
Small land holding & house of their own

in village or hutment in slum area in city

C

Low Middle Class

22

Permanent job in organised Pvt. I Public
sector, class IV govt, employee, small
business holdings
Steady source of income

Financial balance & security present
Lives or may own pucca house.
Moderate size land in rural area from which
there is steady and regular income

D

Middle and Higher Class

10

11

As per Planning Commission the criteria set up to determine poverty
line are incomes less then Rs.12000/- per annum per family or Rs.2,600/per annum per capita. Application of this criteria to our cases a was difficult

because poor persons belongings are

also as poor as he is to yield an

income. The job or work he does is not permanent, is without any financial

security, odd and inconsistent. Still the group A and B identified by us fall
below poverty line. In the past we have found 52 MM cases were below the
poverty line and in the present study it has gone up to 68. We have found a
gloomier picture in relation to per capita income of our MM cases below
poverty line which has shown an increase of 16%. With this limited

experience we still doubt the economic and technological development
benefits reaching the poor and particularly the mother who is the poorest of
the poor.

This also indicated the number of MM cases is inversely related to per

capita income. Elevation of people living below poverty line will certainly
lower the number of MM cases.

NUTRITIONAL STATUS AND ANEMIA OF CASES
In relation to anemic status in follow up study we have found that

overall Hb status of MM cases was worse than what we have observed

a decade back. In the present study there was no case in nonanemic
(Hb > 10.1gm%) group as compared to cases in the past. A much shocking
picture is seen in relation to severe anemia group ( Hb < 4gm%), It's number
has gone up by three times. Over three fourth (78 cases) cases had moderate
to severe anemia ( Hb < 8gm% ).
Table IV
Hb STATUS OF MM CASES

Hemoglobin
in gms %
<
- 4
4.1 - 8
8.1 - 10
>
- 10.1
Not done

Past
No. of cases

Present
No. of cases
34
44
17
0
5

11
38
34
10
7

12

What caused this change appears to be complex and perplexing
loopholes of modern technological development. Majority of us are carried

away by its glitter, material comforts, quick results and high speed. Very
few of us want to explore the loopholes in it. Some of the loopholes are
realy dark and complex. A sharp rise in severe anemia cases leading to MM
is one such dark loophole. It is the prepetuation of earlier factors like
poverty, chronic malnutrition, poor sanitation and living conditions of majority
which has remained as such in the spate of high speed development and

now joining hands with resurgence of malaria and other infections.
Because of the agricultural and technological development we have
attained a self sufficiency in food grain. Now the starvation and famine

deaths are unheard of but major degree of malnourishment is still rampant
in poverty stricken 40% population. When this severe malnourishment gets

associated with unplanned pregnancy the body immunity (due to severe
hypoproteinimia) falls an easy prey to various infections including malaria.
Malaria is also known to damage the immune system and therefore makes
the body susceptible to multiple infections. Malaria co-existing with every
other or multiple infections is a very perplexing problem in front of the

clinicians. We have found that there is a total change in clinical presentation
of the case. The typical picture of different types of infections and fevers all

have changed for the worse, li has thrown a challenge to planners,
demographers, administrators, medical researchers and of course to the
clinicians.

Recurrent attacks of malaria causes progressive anemia. The possibility
of developing severe anemia because of untreated malaria in females is

otherwise also high because of their poorer accessibility to health system.
The anemic status further worsens during pregnancy because of increasing
fetal demand and other body changes. That explains such a high incidence

of moderate to severe anemia in MM cases.

In 9 cases the diagnosis of Malaria was confirmed and in 17 cases,
though there was high clinical suspicion, could not be confirmed because of
number of reasons (Table X ).

Severe anemia complicating obstetric emergency needs immediate blood
transfusion. This is also a well known fact that such cases also have a very
high incidence of blood transfusion reaction. Added to it is the risk in nearly

13

40% of cases of going into sudden heart failure when Hb level is less than

6 gm%. Over and above this are now newer risks of blood transfusion i.e. of
HIV, Hepatitis B, syphilis and malaria from donors. Screening of all this have

become mandatory and rightly so. But then there are frequent incidences
where we face that blood could not be supplied because various mandatory

test material was not available. ( Management or administrative failure).
Social delay in blood donation by the relative in obstetric emergency cases
because the recipient is female as highlighted in the past still continues. All
these have influenced MM in severe anemia cases. We have found the
change in SE status also influenced Hb status of MM cases in the present
study.

ANTE AND INTRAPARTUM CARE

When compared with the past we find that nothing much has changed
in relation to available ante and intrapartum care of MM cases prior at
hospital admission. TBA still remained the sole agency responsible in
majority(60%) of MM cases. As per Census 91 only 20.45% of the villages
have medical amenities in the village itself. It means majority of population
in rural area will have to move out of village even for basic primary care like
antenatal care.

Table Va

PERSONS/AGENCY RESPONSIBLE FOR ANTE
& INTRAPARTUM CARE IN MM CASES

Person/Agency

Past
No. of cases

Present
No. of cases

T.B.A.

68

PHC, SC, CHC, staff
Other practitioners, quacks etc
Nil or had no care, came on
their own

37
5

60
40

12

21

Had proper reference slip

11

15

14

10

On further elaborating on the primary care we found following.

Table Vb
TETANUS IMMUNISATION STATUS OF MM CASES
T.T.

Past

Present
No. of cases

No. of cases

2 Doses
1 Dose
Not recd.
Pregnancy <16 Wks hence
immunisation not initiated

22
10
60
8

36

16
31
17

More number or MM cases had received TT in present series as compared
to previous study ( 22 vs 36) . But there has developed an undesirable
concept amongst people as well as health workers i.e. what is required in

antenatal care is only TT. Therefore after giving shot of TT, FHW does not

bother to do

necessary examination of the case, give advice, iron folic

tablets (IFT) etc. As a result linkage between mother and health worker
does not develop. Analysis of other components of antenatal care in the
past and present are shown in Table Vc.
Table Vc

TYPE OF ANTENATAL CARE OF MM CASES
Past
No. of cases

Present
No. of cases

Booked cases of our Hospital
Proper ANC elsewhere > 4 visits,TT + IFT +
Inadequate care, Erratic or 4 < irregularly

10
6
4

5
10

paid visits, TT +, IFT +
Pregnancy <16 Wks ANC not initiated
No care at all

8

17
50

Key - TT - Tetanus Toxoid

72

( X2 = 4.33 df = 1 p

IFT - Iron Folic Tablets

15

18

.05 Significant).

In one of the survey carried out by undersigned it was found that only
in 21 % of the antenatal cases FHW or LHV did proper examination (Hb, P,

BP and PA) alongwith giving TT and IFT.

But the main problem is that of monitoring system at PHC which
exists only on papers in good number of centres. Those who are responsible
for this have shown apathy towards it for the reasons which are irrational
and unreasonable. Actually, the safe motherhood programme at various
levels exists still on the paper only.
There seems to be lack of seriousness in it's implementation. One of
the reasons behind it is that to the majority in administration it's a programme
for the benefits of the females. And it is here that the change in this
attitude is desired. It is the change in the attitude which takes time and may
life time.

DISTANCE COVERED
In relation to MM the distance covered by the case to reach the hospital
becomes important for several reasons. The average distance covered in
present series is72 Km. The number of cases from the city have halved.

Cases who has to travel more than 100 Km and above have doubled.
Table VI

DISTANCE COVERED TO REACH HOSPITAL
Distance in Km.

Past
No. of cases

Present
No. of Cases

7

City limit

10

30
60
31
61
100
101 - 150

151 -

>

15
14
26
34

10
1

7

20
30
23
14
6

More is the distance and time taken for the journey more is the
deterioration in the condition of the cases.

16

But the most pertinent question still remains is why they had to travel
such a distance ? The answer remains the same as in the past. Zanana
Hospital, Udaipur remains the only hospital to render proper specialist
obstetrical emergency services to the people of Udaipur City, Udaipur district,

adjacent districts as well as for the adjoining state like MP. With increase in

the population density ( 129[sq Km in 1981 and 167/sq Km in 1991) and
more frequent transportation available, increased number of emergency
cases from far off places have started rushing to this hospital in the hope
to survive.
In past one decade there has been no expansion or further development
in obstetric emergency services in the district. And it is still in bad shape.

Even in Udaipur city the govt has failed to develop alternative maternity
services despite having surplus manpower in two so called satellite hospitals
which run only OPD services.
TRANSPORT

When distance covered by the case is considerable ( average 72 Km) it
is important for us to know about the nature of transport used. Because of
the hilly terrain and improper roads for such a distance delay in transportation
leading to deterioration of condition was a strong possibility in majority of
the cases.

Table VII

TRANSPORT USED BY MM CASES

Transport

Past

Private jeep

No. of cases
37

Present
No.
of Cases
__

Autorickshaw or 3 Wheeler
Private bus
Truck

25
16
12

23
12
7

Bus - State Roadways

10

3

Private Bus or Taxi
Bullock Cart
Ambulance Pvt
Ambulance Hospital
Tractor

9
6
3
3
3
4

6
4
6
2
4

Carried on back or cot

8
(More than one mode of transport was also used by some)

17

Over a decade the mains change that has occurred is that more number

(62) of cases were transported by private jeep as compared to those in the
pase (37). We have found this fourwheel drive vehicle is suitable for rough

and bumpy road and has a speed. It has gained popularity as well suitability
in tribal as well as non tribal rural hilly areas and proved helpful though not
ideal.
But what surprises us is that more than one third cases had to use more
than one mode of transport to reach the hospital. She was carried on a cot

or on back from her home in the remote village ( which had no proper
motorable road) to the nearest point on motorable road. There they hired a

passing private bus or minibus or truck which has transported her to the
fixed point in the city and from there to reach the hospital they hired a three
wheeler or autorickshaw. If this is the state of affair in relation to basic
facilities available for emergency situation then how one can think of averting

death. The distance and transportation both jointly must be taking more

maternal lives than what is observed by us in the hospital. We have analysed
the transportation history of only those few who were destined to die in the
hospital.
Desperate situation like this invariably occurs at odd hours in obstetric

complication cases and that makes them accept any available transport at
hand. There was history that in 4 cases age old bullock cart was used and
8 cases were carried on back or on a cot. There was no question of choice,
suitability and option.
From the Govt, side we find that still there is no policy or directions
from health department regarding transportation facilities (transport + POL)

for serious emergency maternity cases from rural area to the city govt
hospital. Repeated reminders in this matter have failed to draw attention.
EXPENDITURE ON TRANSPORT

Because transportation facility is not provided by the govt, even for
serious referred cases from PHCs, private transport was hired by all but 8
cases. These 8 cases were mostly from the city who were carried by our
hospital or public sector hospital ambulances. Rest all had to pay from their

pocket. But what concerns us is how much they had to pay because 68% of
them were very poor (below poverty line).
18

Table VIII

TRANSPORT EXPENDITURE BY MM CASES
Rupees

Past
No. of cases

Present
No. of cases

<

- 50

42

21

51

- 100

6

101 - 200
201 - 500
501 - >

23

10
11

26
3

38
12

We have found the average expenditure on transport borne by the case

has doubled (Rs.300/-) over a decade. Some of the factors responsible for
this rise were distance covered was more, price rise on diesel and other

things etc. What concerns us was they had to pay exorbitantly. Viewing
emergency desperate situation, illiterate and ignorant clientele they were
made to pay out of proportion. Since majority of them were very poor, had
no cash balance and therefore had to get it from money lender or by pawning
some belongings before undertaking the journey to the hospital.
This whole exercise made them more and more poor both materially as
well as emotionally, A maternal death really crumbles the house and wrecks
a man. In the recent oast prices of everything have shot up than what there
were ten years back. We have driven a poor man more towards poverty in

the event of MM than what it was in the past.
This all also shows how important it is to develop emergency obstetric
care at few more centres around Udaipur and frame a policy of providing

transportation at reasonable or controlled rate to emergency MCH cases if
we really want to achieve ' Safe Motherhood and Poverty Elevation'.
DEMOGRAPHIC DISTRIBUTION OF CASES

Demographic distribution of cases has remained more or less same as

no further health facilities were developed anywhere. Malaria resurgence
was observed at all the places but was found with more severity in tribal

19

belt which was more hilly and had more rocks, water, valleys and dense
vegetation.

Because of the growing

awareness of antenatal care and hospital
delivery in urban population, the number of MM from city area has markedly
gone down. Many newly started private clinics and maternity homes in the
city also have contributed to it.

Raj Samand ( Part of Udaipur district in the past) was declared as
separate district in 1992/93. It is a very rich area of marble mines and is
undergoing rapid development with growth of marble industry. Hence we
hope that the process of development will include that of emergency obstetric
and other medical care also.

INTERVAL BETWEEN HOSPITAL ADMISSION AND DEATH
Admission-death interval scenario of MM cases has shown a trend
towards more desperate a situation than what it was a decade back. Twenty
one i.e. 1/5 of cases as compared to thirteen in the past died within two
hours of admission indicating that their condition was so bad that nothing

could be done. Contributing to such poor condition of MM cases were two
category of causes i.e. direct and indirect. Direct causes invariably were

obstetric like Eclampsia, APH, PPH, Rupture uterus etc. Indirect were the

not so well defined, poorly known, poorly highlighted but more deep rooted
and dangerous causes. Therefore these need more attention. These indirect
causes have been discussed as various factors in this study.

Table IX
ADMISSION DEATH INTERVAL IN MM CASES

Times in hours
<

2
4
8
24
2
7

Past
No. of cases

13

Present
No. of cases
21

12

13
14

7 days

21
26
4
14

> days

10

2
4
8
24
48

20

25

8
15
4

y we could not save her on hospitalisation is not because of direct
cause of death or fa.lure on the part of hospital emergency care but because
o

number of factors like severe malnutrition, poor

or nil primary care

.stances and delay in reaching the hospital, no timely or proper referral
unplanned or unregulated pregnancy , pre-existing poor health status and
poverty etc. were found complicating and depleting body's last reserves as
.t was found m the past. But in the present study magnitude of these
adverse factors was worse. Adding to it this time was moderate to severe
anemia in majority due to resistant malaria and malaria like fevers and
Th,s has been hi9hl'ghted in Table IV also. Hence number of
cases with moribund condition on admission
------------- 1 was much higher.
This also indicates at'
strong possibility that good number of them must
be dying on the way or before hospitalisation.

This indicates that MCH programme at PHC

and SC level should

incorporate malaria prevention and treatment with it because malaria was
found both directly & indirectly responsible for good number of MM.

CAUSE
We have observed some change in the pattern of direct causes of death
over a decade.

Though the number of PIH ( Pregnancy induced hypertension ) cases
admitted to the hospital remained more or less same but mortality due to
Eclampsia drastically came down. Thanks to Magnesium sulphate therapy,
imely administration of Magnesium sulphate to all suspected cases of
Eclampsia except those who have received other drugs prior to admission is

a policy followed in all 4 hospital units since 1990. This has greatly helped in

avertmg MM. We consider Magnesium sulphate a wonder drug a blessing
for survival of eclampsia cases. This made Eclampsia to slip down from

commonest or number one position to number four position as cause of
death. It has dropped down from 28 to 1 3 in the follow up present study.

21
?.

Table X

CAUSES OF MM
Direct cause

Past
No. of cases

Present
No. of cases

Eclampsia

28

13

Hemorrhage

23

31 ( Abortion 4)

(APH PPH 27)
Sepsis

Severe anemia
Rupture uterus

21

21

( Septic Abortion 4 ) (Septic Abortion 11)
4
7
4
7

Malaria provisional diagnosis 17 cases

Malaria confirmed diagnosis 9 cases

There has been six fold increase in cases who succumbed solely due to
severe anemia (Kb < 6 gm%). The number of cases who died due to rupture
uterus also increased from 4 to 7. All these cases had rupture uterus due to
obstructed labour and unskilled and bad handling prior to admission.
In the present study the most disturbing trend was found in causes like

Septic Induced Abortion and Malaria. Number of MM due to septic induced
abortion have increased three times (4 vs 11) .
Abortion was legalised in India by MTP Act in 1972 with the objective
of protecting maternal health at large and giving an opportunity to free herself
from the burden of unwanted pregnancy. What tragedy of maternal life and

health can be bigger than Septic Induced Abortions and that too twenty

years after legalisation of abortion. It appears the MTP act is defeated in it's
very objective. Without proper implementation and service infrastructure
the Act alone carries no meaning.
Septic induced Abortion was the result of illegally or criminally induced
termination of pregnancy by unregistered, unskilled persons mostly in rural

area because MTP facilities nearby were not available even at referral hospital
level. Therefore where a rural woman with unwanted pregnancy can go?

22

She invariably goes to a quack or to a local abortionist who is a semi or
unskilled unregistered person. Quackery abortion practice by unregistered

persons has flourished everywhere in India andmore so in rural areas. Majority
or women in remote rural areas are hence exposed to risks of such abortion
practices and some of them in turn pay with their life. Lack of awareness
and illiteracy have perpetuated such practice. But then even if persons doing

illegal abortions are identified they are rarely caught and punished. Judicial
procedures have become tedious, tardy, costly and out of reach of poor

people. Even if someone responsible for criminal abortion has been caught

and legally

punished then also problem will not end. The basic question

remains i.e. what alternative MTP facility the govt.will provide in rural area ?

Government's failure to provide MTP facilities in rural areas is the main
reason of illegal abortion trade and large number of MM due to septic induced
abortion. With recent slackening of law and order situation such illegal trades
are further flourishing. As the population increases so also the number of

unwanted pregnancies as majority do not accept FP. ( Couple protection
Rate 43.3% 1990) . The higher number of MM due to septic abortions in
the present study is indicative of this. At no time the reproductive health of
rural woman has been at stake as it is now. Another sad part of the story
is that very few or almost no NGO have taken up the issue of criminal
abortions in rural area.

PREVENTIVE ASPECT

To have an introspection and to bring improvement in relation to hospital
emergency services in such serious cases we started having MM meeting
every month since the year 1987. At such meeting MM cases of that
period are retrospectively analysed and discussed. The resident P G student
who was on duty at the time MM occurred is made responsible to collect all

possible details of additional history of the case from relatives and to present
all the details at the meeting. The teaching, nonteaching staff and P G
students attend the meeting and participate in discussion. The ultimate
objective being to find out the avoidable/unavoidable aspects of each MM
case.
We have found this medical audit constituted excellent teaching material

23

V COn,ribU,ed ,O Preventiue — -

hoX..

" in the

s°™ °< ‘-e
Table XI
AVOIDABE FACTORS IN 100 MM CASES
Factor/ Variable
No. of cases or %

Proper Antenatal care and screening of HRP
Sterilisation after last delivery

Timely treatment of Malaria & Anemia
Timely referral
M T P

Simple aseptic technique of delivery

70
39
28
32
11
10

Proper management of III stage

9
Avoidable MM
Unavoidable MM

87
13

i.

GENERAL REMARKS AND SUMMARY

■To

xxv r —«

pur-R—

24

Parameter assessed

Present
No. of cases

Change

12
22
22

T

Illiterate status

7
21
21
82

88

Cases belonging to higher castes

53

22

?
U

Cases belonging to SC/ST

45

75

income below poverty line
Cases with moderate to severe

52

68

anemia (Hb < 8 gm%)
TBA -Main agency responsible
for ante & intranatal care

49

78

T T received

68
32

60
52

Booked cases

10

5

No primary care
Cases from Urban/from city

72

50

15

7

11

20

Age 18 years
Age 31-40 years
Gravity 5

4

Past
No. of cases

n

Cases belonging to Per capita

<,

Cases who travelled
100 Km to reach hospital

?
n

T
(-2)

I
I

(^-2)
(X 2)

n

27

62

Rs. 201 - 500
Transport expenditure

26

38

T

Rs. 501 >

3

12

T? (X4)

of hospital admission

25

34

Eclampsia cause of death

28

13

T
i (-2)

Hemorrhage-a cause of death

23

31

Severe anemia-a cause of death
Rupture uterus-cause of death
Septic Induced Abortion a
cause of death

4
4

24
7

T? (X 6)

4

11

T? (X 3)

Transported by Pvt.jeep
Transport expenditure

Mortality within 4 hrs

k

Key - (?) Increased ( TT ) markedly increased ( J-) decreased

(
(

) markedly decreased ( X ) number of times increased
) number of times decreased (^) unchanged
25

The adverse socio economic circumstances which lead to Maternal
Mortality are still more tragic but we like-minded persons should continue
to highlight and fight for their improvement.
A study & struggle to continue
«

t

This work is dedicated to
Thakur Sri Ramakrishna

1

*

*

Safe Motherhood is related not only to
the Mother & her baby but to the
existance & welfare of human race. To
achieve this is a global responsibility &
duty.

Dr. Vinaya Pendse

26

I

critical time and have been doing so since past ten years.

A Dai from Gujarat State

Some take up this work during their association with an NGO.
A NGO SARTHi trained me as a barefoot veterinary worker When i started giving
treatment to cattle, I realized that there are no trained Dai in my village. The PHC is 15
kms away from my village and we had to go to the CHC in neighboring towns in critical
condition. Women and children used to die on their way to hospital. I participated in the
Dai training. I went to the PHC and CHC to enquire about the services and when the
ANM comes to our village on that day, at fixed time I collect all the pregnant women for
check up and immunization, i also motivate them to go to the Anganwadi and eat the
food supplied. I also participate in the monthly meetings at PHC and collect Iron tablets
and Disposable delivery Kits. Through my efforts, I have been able to motivate women to
avail the services and health care provided by the Government. Earlier, they were aware

of the services but did not avail them, but after I started this work, they have started
availing them.
...Chanchiben, a Dai from Gujarat State
Most Dais are engaged in labor or agricuiture. They support women in labor as a duty
towards their community and assist women in need. The communities happily reward
them with gifts like a sari, grains, and coconut etc. and at times little on no money. Most

of the time it was left to the family based on their capacity as to what they would like to
provide. Many a times they wait for hours together barely get a cup of tea and usually
end up receiving hardly any money. At some point of time, a Dai may negotiate and
demand remuneration for her services Hence most Dais remains in poverty. With more
and more families reducing the number of births, they are getting out of business, which
is affecting their meager income.

We do this work because it is dharam ka kaam (religious work). We walk distances, wait!
for hours together, and risk our lives to go at any time and toil with the woman. Still we ;
: get very little money. If the family is poor we may not even get a cup of tea. Some i
families give us some grains or old clothes whereas some give us Rs. 50-60A
i (A Dai from State of Maharashtra during healers meet organized by CHETNA)

CI1E1NA

:

5

CIIETNA

5

The Dai speaks the same language and is an integral part of the religious and cultural
system of the society. They enjoy the full faith and trust of the village women with regard
to her own health and that of her children. When any problems related to gynecology
arises village women consult them. They possess midwifery skills and are trusted
counselor to the women during pregnancy, delivery and for the care of newborns. In
majority of cases, Dai is the only person available at any odd time and all the time, round
the clock in the village.
/ am a skilled person, i have delivered most of the young people of this village. The
entire village respects me and listens to my advice. Once there was this family whose
daughter was very young and weak. I assessed that she would not be able to 'take the
pain. I advised her for a hospital delivery. Her parents agreed and today she Is happy
with three daughters.
...A Dai from rural Gujarat State

i Ayurveda, the ancient science of life, one of India’s formal, institutional traditional health i
! systems, strongly recommends birth in a home which is equipped to take care of birth i
and birth related concerns and with the help of experienced and wise women indicating i
the critical role of women in childbirth processes.

About Dal's healing knowledge and skill
Complaints of Dizziness, lump in the body, morning sickness, broken heart, internal
disagreements and worries, birth of a calf, property problems, delayed labor and so on
are handled by Dais Women and men, share and consult her with varied problems and
so is the guidance and measures provided by her. Most women experience her warmth
and the deftness of her skills during a trying period of their lives, when her near and dear
ones decide to be aloof.
The Dai performs elaborate to modest ceremonies to invoke the goddess of birth and
pray for a safe delivery, to pacify a violent husband and engage him in caring for the
mother and child.
A Dai is an original practitioner of comprehensive medicine. According to Anthropologist
Karen McCarthy Brown they combine the skills of medical doctor, psychotherapist, social
workers and priest. By specialty the Dais heal women and men in the context of their
family and society, taking in to account the culture and the physical, spiritual context.

As soon as / enter the house o: a woman in labor. I ask the family to bring warm water to
wash my hands and feet. Then ask the women to open al! the locks of the house. (This
symbolizes the opening of the cervix), i assess the womans condition and inform the
family accordingly. I also ask the woman in labor to separate a mound of flour with a coin
(symbolizing separation). Then i coax the woman to squat and than to bear down, i also
massage the cervix and the passage with edible oil and hold the baby carefully when it
pops out. (A Dai from the State of Uttar Pradesh)

CIIETNA

6

The community believes that the process of childbirth is understood not only in
physiological terms but woven in the cultural and religious ideologies of women and
communities. Childbirth is viewed as not just a physical process but also an act of
bringing soul to the earth. Childbirth processes are facilitated by special goddesses. Each
community has their particular deities whose blessing expedites childbirth and protects
the women.
It is within this matrix that the Dai’ works not only as an accomplished herbalist and a
childbirth facilitator but also as a ritualistic. She adopts various rituals to invoke the
goddess of birth and seek her blessings for a safe birth. They function as a mediator of
the divine energy, channeling energies or ancestral knowledge.

Labor therefore is a process of opening up and separating two souls and various rituals
are performed accordingly. For example, the Dai may ask the women to open her plait or
all the locks of the house or separate a mound of flour in to two using a coin.
Rituals are helpful in facilitating 'woman's labor...imagery acts as interface between mind
and body and bringing what in physiological- terms have been considered involuntary
system, at least partially under conscious control. (Chawla, 1994)

The client-woman in labor is actively engaged in childbirth processes through dietary
modifications, rituals and regimens. They are measures to connect oneself with the
internal and external environment. One would expect love, respect and benevolence from
. the community for this noble work. But many times, down trodden by the casts, tradition
and humanistic values bind her to serve others. Her work is often taken for granted,
without recognition or reward with some time scarcely more that insult. Often her work
has been ridiculed and her role and services ignored.
Dai is thus rich resources available at the village level for addressing women's health
concerns. On one hand they possess the midwifery skills and on the other command the
respect and trust of the community. In addition to this they are the only persons available
at any add time and al! the time round the clock in the village. Being a resident of the
same village they Know each family personally, the status and the history of most of the
women of the village.

VVfe hsve to he reedy to go out any time. We have an additional threat of violence
from some villsge people, particularly when we pretest Many a times, after a long
day, when we are about to retire tor the day, a caii comes. At night it is even more
dangerous as the area is fuh of snakes and there is no light or roads. At times! ask
the person to bring a cycle. If they cannot. I have to walk for miles.
.(^ Dai (suin) from a rural area in the State of Maharashtra)

_ J1____ JIIIHLH JB II

Page 1 cf1

Main Identity
From:
To:
Cc:

Sent:
Attach:
Subject:

"chetna" <chetna@icenet. net>
<phmsec@toucntelIndia. net>; <secretariat@phmovement org>; <sochara@blr. vsnl.net in>
<gk@citechco.net>; <maria@iphcglobal.org>; <vhai@unv.emet.in>; "Meera Shiva”
<vhai@vsnl.com>: <vhai@del2 vsnl.net. in>

Wednesday, January 28, 2004 11:56 AFv1
> « \7 7
Daigadh2.doc
(
Paper on safe homebirths and traditional midwife

’('pKn^^^ty^hfefindia'nef . secreiaYiattWphmdvement'ordP'

'pp08??>8pp'

Dear Dr. Ravi Narayan,
Greetings from CHETNA!
« hope you must have received my comments on the dai (traditional midwife) issues for the Mumbai deciaration.

Based on the discussion with Ms. Maria and Dr. Zafarulla Chaudhary, it will be an excellent idea to cal! a meeting
of traditional midwives during December 2-4; 2004 at GK Savar. This would be a valuable contribution of PHA in
protecting the rights of Dais who are generally from the vulnerable groups.

I am attaching a paper on safe homebirths and traditional midwife. Please send us your valuable feedback and
comments on the same. This could be modified to include traditional midwifery from other countries.

Looking forward to hearing from you.

With warm regards,
Vd. Smita Bajpai

-■’t

Coordinator (Traditional Health and Healing Practices)

WHDRC-CHETNA

Cc. Dr. Zafarulla Chaudhary, Gonoshathya Kendra; Sava; Dhaka, gk@citechco.net
Ms. Maria: International PHM : Brazil. marja@iphcglobaLorg

Dr. Mira Shiva;VHAI;New Delhi, vhai@unv.emet.in, vhai@vsnl.com, vhai@dei2.vsnl.net. jn,

vhai@WBn_Lcom

CHETNA
Lifovatiben Lalhai's Bungalow, Civil Camp Road, Shahibaug, Ahmedabad-380004, Gujarat, India
Phone: (+91-79) 2866695, 2868856, 2865636; Fax: (+91-79) 2866513
Email; chetnaOicenet.net: Website: www.chetnaindia.QrQ
Realizing the right to accessible, affordable and quality healthcare for every human being, CHETNA being a
Communication for Health India Network (CHIN) Secretariat has coordinated the process of oeveloping the CHIN
website. On behalf of CHIN, we welcome you to visit our site www.chinindia.grg

1/28/04

IBHIHII

IM

Ensunng Safe Homebirths through
Das (India's traditional midwife)

/t case presentation by CHETNA
In a rural village in Barmer, Rajasthan, Rasi was giving birth for the sixth time. During her
pregnancy she did not receive anycare. lichen she went into labor, the village Dai was called. In
the midst ofher labor, she started to bleed. The Dai immediately asked her husband to arrange for
a vehicle, as Rast’s condition was critical. lie walkedfor half an hour to reach the nearest town
where the nurse stayed. Somehow he managed to find her and by the time they both walked back
to Rasi’s village, it was too late. (As told by Rasi's sister in lav.' to an NGO worker in Rajasthan)

The above real life incident sensitively depicts the reality of most of the women from the
remote, rural areas of Rajasthan. The story also resonates the national level reality of
births in India. Recent data calls for urgent action for safe motherhood.
*
*
*
*

About 65-70 percent of births in india are conducted at home.
Non-medical professionals conduct 58 percent of births.
Dais conduct 35 percent of home births,
90% of ever-married rural women (15-49 year) do not have a hospital in their own
village and 40% are more than 10 kilometers away from a hospital.
46 percent women in India are not involved in making decisions about their own
heaith.
40 percent women do not have control over some money.
40 percent women in india do not have regular access to mass media

I Maternal mortality rate has not reduced in the past decade. The Maternal Mortality rate in i
India is estimated at 540 per 100,000 live births.
'
...National Family Heaith Survey (NFHS-2J597-98) j

The situation becomes grim when we come io the state/region ievel. in difficult, distant
areas, these figures cause even more grave concern. Community needs assessment in
six rural, desert districts of Rajasthan indicate
*

*

*

81 percent of births were conducted at home.
Dais conducted 75% of deliveries, ANMs conducted 9%, and family members
conducted 7 % deliveries.
On an average, the distance to the nearest Primary Health care Centre ranged from
1.5-50 KMs and 20 mins- 6 hours.
The distance of the district hospital, where emergency obstetric care was available,
was 13-100 kms and 1- 7 hours far away.
(CNA for implementing Safe Mothemood Campaign; CHE i NA-June-December 2001)

CllEINA

i

In most of the rural remote areas and urban slums of India, home births by Dais is a
reality, Institutional deliveries seem a distant reality, as the infrastructure required is
immense. The need of the hour is to make home births safer at the same timework
towards improving infrastructure. In the Indian context, Dai is a critical actor to make
hnmehirfhe

i<= ?v?il=fh||?

J1U___

the community levs!
Who is a Dai?

The Dai is India’s traditional midwife, also known as Dai-maa in Hindi, daayan in Gujarati,
suin in Marathi, Pathichhi in Malayalam and so on. The word Dai stems from daayimeaning'one who gives'. It closely resonates the term Daakin, daain-meaning witch.
----------------------------------------- -----------------------------------------------------------------------------------------The socio-cultural profile of Dai

The village Dai is usually a middle-aged woman of dalit or oppressed caste and poor.
Certain subcastes of oppressed and dalit castes perform the role of Dai. The Dai is a part
of the balutedari or the Jajmani system, by which trades in Indian societies have been
traditionally arranged. In many states, women of naai (baiter) caste may perform Dai’s
role, perhaps because of their relation to surgical trade and instruments. Like wise
women cf chamaar (leather-worker) and based (bamboo 'weaver) are dais. Among Rajput
Thakurs, elderly experienced women of the extended famiiy support woman during
childbirth; Dai are often called upon to cut the cord and dirt. In most communities, she is
an integral part of community health and healing systems.

Most Dais have not received any formal education. Hence they are usually illiterate or
barely literate. But their learning from life experiences provides them the necessary skills
through apprenticeship and experience.

A Dai usually has the experience of giving birth to several children of her own. As a
young girl, she accompanies her mother, mother-in-law or an elderly aunt and learns
while observing the process. Gradually she starts assisting her and later on, when she
gains confidence, delivers the child on her own, with another learner by her side. Usually
the Dai have 5-10 years of informal learning exposure, before she practices on her own.

My kaki sasu (husband’s aunt) used to do this work. When I married, she assisted me in
giving birth to my children. I got interested and started accompanying her. She used to
tell me how to manage the labor and I used to do the odd jobs. Then she used to let me
handle the labor during various stages. She once asked me to conduct a delivery in her
oresence. as she was not feeling well. This aave me the confidence and form then

onwards since past many years, I have been doing this work
.. .A Dai from rural Gujarat State

ClUilNA

4

However, some Dais start the practice earlier when they are compelled to support a
woman in crisis and then continue the work in interior communities as philanthropic
gesture.
On my way to the nearby village. I saw a woman going in to labor by the roadside. I
panicked but could not leave her in that state. I am a mother and have the experience of
giving birth-1 said to myself. I supported her to a secluded spot and assisted her. I used
ihe ciesn srsa of my sarj iQ r/e the cord and the blade of a grass to cut the cord. The
mother and her baby are hale and hearty even today. From that moment, I got the
confidence of condiictinn deliveries I decided, tn heln women of mv villsnes in such e

Page 1 of 1

Main identic
From:
To:
Cc:

Sent:
Attach:
Subject:

"chetna” <chetna@icenet. net>
<phmsec@touchteiindia.net>: <secretariat@phmovement.org ;>: <sochara@blr.vsnl.net.in >
<gk@citechco.net>; <maria@iphcg!obal.org>; *vhai@unv.ernet.in>; "Meera Shiva"
<vhai@.vsnl.com>: <vhai@del2.vsnl.net.in >
Wednesday, January 28, 2004 11:56 AM
,
- n
Daigadh2.doc
Paper on safe homebirths and traditional midwife

pi tYq^j^^iy^chfei ind ta. net secreiariatt^phmovement'ordR

*

' 'pp08??> 8pp ' '

Dear Dr. Ravi Narayan,

Greetings from CH ETNA!

i hope you must have received my comments on the dai (traditionai midwife) issues for the Mumbai declaration.
Based on the discussion with Ms. Maria and Dr. Zafarulla Chaudhary, it will be an excellent idea to cal! a meeting
of traditional midwives during December 2 -4: 2004 at GK Savar. This would be a valuable contribution of PHA in
protecting the rights of Dais who are generally from the vulnerable groups.

I am attaching a paper on safe homebirths and traditional midwife. Please send us your valuable feedback and
comments on the same. This could be modified to include traditional midwifery from other countries.
Looking forward to hearing from you

With warm regards.
Vd. Smita Bajpai

4?

Coordinator (Traditional Health and Healing Practices)

WHDRC-CHETNA

Cc. Dr. Zafarulla Chaudhary, Gonoshathya Kendra; Sava; Dhaka, gk@citechco.net
Ms. Maria: International PHM : Brazil. maria@iphcg.lpj?a!,.org

Dr. Mira Shiva;VHAI;New Delhi, vnai@ijnv.ernet.in, vhai@vsnl.com, vhai@dei2.vsnl.net.in,
vhai@V6nl.com

CHETNA
Lilevatiben Lalhai's Bungalow, Civil Camp Road, Shahibaug, Ahmedabad-380004, Gujarat, India
Phono: (+91-79) 2866695, 2868856, 2865636; Fax: (+91-79) 2866513
Email: chetna@icenet. net; Website: www.chetnaindia.org
Realizing the right to accessible affordable and quality healthcare for every human being, CHETNA being a
Communication for Health India Network (CHIN) Secretariat has coordinated the process of developing the CHIN
website. On behalf of CHIN, we -welcome you to visit our site www.chinind.ia.org

1/28/04

/‘Ila, 1 /V < a
A IX- X X ‘•X X

What tarnishes a Dai's imaged
The reality that the Dai is locally available and affordable resource can be flagged in any
corner of the country. I heir social and cultural bending with the community is a great
asset, which enhances their accessibility. The or. the job training which she receives
from one generation to another is an ideal learning situation to build up her skiiis.
However little is known about the standards of this training, which authorizes them to
function as a full-fledged Dai. We are also unaware about what are the gaps and
limitations in this training.
Often field evidences are used to condemn Dai practice and holding them responsible for
delay in referral. This delay is due to the fact that the Dai often operates in absence of
institutional health care and when she makes referrals they are largely not respected by
the health system. Hence most of the time she tries to do the maximum and then gives
up when there is no hope.

Holding them rAspcnsible for high maternal mortality and morbidity is unfairi Dai is the
one tfvliu vvGiko in absence of adequate, accessible, affotdabie and acceptaole iieakii
care support. I imely obstetric services are essential to avert maternal deaths.

She can support the institutional health care system to reduce maternal and infant
mortality However, she is hardly integrated in the mainstream health care systemignored and neglected.
Ironically the Indian Systems of Medicine (Ayur/eda, Siddha, Unanai etc.;, which has
strong iinkages with Dais practices, rias made hardly any efforts Io promote this crucial
resource at the village level.

In any field there are rational as well as irrational practitioners. Blaming the entire system
due to irrationality of few individuals in unethical. There is a need to give a fresh look to
roie of Dai in safe motherhood particularly for normal deliveries at home.
l^wic VI

itt OciH’Ci ■ IVi t

II

In the Indian context and other South Asian and Southern countries, local healers and
Dais have been taking care of community health for thousands of years. Home births are
a reality and a strong backup support is esseniiai io utilize their input up to the optimum
level.

Various international authorities have emphasized on integration of Dais for community
health.______________________________________________________________________
where the use or i BAs is strongly rooted in local customs, it is beneficial to:
- Train TBAs to avoid harmfu! practices during deliven/ (and use good practices), ,
recognize danger signals and refer complicated cases io higher-levei care.
» Establish or strengthen linkages between TBAs and the formal maternal health care
Sy S k—.. i



Ensure that health centers and hospitals will accept referrals from TBAs. (WHO |
information for Safe Motherhood, World Health Day, April 1998)■

CIIETNA

8

Where TBAs account for significant proportion of deliveries, safe motherhood programs ;
should include aciiviiies aimed at providing adequate supervision and integrating them in
in the hpalth system.” (Wnrl'ing group discussion OP the mie r.f TRAs -report on the -Safe !
ivtvu ici iiood i tiCi’inical uoi'iouHciuori, Gctober 5997, Colombo, Crilanka)
I
However, most Dais continues to work in isolation, without being integrated in the
Primary Health Care System.
Experiences from SriLankaJ China, Cuba and Malaysia indicate that establishment of
community based maternal health care systems comprising pregnancy, delivery and post
partum care and a system of referral to a higher of care in the event of obstetric
complication is the key to ensure safe homebirths. In Bangladesh, services of a trained
midwife (not Dai) have been usefui io make some difference in maiernai mortality.

Community Midwives and Skilled Birth Attendance in India
Realising the need for skilled birtn attendance at every births, the Government of India
has piloted the project of community midwives in selected states of India.
While there are no doubts about the need for skilled birth attendance for every births,
there is a need to keep the filed realities of most rural areas in view. Dai is a available,
affordable and skilled resource in distant, difficult areas and where there is a tradition,
there is a need to strengthen ppH support the Dai through the services of skilled midwife.
Gome concems in this approach are
s meeting the requisites for training at the village level,
* caste and class ievei issues that influence access to care- a literate woman may
come from UDoerclass and caste and may not want to serve the lower caste
women and vice a versa,
ensuring ner ail time availability in rural areas
providing back up emergency obstetric services
• ensuring supervision and rationality in her practice

in most of the rural and remote areas in India, the services of skilled /medical
professional care is not available and traditional midwives are the only source of care.
While on recognizes the right of every woman to access quality services in the public and
private health systems, one cannot afford to ignore the Dai.
Even the WHO recognizes the role of dais for ensuring safe homebirths.

The traditionai midwives can provide culturally appropriate nurturing in the community
setting; offer a first line or link with the formal health care system. However, Dais alone in i
absence if back up frem a functioning referral system and support from professionally :
trained health workers, cannoi ensure safe motherhood. (Heauciion of Maternal Mortaltry
i ioint WHO/UNFPA/UNICEFA^Vorld Bank statements

CIIETINA

9

•China, (rural) and rortezeia, Brazil were able to bring their MMR to 115 and 120 using
the following model:
1 Dai (lay provider) recognizes complications
2. Family or provider organizes access to essential and emergency obstetric care
3. runctionina emergency obstetric care available
°Dai's ni in iiii^ mud&i emd con biyfiiiicdniiy reduce maternal mortality -up to 100 but not
fess thsn 100 per 100.000 live births, if backed up by appropriate referral systems

Ds/’e ensure Safe home births m rural, remote, diffioult areas of India
Dais provide affordable basic maternal care at home
in half of the major states of India- Assam. Bihar. Haryana. Himachal Pradesh. UP,
Kerala, Punjab, Rajasthan, - the total cost of treatment to the patient (excluding cost met
by the government) is higher in the public hospitals. While public institutions provide
normal deliveries with a nominal or no cost to the user, the opportunity cost (Rs.57 per
day per patient (source: NSS, 1992,418,437,516; Independent commission on health in
India) makes deliveries in a public iiisuLution. costly. The poor utilization of public sei vices
further increases the cost of care. Loss of daily wage, transport, supplies and other costs
to the user make institutional deliveries almost unaffordable.

Dai’s are available in the village level. They are the critical link between the community
and the public health care system. If trained properly, the clients can avail essential
obstetric care at home: leading to minimal disruption in her day-to-day living A normal
homo delivery by a trained Dai would cost anywhere between 100 500 rupees. Hence
strengthening Dais is an affordable strategy tor safe home Dinhs.
Dai supplements the human power needs for safe motherhood.
Considering the population covered by a PHC in rural area, it can be estimated that the
PHCs conduct 000 deliveries per year. Of these, 15% have chances of going in to
complications. Therefore a minimum or 750 normal ana 150 complicated deiivenes are
estimated at the PHC. Most PHCs in India are hardly equipped to cater to this load.
Estimating one Dai per village there are about six lakh Dais who facilitate 80-90 percent
of childbirth in remote, rural areas and 40 percent of births in urban slums.

?n a situation where distances to public health care and availability of trained gynecologist
and obstetricians in the rural, remote areas is a concern, Dais can supplement to the
human newer needs of nubile health care.

/-«TTr?TXT A

10

R^ic heir, ir. rerinr:nr, HeS^vc in referral?;

’ he District Hospitals, designed to cater to obstetric needs are allocated according to
population. The difficult geographical terrain and poor infrastructure obstructs easy and
■.IIIIWIJ

I LIOUlUI I J

III

VZ I

I I

I VKZVl ,

The Dei are acceptable and respected in the communities They play a critical role in
mobilizing the community at the time of complications and it empowered through training
and supportive sendees, can helo in early identification ot complications and ensure
timely and appropriate referrals.

Where Qnppnrted by hmetinn^ efficient, sensitive referral services, traininn Hais has
resulted ir. identincation of complications and ensuring early referral, thereby reducing a
number ot deaths due to two - identification of comolications and timely referral ot the
three delays.

Raiiid was yivhiy bit in fui n'/e? iiiiid iime. A lUCeti Dai faciiiiaieti her iabuf and Sna ga ve |
birth to a beautiful baby immediately, she started to bleed The Dai alarmed asked the i
family to rush for a vohioio and shift her to the hospital, which was an hour's journey.
Meanwhile sne siarrea giving Rama nuias ana assuhng ner. She was trained to keep her |
foot end elevated, which she did and wished she could somehow block the blood from
flowing.
i
: Ramu s family was running helter-skelter for a vehicle and half an hour's efforts yielded
no results. Luckily for Rama, the local NGO came for a visit. Realizing the threat to i
Rama's life they readily provided their vehicle. Rama's life was saved.
I
....
(A case study horn Aiwar disirici in Rural Rajasthan)
i / was trained as a Dai by a local NGO. A family called me for a consultation when their j
I daughter was seven months pregnant for the first time ef 17 years age. A look at 'he J
daughter and I could make out that she was very weak. I informed the family about the
: care to be taken and advised them for a hospital delivery. Respecting my suaaestion the I
2nd her baby
beby ere
family did so and their daughter and
are healthy and happy.
- Pa* frorn 9 rtfra-

State, j

I hfsve undergone training organized by my organization an wel! an by the primary Health
i ii^j'

jf<v»

tw fc/v* tunui/ iw'V'Mi

nti I

w

miiv*

*J i\>

/ vif */w» »s>_.

wmcn pose inreat to woman s are. wnen ! come to me nospitai me doctor always greets
n-,c> Grir-ii iif£»<■ ahrtHt /vr.7 bi/Ar/r shw s/i/h-Qf/wr /
?,ny problem In the field. 1 fee! good.
Vvii&u ihtf Ahiivi ^umtJ s> iu my vin&ge. she m^i cOfties iu my iiOu&e tmu iheii we yu
together for a visit to women that i have identified Some times i cal! the women to my
house ot or, appointed timo and wo discuss about pregnancy cure and numerous other
things. I ne nurse is very tnenoiy eno come regularly ana taKes good care of my women.
A Dai from nirai a area of Gujarat State

Ri5: MnhilhFPC th** Cr.mmunitu fr.r

IWothPrhnnd

vVunien die while Umiig bn*til iioi viiiy because Oi HiediCai uOiTipmuatiOiiS Out Que

iO

cumulative effect of social injustices which begins with female foeticide and infanticide in
to teenage pregnancies, distorted abortions and too many, too close and frequent
pregnancies mrougn out me reproductive years.
Fvf example,

F elvic disproportion

Ocphaiv-

vaii

be treated

medivally

but rvr

its

orevention there is a need to ensure proper nutrition to aids riaht from childhood. I his
calls for action to eleimate gender biases at the community level.

Dais being >odeny-cnHur^lly sccent^hle end helping communities end women during
their critioa; period odd times enjoys the status of a power figure. She can therefore
command respect from the communitv and become an agent tor social change.
I was a contract laborer in nearby quarry. The need for doing something for numerous
women and children dying during deliveries prompted me to participate in the Dai training
organized by my NGO. I started working as a Dai. Women and men were not aware of

me care dunne oregnancy. childbirth and after childbirth. / conducted meetings to impart
edacaf/on on these aspects. Now families in my village accept immunlza^on. Earlier there
i/i/cft? many food taboos but now limy have changeu end pay more attention to the

woman's nutrition

earner, tne ANM was irreamar: / went to the
1-4&

■JCi'i :'^r!

rri&i

fr>~ir~ri

irtrWSi

ano informed the medical officer about
thO A.N M

'.''/‘.H 00^.0 fe^U

hilt /Oi.'

WH!

how iu uuuutifiait! wiiii ritti, i wtini uuuk and uonduuied a 'jumrnuniiy rn&tiiifiu. i informtiu

ahwjt mv rote a.*? « r)=s:
<4

fl

.^Iso my discussion with th<=? Modins! Offinsr They took this as

----At

A fl! it/i .•=T^r^4=<’

qi/

f-.i :r

snd children get immunized

um t&'fieduie. vvUihchi ytfi it on iediets end DDK ieguiaiiy.

...A

ujai a i otaiC

n

i

CIIETNA

12

ScVq nnciu idii -vuciipui’ -ixcijaoi.i icih

The organization has implemented a comprehensive Dai training Program with a
continuous training component as well as service delivery by the organization. The efforts
for over a aecaae nave been posmve in Terms or reduced inciaence of neonatal Teianus,
irr.nmv^d rare dnrir..- nr^gn-enry and labor improved birth weight and extended outreach
in remote and inaccessible areas of five blocks or Uoaipui uistricl.

Sure -Barmer
The Organisation has been working in me desert districts. Dais is trained regularly,
willan^ level ^unnorf ic pnciirwt thrnngh women’s groups They have identified a male
iieCTi’.ij Imikz'.ioiidiy to assist uie Dai in referral sitOauonSj who aioO avMompanies> ii@i to ibe
reterra! center.
rzcai- i^auaur
Hswa HosaIv worked with shout 100 Hsis of their area Are. involved in regular capacity
L/uiiviMM and nave m ic?n I’.an ic»J ioCviuo Oi i ivi'i ic; biftiiS My liaiiicM u/aio.

arh
a Heid NGG in i’ne RCH programme, have (rained Dais in Bhiiwara district and
nrnaniTed them They heve maintained regular follow-up and refresher trainings for Dai

ms

UKMUL-biKaner
Have been associated with CH ETNA and build capacities of Dais in desert and difficult
districts of Bajju, Lunkaransar The activities are being monitored through their citizen’s
groups to ensure services

CHETNA-Ahmedabad
Dai is regaided as a sociai resouice. available, acceptable, accessible and aiioicable io
most women in remote rural and tribal communities Hence efforts are made to
strongthan this rsoGvircc base My MUMuing theii oapadiios.
1 QtyHu of n-si n>f5Ct?Ce£
in coiiaboiaiion with me Lok owasihya raiampaia Sanivaidrian Samiii iLSrSS) and 26
field based NGOs C.HFTNA conducted a study of traditional practices during pregnancy
iabGr after CrtiiMMjini, MOW mOim c«mm vmiim oe»io. moMMM z.0w won io« i csmm ucjio iiUiii id.
states ot India participated in the study to snare their traditional Knowledge and wisdom.

lilt? study deaiiy iiidiuaied iiit? pusiiivt? ivit? of iiadiiiunai piauliues and Dai iiadiiiun iu
strengthen health of women and children at the same time laying emphasis on supportive
primary ncaith caro and
vivGo.

!

f'in~TM \

11

9 Training r.f Raic
Duiiny ISuT-cm, Ci SET NA ucshicu wc:f 1Guv dais ii’vfii liic stales Oi x-jujafat, rXajastiiaii,

Uttar Pradesh and Andhra Pradesh. This was done in collaboration with local NGOs who
than tookk up the task of follow up and monitoring. A genera! feedback of these training
was in terms or:
s

Inor^^S^H sr-r.iar.fzihilitv anH nnadihilitw of Rai’c in ih^ nnmmi initw




a

inoieased seif-coriMdenve and seif-iespecL
Utilizing positive traditional practices durina pregnancy and labor viz., birthing posture,
herbal remedies, massage and so on.
Adopting aseptic measures during labor viz,, scrubbing hands, cutting naiis etc
I inking up with ANM and l-tealth center for care and referrals

-

AddreSSiiiM icpivuuCuvu irodiui piOOieiiio Oi vvOiitoii aiiQ iiioii.

i

j

j

These experiences rofiootod throe strong aroae for inten.'ontion viz.
i Recognition of Dai by me Governmenx heaim cares sysiem, integrating her with the
nrimarv hpsdth nsrA
■wM imoi ouai mil im
Q

i_/cii

pi av.iCco

iiviii

a

pMomvc

vievvpOiiii

ciim

ou ci ly u lofii! ly

1101

Knowledae based on traditional medical systems like Avurveda.
Suiiding capacities of trainers, supervisors and managers of the health p
develop a positive appioach iowaids me uai iradiiion.

A Argi uiJ ZSXl
Ihree IO Is were conducted durina which caoacities of 50 trainers' were strenathened.
These treinere’e have then oenduoted Dei training in their respective organization. Efforts
were aiso made to strengthen the capacities of trainers in the Government of Rajasthan
that they ran imnlemAnt
R^i training Program
i his strateay has enabled oraanizations and individuals to view Dai's from a oositive
engie and integrate her in the existing health programs.

d Lnnn term llnkafips
Long-term involvement with organizations implementing Dai ; -rogram was forged and
efforts were made to supoort and strengthen their activities, lhese include UKMULFalaudi, Sews Mandir-Udaipur, CUTS-Chiftorgarh, Astha -Udaipur and CASA Udaipur in
Rajasthan state. SAR i Hi- Godhia. uHrxuv-uhaiampui in vjujaiai.
These efforts enabled appropriate implementation of Dai training programs and
strengthen nGOs in training or Dais for women's health and development.
3.

DcvcivpHiM icdtiimu A>us

In order to support trainers to impart training to Dai’s CH ETNA has also developed a
mult; media training kit, whioh is useful to impart training to Dai:o, and has boon widely
distributed

I

Given the critical importance of Dai training Program, various NGOs alt over the country
are investing resources to strengthen the tradition.
SEARCH- GadhchlrolblViaharashtra
J
HIIMIMUIMI 11.00

a
4
>->.. <,
4!
> t i «Xzl VV> I I MOI I

r
OMIlt.llMUKV

5 aa4f->
VVIMI

4U»
Mikx

U
tIKxl)-/

T a- —
54, <>
I
» I UU I LIVZI 104 I

€>.,!►»
MH I

OJao^U
Mil U I

rvcntion with the heip of
Attendants, village health workers and paramedics.
team of 30 paramedical workers
i M-4v>

/P^4\A/c\ 9^ villpric> health wnrkc»re

pr,H



Trad if in n a! Ridh aHcindantc- haw£> Kc»c»r.

udiubu. .*-> neiu udSeG indi HiuiCaieu feduCiiUii in pi’icuiiiuina 11lUildiiiy cjiiu toicsi ciiiiuiiOud

mortalitv in 58 villaaes in Gadohiroli district.
Asnish i^ram rcychna nusi- Auranysbad-Manarasnira
H-sic- arid farnala health WArkorc arc> irainArl fr»r rr>rnnrohor.ci-v* Mr'r-I nrrinram TRAs are.
lesporioiblc for pfeyncmCy, denvefy and pOSi uenvci y C-afe.


5-2

■ ■

T'r-J 1«'T




w

r4 Jri •?> »-» »r* 1 t »-

D
■ «*««■

1 » « •-» r-'r»T

■ ■ >w>- • ■ av am • l|

**’•11 *** ■

*•

TBAs are trained ana providea with safe delivery kits. They are supported by services of
njrsii health workers and a rpfprral hnsnifal
I

Aga Khan houndation- Junaaacm, Gujarat
The principle objective of the project was tc establish a community based PHC program
to cater to the heaith needs of women and chiidren. The project was impiemented
through a thrPA tipr health service system-the first was the VHW and TRAs. An
evaluation of the Foundation project indicates that by training and utilizing the TDAs who
were already providing services to the community and in whom the community had
confidence, the services provided by the project became mere acceptable to the
communiiy.
* • .rn

n.

m

*,r"* - ? ? «•. «•

^vyyci IXUIUI- oiiuMUMiu-aujuiUs

In coordination with the communitv members. Anaan Wadi Worker and Primary Health
C
r-» 4
r
x-» %> ■ ■ «.» ■ > awtO

At>-ZAr>^tho’'''4 4|>XS
«.■
i •s-'--. a. ■ a >w»

rtnry
Jz ^>1 I V I I I lun iw V

Ma

•_zMz>>a_r.

a a ux «_»

»•
1 •-< •• ■ •_• Mz m«

vl Mi a I a a • I sj

4zx
a.>z

OSOZ.z^f T" D A
vz w t m m a
a a—«a »a_»

in iheir area and provided them with safe delivery kit. Training Dais has been positive in
fArms nf adnnfinn nf
>

mAssnr^s

I

<

incrAacpd nnverane of nrAnpts»i and nnctnatai
vz

I

«

vaic, IIM ’OGOCU ICHCHICil CTIIU IC'JUiJ.IVII HI I I CTO I ICT '.a I mOilCTli'.y.

QEZio* A r A

Have made linkages with the government of Gujarat for remuneration, recognition and
training nf dais Rai cnnnarativa.s have been formed to enhance their solidarity and
oil ci ip'-i 101 < Dcti oyoioiii.



l/cti

iiCTiiini«4 0G1 iOOi vvcto

ctioU

biaitcu ii i

vviiiC'ii piOvi'Joo a

3 month course tor women /uais who are interested in takmq up this work.
SARTni — Gwuiiai- Gujai csi.
The organization has trained RO local women, including Rais as MCH workers; Ruring
13S7-33; SAiTTH! Dais conducted 1430 normal home deliveries and referred 32 for
hosoita! deliveries. I hev also facilitated immunization of 810 children.

/•Ilf

IV

15

4

K.- A AJ-/ A x • A A

v.

«. a-t w jre

r^. s-» j*3

4-- e p

» -rf f ? «.

onqiiiiM ciiim tmczvyvMi rxti 154

CHETNA has been sharing experiences of Dai training at various national and
I> I

■ I > vx m s>-<



>«>z<vtl«>w

-Jix m I y





• sx sx

<a

v« • i» <-• k> > ■

n »

vz^> ■«-»

•_> jr »_» vsx i • > x_>

i

Ml

I I I^

ana policies. Related reports ana aocuments are shared extensively at the national ana
Qtate level with thp MGOs a.^ariamic.iar.c and GOs and linkages ar& established

!
A CC/i i ii i littee V/1

w^irxnig

iwi

i uvvyi ii ’.fOi i Of Dai i icio L/ooii iC/iiiiC/G

iUi

I
I
uic oiaio Of

Guiarat, which aims to work for recoamtion of Dai at the state and National level.

I
.
, - .
. ,
I
Seva rxuiai, a ieadiny NGO and rnernbei oi" ibis coniniiii.ee nave oigauizeu uai rneia ioi

IHa states nf Gniarat Raia^than and Madhva Pradftsh

un ii"' Adhi. National bate Motherhood Dav, a Dai consultation was oraamzed in
i ne GOG has studied trie cunicuiurn and recognized the Dais iiained by Seva-rxuiai.

Rhansali trust and .RFWA-Ahmadhad

I

A state level tasK force to look in to Dai and sate motherhood issues has been formed.

I
I

I

As a uuuidiiiaiiiiy agency, CHETNA is suppiiiiig NGOs in Danaskaiiiha, oabai kaniiia. j

Panchmahal and Mehsana districts on Safe Motherhood issues

in the state of Kaiasthan.
•—;h, I
CHETNA coordinated a campaign in seven districts of Rajasthan through 11 Communi
i based NGGs during 2001-2002.
I
/’•UCTM A

v>! !i_ iiin io

-

--------------- -x

C-»I ih.-i.n

CmVCmsCii lai v/i

XI---------

ivtr-1-! xcqtaou jcsm

< >.-----vui arxoi nt,

IVIQll iw«d

white RiPDon Alliance tor bate Motherhood.
A

siaie ievei puiiuy diaiuyue was uiyanized by oUMA on rxoie

hirthft at .Jainur nn 141" Nnvamhar 2003

CaiMMariCihari, statG

I
, ~ . . - . .
I
oi uei in oaie iioine i
I

UHt iNA is tne coorcnnatina aqencv tor impiementinq tne kQH awareness activities in
A i-.»;nr

nn

Knnrninnr

/-4 : eV inrr r>f D ri i«->

On

Advnrarv eff.nrtc

Since 2000, CHETNA is auvccHting for recognition of lire roie of Dai in safe
homo births. Thio inciudcs presentations and showcasing Dai tradition at
national and state tevel policy meetings and dialogue, media advocacy,
developing ari duvuCdcy paper . building capapcities of CBOs to implement
evidence basSed dai training programme > Since Januray 2004. CHETNA. is
advocauna for safe nome oinns at tne Internationa! ievei through the
; ’eOpiG 3

health

mOVeiTieriu

qjjG

qi

the

i^gik/nai

!S?v<7i

unvugii

ii

v v.

I

O

fr i

«=»

« *r-

P'jh’q foU? ‘e c^ntrsd for onsuriritj ssfo homo births in romoto. ruroi, difficult sroo.s of Indio,
hi lilt; ueOyiapiiitcii, SOvicti cinu uiiiluieii CuiittfXi of the iiiuiaii nlined DaiS foie aS a change

agent and in safe homebirths will be more powerful in the coming millennium

1



w

nf trad if Inna I mldwiferv*

While the tradition of home births by Dais operates informally, a need to develop
folt The National Institute of Health and
a
midwifery standards for Dai s is strongly felt.
Family
welfare nas pioneerea in developing a curriculum for Dai Training. Several NGOs nave
C’lrffel imc> Triesining Schedule etC.

-•!tii/C>lr-.r5or’ feftir r>tA/n cianHarrlc

There is a need to review the existing efforts and based on the experiences, develop
otandardo for traditional midwifery. The ISM and H department and Ayurvedic universities
ana insxiiuuons aiso neea io oe actively engaged in mis process.
2.

Li"u'iai’iCO SwnMui ity

cAaSui!L*t«»

At oresent. about six-lakh dais are workina in isolation without any formal oraanization.
There io a need to form Dai aooooiations, networks and cooperatives/ mandala. These
associations can become piatibrrns for building capacities of uai, regulating,
cipndarHi-in.- pnH : snnrsdinn th—ir prnfessinnpl skills And advocacy

3. strenothen Dais role in Safe homebirths
The Dai tradition differs from place to place. Various training and follow-up efforts by
NGOs nave snown mai mrougn adequate Training ana foiiow-up, me Dais can become
pHivp AHAnts r«f social nhpngp pnd provide hasfe hpplfh cprvines dunnn pr—gnpnry

childbirth and after childbirth as weii as other socio-medicai issues of womeri and
adolescents.
i he traditional roie of Dai needs to be expanded for health concerns of women and men
ac \A/pil as n&whnrn

l-fer Qrrppfenrp in thp community r«n hp nspfel in rpQ.-hing out

iO vvOi i ici : nt mih’ivuu oiiuaiiUiio, auvicbvcii*. mvvS dnu yiiiS aii’J iiicii hi pai nCuiai.

3. Ensure ucHiirriuniiy support

Whife Dai’s rnfe is nritfeal in Rafa hnmahirths; due to bar low stotns in tha sncial r.lass.^ote mi dci and hei sewiidcu y status as a woman, she is often iu treated and .gnored in
the society. Beina relocated a secondary status; her own health at times is at stake.
“T'^ro io a nood to support Dais in rural communities.

Fftrlipr naanle? r.firtiniilfsriv thes unnar clasa Rajputs useri to treat me fiddly snd ignore
-c. After I too:;'treeing form, thioorgonizstion'ond vroo supplied vritd o kit bug, people

nave stanea recoamzinQ ana respecuna me. (...A Dai mom a rural remote area in

*

s-

U25-

Flit: public dfiu pnvaie iiccaiiii vaic oyoicmo need iO be Sensitive iu ine owidS dopeCii? Oi

the Dai tradition, recognize the tact that she is a member of the oppressed and
disadvantages section of the society, '.yhich has been systematica!!y marginalized .There
is a neeu io proviae community social support systems through ensuring their social
eponritv greeting support groups end structures

4. provide oacK up services
:t :s evident from the existing scenario that unless the Dai is involved, it will not be
possible io provide a trained person for attending normal deliveries. Hence it is important
thflf Dpi’s pca rp-onnnizpH and infprwnvpn with thp pyisting primary hAPlth nPrp cyciprn
and avuv6iy iiivOivcd lit uic Pi SC.

Trained Dais need to be supported by functional and effective public and private essentia!
and emergency obstetric care services. Keeping Dai as a central community service
r.rnwidAr ^npripi Hnk-poAs with thp ANM. A\A.W pnd thp r^f^rrpl hApIth care system need

vvhiie Dais deuiy inaiiayes iiumt? biuiis. i-heie is a need io supeivise hei woik.
nerticularlv in tha ram.nto rural ar^as This would ^nahlis mr.nitonr.o .of nhiiHhirth pr.rl
i ov-uui»

k>i

anv

'jiol;! c}jg i iL/iuo

hi

uai

pi ci CiiCe.

111

«un laiiMi i

wvi’.ii

'.nc

mi^aiivvaui

worker and the ANM. She should be made a cart ot this system
6. rruviue oe^uriiy uiiuuyh invoivirig rariuhayai Raj in^iiiutiufi^
TKa rnlA of Panchawat is r.rstir.al tn strAnnthAn Rais rnln SnmA of IHa mAasurAs taken up
mv

i GiiL/i laya1. u'juiu

mo.

■ Makina a local inventor/ ot Dais and women /men who want to take ud this work
Forming and recognition of local Dai cooperatives
'

Ci eating untiling iiumes fui women
Allocating and land ^nd rAsonroAs for r.i iltiwAtion of herhs hsaH by Rais

'' OoGidinatiMM with the Public Health System to aiTange for training, supports and
supplies
' Displaying contact numbers and addresses of the doctors and referral centers as well
as utiiei leiated peisons at a common communiiy point.
z Fixing up a standard rAmur.Aratinn system for Dais
LiaSiOnii IJ^ VWIlll II IO |Om€4I II I I I CiOll MVlV! o Si iu diSiHOt aMinit iiou ctt!v/11 tQ Gi”iovii^ ci^pi wciOi i

ana communication facilities
/

C nc'i ir:n rr U/-?. ““i I tr~> ^r~. z-1 c'z-»z-»i

I ociz'iiriru to

f-’T-nr-T-XT A

0

X

i C-

X X I J. J.

*
/

<*

P&forsanrec*


udjpd! Gifiiua afiu

-Z

A \ £ O
I
r\ « «l«x
rv*>
] C? O O O
"T" irr> /-! «4«
•*>
• n« w ■
XAIU4

• S-* a • ax^ • a • *-* • a x^a a a •-« x-« x^ a
x-x x-r «
at <-« x* a vaxx « a

jl_ s iv.*-i

> cam. i tan'iing 0» I umo -Ct fi_ t Nn o LApciiciiuco.

CHE I NA: Ahmedbad: February-1989
i O yo
<% « ►**
a a a xx Vti WV a • a

<j*%
tl tx»

*-> **<x z>
xxa a xx vxi

XX a

K
O •*
I**
a e a xz VI a x-* I xza a a •••

Cniia Heann; lok Swastnya rarampara Samvaranan .^amni ana ChETna; madras;
■’

A.T.Ddtiy. ix.A.Datiy.C.C.CviHcikkcGtuic: Gcai vi i icain, iviaiiayun’ici n ’Jt vi hiui tuvu

Pneumonia bv traditional birth Attendants. Bulletin ot the World Health Oraanization,
4 004 “TO/^A.pOT HOC
a VZ W

t | t

4M •X^XZaXZXZX

XZWXZ

v

Bajpai Smiia and Sadgopai iviira; Her Heaiing Heritage; LSrSS-CHETNA; i995

Z

lUlsin-a R^hr.rah-* cpfa Mnfhi&rh.-.nrl PmnrArnc*
Ohfinnc pnrl Iqqijpc*- Cpnfrs* for
v/
4. . t
i v/puiaiiv/ii ciiru i Qi! iiiy i icciiui, ovi ivoi Oi i

umiiO i

icaiui, i avutiy v>i

iviovji^zii ic,

KyOtMi i iDia

University. New York: 1993
,/

*

>• Qz-s+<x r.nA4knrr>r»Ar!. n

a ’ z-xi * 11- • atta r a.t

lA/ssrlrl \A/iz4z-x A zirk ziti .-•

IrmiA OO-

-1 Ofifl

. VvhG, A juini VvHO7ui\irFA7uNiGEF /Vvurid Dtifik siai&mtjni. Ftruuuuun ui iviaitsintii
Mcrtfility; 19^^



L/cpaiti i tci it Oi 1100101 ciiivi i csiiHiy yvouaic, GOVei i it i tci ti mi k^ujaiot; Ovate?

imoiementation Han; Keoroductive and Child Health Froiect- (Juiarat state November
1996
*z The Safe ivioiheihood Auiion Agenda. Fiioiiiies Foi the Next decade- A repod on trie
Safe Mr.thprhnnrl T^nhnioal Cnnci Htptinr. Colombo,

I onks Ootohpr

1QQ7

Katidiiaii Gdiidiiya, nuiu ui ti id viiiape i icdiui vvOi n.diio/ uui iiiiiui iity i icakh vvMin.did ii i

HCH: a brief Overview: Indian Institute ot Manaqement- Ahmedabad- 1998
WHO; Worid Health day Safe Motherhood information kit 7 Apr:’ 199S
Naik Karnai; Concept paper t^ai training School at Seva; October i 999
Z

Npik kamsd* Rfannrt r.n Rai trpininnKZiiiata*. o’.atc,

<

MMI MCil V

Ir.QtitntA r.f HA^Ith and Fsmilv'A/olfer^

I O-C’C-’

baiDdi smita and ent I na i earn: I he Dai I radition; CHE I NA-Ahmedabad: October1999
Z Teuiinicai Support Unit, uNrrA-Deiiii- irnpaci or” Dais iiaining on leproducdve Heaitii.
Fvid^nr.*^ frnm Indian Fynsiri^ncA^ ,lnlv 1QQQ

Tadh«on; lOtfi /-Ipt II- £-\J w
che I NA: workina towards recognition or Dai tradition: a reoort on a consultative
K_>! | IllCi C<I IM

Ml I L_ 1 11/—.

mz-ir^Tir-.z-r <~.r MiOOic* •. a

/

I MCI I II,

>L: n z i

M U MI IVj tl I Ml 111 I^J

PiAiC tir

11 IM

JTIO Atl I mirljunwooV Z

J £T\i "i

z*

r*?<~ad r?

iviay -20uu
Raioai Smita- Patel Pallavi and CHFTNA Team1 Role of Rai in the RCH program/■*i if----- Twin
AL—,----- s_i----- ------------------------------ v. - ----- A Cxtyfy
Ml IS— I I*/—.“/“^l II UMMGIMCIM, |■»MWMillMMt
I MMM

y

capoor mou. NaiK carnal ana baipai bmita: Fcoie or Dai in women s neaitn: cheTna-



Ahmeda^ad" Nevernber 2000
Capuui iudu and oajpai Smile, i he Suuiu-CuHuiai Context of Dai, CHE i NAAhmedahad- Nnvamber 2000
Jridicirs IIIOIIIUIV Ml »

MpMtM tlMI ■ V-«MIMI IMMM,

I I IM I'fUtlvi itll I

Mumbai. 2uuu

v

M

» C"* • •
i

r
• ' ? • • ! 1?5

v\sy a., in x>,

IT-

LIVING

CIVIL SERVICES

Give Dowry, Will Marry
The latest crop of probationers don’t believe in pretences. Not only do they see their jobs
as a passport to higher bids in the dowry market, they are also brazen about it.
By CHARU LATA JOSHI

I

I paid Rs 18 lakh for my sister's
wedding. And since I am in the civil services.
I will command my price. I will take at least
Rs 25 lakh.
—Neeraj Singh. Indian Revenue Services
probationer, batch of 1993.



The rates for an ias or ips officer from
Bihar are astronomically high. Parents
ki to lottery lag jati hai (the parents
of these male officers literally hit
a jackpot).
—Manisha Chandra. Indian Revenue
Services probationer, batch of 1992.

I
I

‘I I

I

Probationers are taken out for dinner
and lured by open offers.
—Sandhya Manchanda, In­
dian Administrative Service.
batch of 1993.

I

"Tl
fhe new generation of ad11 /I ministrators: all set to join an
I 1 / I elite tribe known for its high
-L V JLideals. elevated standards of
intellect and concern for social justice.
They clearly knqw their worth. But
these confessions are hardly a revela­
tion. For years, civil servants have com­
manded a high premium in the mar­
riage market. What is new is that they
are now brazen about it. And dowry
rates are going through the roof.
Each time the country performs its
periodic wallow in the primeval slime of
casteism and communalism—as is now
happening—there is the ritualistic ex­
pression of righteous concern. But an
equally ancient and debilitating social
evil like dowry continues to perpetuate

itself as a parallel blot on the country's
social fabric.
Shamefully, the practice is rampant
among the very people who bear the
burden of translating progressive gov­
ernment policies into reality—the elite
core of administrators. Here, rates are
determined by the highest bidder and
they vary from state to state—ias grooms
in Andhra Pradesh and Bihar could cost
between Rs40 lakh and Rs 1 crore—and
within states, prices depending on cadre,
branch of allied service, caste, status of
the families, and most recently, the vir­
tues of the bride-to-be.
Interviews conducted with proba­
tioners from the 1994 batch of the civil
services show a disturbing trend: the
essence of marriage for many of these

I

“Once in the field, frustration leads to cynicism and

i'

Photoqraphs t>v BHAWAN SINGH

11
‘ I.-:

J

■i

' TH

J ’

i

'
.

..

%

i

II
I

1

“Most dowry deals are made in
cash or consumer durables.”

Men who make it to the allied services
aim for a Rs 50-lakh dowry.”

ALOK KUMAR, L j, IPS probationer, 1994

RAiNJANA HAND A, 23, IAS probationer, 1994

i ■

'

'

122 INDIA TODAY ♦ OCTOBER 15.1994

J

’’t

4
I

J

bureaucrats-to-be is based, not on con­ instances of unattached probationers
cepts of equality and sharing, but on being approached by highly-placed
suitcases of cash, acres of land and tolas businessmen."
of gold. And this pro-greed creed is
/^ENIOR bureaucrats explain that
justified by the social sanction it gets
the new hike in dowry rates is on
from groom-seekers.
two accounts: firstly, for most salaThough each community may have x .
ried employees, resources don't
culture-specific prejudices regarding
dowry, the civil servant enjoys universal match desires, so a handsome dowry fills
appeal. An upper caste Brahmin irs in the gaps. Secondly, among the
probationer from Uttar Pradesh talks of unorganised, semi-rural segments
how a minister in Mulayam Singh which hoard undisclosed wealth, the
Yadav’s Cabinet offered him Rs 1.2 crore primary issue is acquiring social legiti­
as dowry: 1 crore for clearing the civil macy in tune with their economic sta­
services exam and 20 lakh for being a tus. They seek to do so by ‘purchasing’
Brahmin. In Andhra Pradesh. D. an ias son-in-law. A fact illustrated by
Pullaiah. a former Congress(I) MP, S.N. Prasad, a retired civil engineer from
openly says: "The sendees may not hold Lucknow, who recently married his
much charm for the youth these days. daughter to an ias officer: "You don't
get an ias son-in-law for free. If you
’ for us Kammas. having a son-in-law
ie ias is a matter ofgreat prestige. An spend Rs 40-50 lakh, you can get some­
body in the Uttar Pradesh cadre.”
ias Kamma bridegroom is guaranteed
Most ias and ips officers from among
more than Rs 1 crore.”
While the civil services have enjoyed the Reddys and Kammas in Andhra
an aura of symbolic power since pre­ Pradesh carry a price-tag of over Rs 50
Independence days, the unembarrassed lakh. Interestingly, these huge demands
pricing of a probationer is new. Says are made by officials working in depart­
D.D. Sathe. retired chief secretary. Ma­ ments like excise, commercial taxes,
harashtra: "Way back in 1941. during civil supplies and municipal administra­
our training, dowry was not spoken of as tion where the chances for corruption
casually as it is today, though there were are high. "The entry of children from

corruption.”

N.C. SAKSENA, director, IAS Academy, Mussoorie

business families in the senices, who
feel a government representative is use­
ful. has led to higher dowry rates." says
V.K. Srinivasan, principal secretary
overseeing public enterprises in Andhra
Pradesh.
While escalating dowry rates may be
considered a phenomenon limited to the
cow-belt and Andhra Pradesh's cash­
crop communities, it still poses a grave
problem in the civil services. A regional
break-up of probationers of the 1994
batch shows why: out of the total 291
probationers, 16 are from Andhra Pra­
desh. while 50 belong to Uttar Pradesh
and 40 to Bihar.
The ‘price’ of a civil servant groom is
decided by market demand. As a proba­
tioner from the Indian Audit and Ac­
counts Service, batch of1994. explains in
economic jargon: when demand out­
strips supply, the commodity gets dearer.
"Every marriage season, the dowry de­
mand rises by at least Rs 50.000. And
once the Mandal recommendations are
implemented the rates will shoot up.
forecasts Hari Shankar Lal. an irs proba­
tioner from the 1993 batch.
Recent trends indicate that as always,
the ias reigns supreme. Close on its heels is
the ips. But in both cases, the cadre
remains the determinant. Sanjiv Dubey.
ias. batch of 1987. was sorely disap­
pointed when he was allotted the Kerala

I

i
I

I

j
I

I

I

fe4
I


•p ■■

ri

I

“Groom-seekers have actually paraded
their daughters in the Academy.”

“When I got married I was just an
engineer, so I did not get my price.”

ROLI KACCHHWA, 24, IAS, 1994

ANAND BABU, 28, IFS, 1994

i

OCTOBER 15.1994 ♦ INDIA TODAY 123

I

I
j

LIVING
cadre. But probably not as much as his
mother. Nirmala. who talks about the
cars lined outside their house in Patna
soon after Sanjiv cleared the exam. "But
he has not married yet because he
wanted to marry a good-looking ias girl
from a different cadre, in order to get a
cadre change." she says.

1

bhawan Singh

1

"T" "TERY often, the bride's parents
1 / make the investment in an ias
1/ groom, not for prestige, but for
V purely mercenary calculations.
| The premium is on allied services like
income tax and customs and excise, as
against audit & accounts where scope
for corruption is limited. But, there are
other reasons which contribute to the
rising importance of the revenue ser­
vices. "The irs also guarantees a more
settled life in state capitals, as against
district postings in the ias," says Ram
Mohan Singh, irs, batch of 1993.
Negotiations
for
accepting
I purposely got married as an engineer.
dowry—a ‘misconduct’ in the Central
services conduct rules—are blatant.
Purely to avoid giving or accepting dowry.”
"In Bihar, the boy’s family is direct.
They produce a list of demands and
YASHWANT JETHWA, IPS probationer, 1994
discuss the cash settlement before any­
thing else. ' says V.P. Sinha, a professor
in Patna University, who recently pro- future."
comes a part of the family corpus. The
cured an ips husband for his daughter.
But if such dowry is meant to be
Subodh Pandey, ias, batch of 1993, streedhan, the way money changes daughter-in-law generally does not
justifies this: "Since government sala­ hands from the bride’s father to her stake a claim to it."
Despite the facade of mutuality, giv­
ries are low, the parents-in-laws make father-in-law appears dubious. Ex­
ing
dowry is never consensual. "The
arrangements so that their daughter plains an insider: "In Bihar, the cash
has some sort of security for the goes to the patriarchal head and be- entire exercise is very humiliating. In one
instance, the ias boy’s father directly told
us that we will not be able to afford them, ’ ’
"Rates in our community are not so high. An ias sonsays Uttara Pathak. a Bihar pcs officer,
who has been Hying to match-make for
in-law may cost under Rs 10 Iakh.”S. THAKUR, Priest
her daughter for the past two years.
PRAMOD PUSHKARNA
In the absence ofdirect
parental liaisons, middle­
men take the initiative.
Pandit Sharada Raman
______ Thakur works as a priest
at a Hanuman temple in
■T54* ' ■■■ ’ '> J De,hi University. But his
real business is to keep a
tab on civil service aspi­
rants in the university,
particularly from the
close-knit Maithili Brah­
min community. He then
liaises with parents on the
look-out for civil servant
grooms in Bihar, fixes a
price and settles the alli­
ance. all for a fee. "Rates in
our community are not so
high. An ias son-in-law
could cost under Rs 10
lakh," he explains.
A direct approach, as
many
groom-seekers
have discovered, works

'

jfthansa’s
hanging
all. That’s

I

I

sgSIS

OCTOBER 15.1994 ♦ INDIA TODAY

127

91/92/93

.............

LIVING

rare
Is like

, five
:arry

r out
ture
oil

to

col.

Ps

0%
illy

ds
ists

col
50
ds
ity

11

tai

at
le

jIs*

II 178-94

best. There are interesting episodes of of the 1992 batch.
tered by the initial rejection. “If becom­
car-loads of match-makers arriving at
Dowry-takers leave their tell-tale ing a civil servant is the barometer of a
the academies in Mussoorie. Nagpur traces. "The opulence in the homes of person's success. I will command my
and Baroda, after looking up the civil married probationers in the academy price.” he says unabashedly.
services list of eligible candidates from is jarring.” says a senior faculty mem­
This social Darwinism is not totally
the upsc centre in Delhi. Says Sandhya ber of the National Academy of Direct inexplicable, especially when pro­
Manchanda. an ias probationer, batch Taxes at Nagpur. “We don't maintain moted by neo-Brahmins in a semiof 1993: “There’s nothing clandestine. such a high standard of living despite a feudal set-up. And in a society where
The probationers are taken out for dual income and 10 years of service,” the red light on a district magistrate's
dinner and lured by open offers."
adds another.
car works as the most visible symbol of
But women candidates are not so
Not surprisingly, in many cases. authority, evoking instant fear and
sought after. Amita
respect. Says Alok
Sharma, irs. batch of
Kumar, ips probationer,
K M, KISHAN
1989, comments wryly:
batch of 1994: “It was
“I have observed that
basically the ability to
among the car-loads
assert authority over
which arrive at the
people that drew me to
academy, none seeks al­
the services.” One com­
liances with lady proba­
monly expressed view is
tioners.”
Manisha
that if a civil servant
Chandra, irs proba­
starts a career backed by
tioner. batch of 1993,
solid dowry money he
recalls an incident last
will have less of an in­
year. She and her fiance
centive to cheat because
Amrit Kalash. an ips
he will then be economi­
probationer from the
cally on a par with peo­
same batch, were ac­
ple he has to deal with in
costed by a person who
the private world. But
claimed to know Kalash.
evidence shows that this
After the preliminary in­
is wishful thinking. Not
troductions, he pro­
only because most of the
duced a blank sbi
dowry is kept by parents.
cheque, asking Kalash
There is another
to fill in his price.
deeper, sadder, psycho­
In other instances,
logical upshot. Once an
the demands are more
officer compromises his
subtle. “The term com­
idealism at the outset by
monly used is a ‘decent’
falling easy prey to cor­
wedding, but this indi­
ruption. he has already
cates that even if there is
ignored his conscience.
no cash demand, the ba­
-The next compromise
sics are to be taken care
|
comes much more eas­
of.” says Usha Jha. who
ily, the logic being that
is presently scanning
even if the dowry pro­
proposals for her daugh­
vides the basics, a regu­
ter. a medical student.
lar monthly income is
An air-conditioner, a

Now
my
son
commands
more
respect
essential to live well.
Maruti 800, colour tv,
There is hardly any at­
than the Maharaja of Darbhanga’s son.”
vcr, home appliances
tempt to disguise this.
and furniture, are con­
Pankaj Singh, an irs
PARMESWAR
PASWAN,
father
of
an
IRTS
officer
sidered the just due of
probationer, batch of
even a clerk. So, the ar­
1994, is blase: “If the
gument goes, why should the civil the lure of dowry that could be pro­ Government pays me a house rent
servant be an exception? In fact, in over cured. along with other attractions— allowance of Rs410, and expects me to
50 per cent of civil service alliances, the instant attitudinal change towards live like an officer. I'll definitely strike.”
these demands hardly change, how­ successful candidates being one—
As N.C. Saksena, director. Lal
ever accomplished the bride may be.
heightens the clamour for the civil Bahadur Shastri Academy. Mussoorie,
But avarice is not all. A small, but services. Neeraj Singh from the irs, puts it: “The probationers are very
sincere lot, does believe that taking a batch of1993. was a flight lieutenant in idealistic while they are at the acad­
hefty dowry will involve a compromise the iaf before he cleared the civil emy for two years. But once they get to
on the attributes of the bride. “The services. He recalls how there were the field, frustration leads to cynicism
trend among the more thinking proba­ hardly any marriage proposals for him and then to corruption."
tioners is to go for an influential family, earlier. “Relatives in Bihar used to call
—with bureau reports
readdaughterofasenior bureaucrat or me a sipahi (soldier).’’ he says. Post civil
politician, instead of a hefty dowry.” services results, the situation changed (Some names have been changed to protect
says Sukhesh Jain, an irs probationer drastically. But Singh remains embit- the identity of the individuals.)

JT

OCTOBER 15.1>W4 ♦ INDIA TODAY

129

t/jH"\ I*I
*>*

POLICY DOCUMENT
GENDER, WOMEN AND
DEVELOPMENT

i



HIVOS POLICY DOCUMENT
GENDER, WOMEN AND DEVELOPMENT

Adopted by the Executive Board (Dagelijks Bestuur)
at its meeting of October, 1996



f’'"'
• -r

i./

Who Remains Standing?

}

I ■

First,
erase your name,
unravel your years,
destroy your surroundings,
uproot what you seem,
and who remains standing?
Then,
rewrite your name,
restore your age,
rebuild your house,
pursue your path,
and then,
endlessly,
start over, all over again.

I:
Andr6e Chedid

Translated from French by SamuelHazo and Mirbne Ghossein*

r

*

-

HIVOS POLICY DOCUMENT
GENDER, WOMEN AND DEVELOPMENT

1.

Introduction
1.2

The need for a review
Hivos Women in Development policy in retrospective

1.3

The shift to Gender, Women and Development . . . .

1
1
1
3

Hivos development vision: A Question of Power
2.1
Marginalisation and the focus on unequal power relations

5
5

1.1

2.

2.2

Hivos Gender, Women and Development vision: women's empowerment and

addressing gender inequality
3.

Looking at the World Through Women's Eyes
3.1
Economic globalisation

4.

3.2

A global women's human rights framework and its backlash

3.3

The women's movement as an actor in civil society

.

The integration of Hivos Gender, Women and Development policy into sectoral
policy priorities

4.1
4.2
4.3
4.4
5.

.........................

Economic self-reliance

Culture and Development
Human Rights and Aids

..................................................................

Environment and Sustainable Development

Operationalisation

5.1
5.2
5.3
5.4

Strategic choices

.....................................................................................

Women's only counterpart organisations

Integration of gender into mixed counterpart organisations

5

8
8
9
10

12
12
14
14
16
18
18
18
19

Gender capacity in organisational development, organisational change and
institutional development

21

(

6.

Hivos institutional strategies
Policy, programme and project work
6.1

Annex I
Annex II

Notes

Hivos institutional state of the art
Statistical Overview (1991 - 1995)

23
23

26
28

32

1.

Introduction

1.1

The need for a review

This policy document is a review of Hivos policy intentions and priorities regarding 'women and
development'. It incorporates Hivos' experience with implementing the 1988 policy, 'Women in

Development, an essential component of Hivos project work' (WID policy). The reformulation of
policy is motivated by new challenges:
Hivos evaluations of the WID policy, indicating at the strengths and weaknesses in the

policy implementation and monitoring processes;
the wish to strengthen the coherence between Hivos gender, sectoral and general policies;
new insights and concepts which have been developed by WID practitioners, activists and
academics;
the persistence of structural inequalities between women and men, despite decades of

agenda-setting, policy formulation, programme and project interventions aimed at improv­

ing women's living conditions;
the impact of economic, social, political and cultural changes at global, regional and

national levels on the power relations between women and men;
the efforts and priorities of women organising for change, illustrated by an overall growth
of women's organisations and networks throughout the world;

the ongoing challenge of how to integrate the interests and needs of women in a coherent
and effective manner into development organisations and programmes which focus on

both male and female marginalised people.

1.2

Hivos Women in Development policy in retrospective

The 1988 Hivos WID policy distinguished four priority areas for intervention aiming at:

promoting the social and economic empowerment of women through their participation in

education and decision-making structures, through formation of women's interest groups
and organisations, and through networking among like-minded organisations that promote
women's interest;
increasing women's

economic independence through

ensuring

access to

economic

resources, as well as control over resources and benefits of their activities;
protection and promotion of women's right to self-determination regarding their lives and

bodies through legal action, education, information and provision of services regarding

women's (reproductive) health, sexuality and countering violence against women;
addressing pre-conditions for the advancement of women, in particular through national
legal reform and strategies aimed at reducing women's workload.
The implementation has been effected in a two-fold manner. A pro-active identification strategy

has been pursued and direct support has been provided to women's organisations, in recognition
of the need for and relevance of women's autonomous space to organise and develop strategies.

Simultaneously support has been provided to the so-called 'mixed' organisations in order to

integrate women's interests in the overall programmes and activities. Hivos has engaged in a
process of dialogue and agenda-setting aiming to encourage organisations to review their policies
and programmes in favour of women. In the identification and assessment of new programmes
and counterpart organisations, the contribution to improving women's living conditions has
become a compulsory assessment criterium.

1

Three internal evaluations have been undertaken, in 1991, 1993-1994, and 1995, which
provided Hivos with feedback on the achievements and constraints in implementing the WID
policy.2
With regard to Hivos interventions vis-a-vis counterpart organisations, the evaluations have
indicated:
1.
Hivos successful identification of, and engagement with, women's only organisations, of

2.

3.

which the majority are at the forefront of the struggle for women's equality and part of the
women's movement in their respective countries and regions. Hivos focus on women's
self-determination has recognised the developmental relevance of issues such as violence
against women and women's reproductive rights.
The need for Hivos to assist women's organisations in the areas of organisational and
institutional development, e.g. management, leadership, finance and planning, given the
relative youth of many organisations. Coordination and alliances among women's
organisations and with other development organisations and institutions deserve strength­
ening. The impact of women's organisations on economic empowerment of women has
been limited.
Hivos dialogue with mixed organisations has yielded different responses and diverse
results:

where alliance building internally with staff or externally with WID experts has been
possible, agenda-setting has led to a change in policies and attitudes. These
organisations are currently grappling with the implementation of effective strategies;
- expressing interest and good intentions but not knowing how to go about it in practice
and failing to identify (external) WID expertise;
establishment of separate women's projects and/or structures, often under-resourced,
which do not have a tangible impact on the organisation as a whole and function as a
token commitment to gender issues;
- overt resistance, in particular by male leadership, and especially in relation to internal
organisational issues of gender inequality.
Given the diverse and at times limited achievements in terms of successful integration of
women's issues and interests, more tailor-made intervention vis-d-vis individual
organisations is required. Without any follow-up gender training of staff as a strategy for
organisational change has only a limited impact.
Weaknesses common to many mixed partner organisations relate to an unwillingness to
address internal issues of male-female inequality (e.g. salaries, career opportunities,
training, decision-making, harassment) and a failure to appreciate the importance of data
collection and feasibility studies in terms of women's priorities and needs before the
formulation of projects and strategies.
-

4.

With regard to Hivos itself the evaluations have indicated:
1.
A positive appraisal of Hivos two-track strategy, whereby the impact of support to
women's only organisations has been the most visible achievement.
2.
Given the more complex process of addressing women's issues and interests in mixed
counterpart organisations, the need for Hivos to further develop its internal instruments
such as:
- operationalisation, including a time path;
integration into all stages of the project cycle and at all levels of the organisation, to
avoid dependence on individual staff priorities;

2

development of its data base for the monitoring and assessment of the performance of
mixed organisations regarding the integration of WID;
- refinement of its intervention strategies to enable sustained impact over time rather
than incidental gender interventions facilitated by Hivos (field visits, evaluations,
workshops, training, exchange visits).
A need to increase WID expertise among staff and the number of staff with specific
responsibilities regarding WID.
-

3.

1.3

(

(

The shift to Gender, Women and Development

Hivos direct engagement with women's organisations in the North and South and the debates at
international fora such as the various UN Conferences in the 1990s have inspired Hivos thinking
and intervention strategies. Of particular relevance is the paradigm shift from 'Women in Devel­
opment' (WID) to 'Gender and Development' (GAD).
The concept of "gender" has been introduced to point out that women's subordinate position is
not due to biological factors and that women's position should be considered in its context.
Gender stands for the rules, traditions and social relationships in societies and cultures that
together determine and sanction feminine and masculine behaviour and how power is allocated
and used by women and men. Gender refers to a social construction of femininity and masculin­
ity which varies over time and place. In other words, it refers to the difference between innate
and learned behaviour. As gender is socially and historically determined, it is subject to change.
The value of the gender concept is that it examines social processes and interventions in terms
of their effects on women, on men and on relationships between women and men. It does not
look at women in isolation, and it enables differences between women (and men) due to class,
race, ethnicity, age, abledness, sexual preference to become visible.
The distinction between the WID and the GAD approach has been artificial at times. Both are
political strategies aiming to improve the living conditions of women. And both approaches have
suffered because of lack of adequate resourcing. The weakness of the WID approach has been
that it tended to be of an ad hoc or add-on nature, or at times even worse as marginal tokenism.
It only focused on women separately. The GAD approach has been problematic, too. As gender
refers to both women and men, it might be mistaken for a neutral concept, denying that, within
the gender hierarchy, it is men who are in control. A gender analysis therefore is not a question
of looking at differences, but an analysis of how these differences have led to inequalities in
power between women and men.
Today gender seems to have become the 'politically correct' term. Gender and Development
(GAD) programmes, as opposed to Women in Development (WID), are seen to have a more
structural focus because they concentrate on relationships between women and men. A
consequence has been that some parties concluded that attention for women only is no longer
justified, denying the rationale for women's organisations.
It is important to remember that gender is an analytical concept. Using a gender based approach
or gender analysis means that organisations assess the (potential) impact of their policies, pro­
grammes and interventions on women, men and female-male relationships. This does not mean
that policies and interventions focusing specifically on women and/or men are unwarranted.
Target groups and strategic interventions will continue to depend on local situations.
For Hivos therefore a gender approach implies the promotion of strategies that redress gender
hierarchies and inequalities. These can either focus on women or men separately or on women
and men jointly. Hivos choice of the term Gender, Women and Development in this policy

3

document is motivated by the need for a gender perspective in all Hivos development pro­
grammes, with explicit recognition of women's subordinate position within the gender hierarchy.
Hivos interventions thus have to acknowledge the socio-economic, political and cultural gender
inequalities. The fact that women constitute a disproportional percentage of the world's poor
motivates Hivos to focus on women's absolute and relative poverty. But the promotion of
women's equality goes beyond poverty reduction, as women's poverty is not the cause, but
rather a manifestation, of women's subordination. Hivos acknowledges that women have to be
part and parcel of the solution to problems of underdevelopment, given their strategic and
essential labour contribution and strategic location in development processes. But this should not
imply an instrumentalisation of women's labour and time as an efficient means to achieve wider
development objectives. Taking the prevailing unequal gender division of labour and responsibil­
ities as a point of departure, the differential input and output of women's and men's labour need
to be challenged.

In order for Hivos to establish future policy priorities regarding Gender, Women and Development
this policy document will:
revisit key concepts in Hivos overall development policy in order to integrate and operatio­
nalise a gender vision;
incorporate features of global and national processes posing new challenges for women
struggling for gender equality;
integrate these consequences into Hivos sectoral policy priorities;
specify Hivos choices and interventions vis-d-vis counterpart organisations from a gender
perspective;
establish priorities for follow-up actions by Hivos.

4

2.

Hivos development vision: A Question of Power

2.1

Marginalisation and the focus on unequal power relations

The central aim of Hivos overall development policy is to address and reverse marginalisation
processes in countries in the South to enable the emancipation and full participation of people in
determining their own destinies and the shape of their own society.3
Marginalisation expresses itself in a lack of control over one's own material environment and life
and decreasing access to the decision-making processes that determine people's material and
non-material living conditions. Marginalisation is the result of a political process and involves
social, economic, cultural and political exclusion caused by unequal power relations. Power
inequalities express themselves in unequal access to and control over resources, production,
property, income, information, knowledge, and decision-making within households, communities,
nations and the world as a whole.
Equal opportunities and emancipation will have to be pursued at all these different levels. This
implies that unqualified notions of the 'poor' dr the 'marginalised' have to be abandoned. Not
only differences in class, but also differences in race, ethnic background, age, abledness, sexual
preference and gender account for differences in power and determine the nature and extent of
the marginalisation process.
Emancipation is a process which requires organised and collective expression of shared interests
and needs of the marginalised. This, however, should not be achieved at the expense of
dominating others, but it does require a redistribution of power and control over material and
immaterial resources. Redistribution and emancipation are more easy to strive for in a situation of
growth and abundance. Under conditions of scarcity a conflict of interests will be apparent.
Emancipation implies a commitment to democratic social relations between people in order to
enable participation by all, irrespective the scale and form of social organisation. Gender, more
than any other manifestations of difference, cuts across all social relations between people and
at all organisational levels, including relations of affection and processes of reproduction which
are considered to be 'private' and are experienced between individual human beings. And yet
these 'private' experiences tend to reproduce, signify and legitimise gender inequalities which
manifest themselves as relations of power and domination in the so-called public and social
sphere. A commitment to emancipation and full participation for all has forced organisations such
as Hivos to reconsider boundaries between the public and the private. Hivos adheres to the
individual right to self-determination without violating this right for other human beings.

2.2

Hivos Gender, Women and Development vision: women's empowerment and addressing
gender inequality

Interventions by international and national development agencies, NGOs and women's
organisations aiming to redress women's subordination have been evident during the past two
decades. Progress has been achieved: commitments by national governments and international
institutions by adopting global declarations recognising the principle of equal rights for women
and men.
But the 1995 Human Development Report paints a rather gloomy picture of the extent and
universality of the inequalities between women and men. Whereas progress has been made in
closing the gender gaps in health and education, as of today there is no society where women
enjoy the same opportunities as men, nor do women enjoy the same protection and rights in the
laws of many countries.

5

Some examples4:

of the 1.3 billion people living in poverty, 70% are women;

in all regions of the world women's wages are considerably lower than men's (on average
women are paid 30 to 40 % less than men for the same jobs);

all regions record a higher unemployment rate among women than men;
women's non-monetised, invisible labour contribution is estimated at US $11 trillion a
year;
women constitute a very small proportion of borrowers from formal banking institutions;

two thirds of the 900 million illiterates in the South are women, and girls constitute 60%

of the 130 million children without access to primary school;
maternal mortality and morbidity account for half a million female deaths each year, of
which 99% in the South;

more than 100 million women are "missing" due to discrimination against girls and
women5;
women occupy only 10% of the parliamentary seats and 6% of cabinet positions through­
out the world;
25-50% of all women experience some form of (sexual) violence, predominantly as

domestic violence by partners or acquaintances;

about 80% of the world's 24 million refugees and internally displaced persons are women
and children.

These examples are an indication of a process of ongoing marginalisation of women, both in
absolute and relative terms (compared to men). Hivos GW&D policy has to counteract the

marginalisation of women. Its objective is to promote women's emancipation and their full
participation in the South. For the social, economic, political and cultural empowerment of
women it is necessary to secure equal well-being, equal access and opportunities, self-esteem,

equal rights and control for women and men. Though not new, the concept of empowerment
requires further operationalisation in order to avoid rhetorical commitment and to assist in the

assessment of development interventions.
Emancipation: a political process of empowerment striving for gender equality
Gender equality is not merely a question of women's equal well-being. It is also a political and an

ideological question (involving both women and men). Equality and emancipation cannot be given

or realised in a top-down fashion and cannot be achieved by third parties. Emancipation has to

be striven and struggled for by women themselves expressing and claiming their elementary
rights, developing countervailing power in order to exercise their rights. This involves a process

of empowerment for women both as individuals and through their collective organised efforts. As
the power dynamics between women and men need to be changed, both women and men are

responsible as actors in the process of change towards gender equality.

For Hivos emancipation and empowerment are central to what development should entail. The
challenge is to visualise a concept of power and empowerment which is not based on domina­

tion. Although emancipation is a self-led and self-controlled process, the commitment and
involvement of third parties is definitely important, in particular those with authority (be it a
government, the leadership of organisations and movements, a donor agency, individual men in
the household and family context).
Empowerment has become a fashionable concept and therefore requires operationalisation. It
entails various dimensions: material and immaterial, which are interlinked and relate to issues of
material well-being, access, ideology/value, participation/agency and control.6 Empowerment
6

can take place with regard to each of these dimensions, which, as they are interrelated, impact
upon one another.
1.
Well-being: the basic material needs, like health, food supply and income. Women's
unequal socio-economic situation arises directly from the inequality of access.
2.
Access: the need for equal access to resources, such as education, knowledge, land,
water, employment, labour and capital. In achieving equality of access women face
blockages such as discriminatory laws, traditions, customs and values. Understanding
these factors means a process of conscientisation and involves ideological issues.
3.
Conscientisation refers to a recognition that women's inequality is caused by structural
and institutional discrimination and not by their own personal inadequacies. It involves an
understanding of the difference between sex and gender roles, and that the latter are
socially constructed and can be changed. A gender perspective and gender awareness are
based on the principal value of gender equality as an objective. For women this implies a
recognition of their own self-esteem and dignity.
4.
Equal Participation in decision-making processes. This requires women's mobilisation,
through organising collectively, in order to push for increased representation. Ultimately
this should lead to greater control. The point of departure is that women are considered as
active agents, rather than passive recipients or beneficiaries.
5.
Control: a balance of power between women and men, without either side in a position of
dominance. Equality of control requires participation, to ensure equal access to resources
and equal distribution of benefits. Equal control is necessary for the realisation of equal
well-being.

This framework acknowledges the interrelated nature of women's problems and how issues of
inequality at one level impact at another level. It enables Hivos to assess the scope, limitations
and potential of development interventions. Most interventions do not address issues of equal
control, as these are more difficult to tackle. Interventions aimed at enabling equal access or
participation are relevant as necessary preconditions for equal control, but are no automatic
guarantee of equal control and a balance of power between women and men. Yet there are
limitations to the extent to which the interventions of organisations in the 'public' sphere, like
Hivos, can change unequal gender relations in the private domain, which is the domain where
many dimensions of power inequality between women and men are manifest. Strategic creativity
is required to challenge the more private dimensions of gender inequality.
As a humanist organisation, Hivos policy focus on gender equality is a principle of human justice
and human rights in itself and is guided by the individual right to self-determination, without
violating this right for other human beings.

7

3.

Looking at the World Through Women's Eyes7

The conditions women face in different parts of the world vary widely, as do their needs and
concerns. Generalised statements about women in the South offer little scope for strategic
action. However, given the increasing global interdependency, a number of (new) trends can be

identified which provide an insight in the overall context of women's lives. New features of

global and national political, economic and cultural processes which have become manifest
during the 1990's are posing new challenges for women struggling for gender equality.8

3.1

Economic globalisation

The rapid and wide-ranging process of globalisation of the economy impacts upon women's lives
in many and diverse ways due to: the global restructuring of production and labour processes

and labour markets, the promotion of open economies, liberalisation of trade relations,
privatisation of services, an increased role of international financial institutions and a weakening
of the nation state as a (macro-) economic agent.
Within the formal economy, new employment opportunities for women have emerged whilst

traditional female jobs have been lost. Overall there has been an erosion of the quality of
women's employment, whereby women and especially young women constitute the bulk of
flexible, cheap and unprotected workers employed in export processing zones, home-working,

temporary and casual work, and the informal sector.
Given the overall reduction in employment opportunities in the formal sector, informal sector
employment has steadily risen and tends to be female-dominated. In Africa the informal sector
has become the most important source of employment and income. In Asia and Latin America

sub-contracting has led to informalisation of formal industrial production and resulted in a boom
in home-working. Women make up 90 per cent of home workers.

Increased competition for cheap labour has resulted in a feminisation of migrant labour employ­

ment both nationally (rural-urban) and internationally. The phenomenon of international migration

of women workers is best known in Asia, whilst internal migration of women workers to the
urban areas predominates in Latin America. In all regions women migrant workers are confined to

unskilled, low-paid jobs in traditional female sectors such as domestic service. Employed in
foreign countries, migrant women are often subject to racial discrimination, treated as second-

class citizens, and extremely vulnerable to violence and sexual harassment. Exploitation of

immigrant women increases under conditions of illegality. Illegal migration often leads to prosti­
tution by choice or by force, as in international trafficking of women. Girls and young women

make up the majority of women migrant workers.

Trade liberalisation and the privatisation of services have led to an increased reliance on
women's unpaid labour. The policies of stabilisation and the structural adjustment programmes

imposed by the World Bank and the IMF in many countries in the South have had numerous

negative consequences, in particular for women, given the prevailing gender division of labour
and of caring responsibilities. Real income levels have dropped due to inflation, price increases
affecting in particular essential goods such as food and fuel, charges for public service provision
(education, health), reduced welfare systems, and rising unemployment in the formal sector.
Women's workloads have increased by having to make up for the loss of purchasing power,
income and services. The loss of jobs has been greatest in sectors, like education and health,
with relatively high numbers of women employees. The advances of the past two decades in the
areas of women's/girls' education and health are being undermined. Globally the number of

8

single women and female headed households has increased, whereas they constitute a majority
among poor women. For rural women privatisation policies, e.g. registration and ownership of

land, have undermined their independent access to land and further contributed to female
migration. Women in formerly centrally planned economies are experiencing the effects of a
weakening of the state as a (macro-) economic agent, combined with the abandoning of the

preferential selection system which provided women with access to education and high-level

employment within the state bureaucracy.
The effect has been a combination of feminisation of poverty (increase in the number or

proportion of women among the poor) and women's impoverishment (worsening of women's
living standards).
The process of economic global restructuring throws up new challenges for women's collective

response to improve their working conditions, income and employment opportunities. 'Normal'

organising conditions as known in the formal sector, tend to be absent. Yet women have taken
numerous initiatives to put these new issues on the agenda, into the public arena. Local

conditions have determined whether self-organisation has been possible. Under conditions of
extreme repression or illegality, women's NGOs and women's human rights organisations have

played a facilitating role.
Women and women's organisations have not remained silent about the impact of global macro-

economic processes. Women's organisations, especially in Africa and Latin-America, have been

involved in monitoring the impact of structural adjustment programmes and have demanded
counter-measures by national governments, aid agencies and the international financial institu­
tions. Demands have been formulated at UN Conferences (1995: Copenhagen and Beijing) for

the reduction of bilateral and multilateral debts and for the building of compulsory minimum
thresholds for donors' and aid-receiving governments' expenditure on social development (20:20
compact).
Another strategy has been to demand the measuring and valuing of women's unpaid labour in
the home, on the land, and in the community. Counting women's unpaid work is a strategy first
of all to secure recognition of women's invisible workload, and secondly to ensure that future

changes in social policy do not increase the unwaged labour of women. A revaluing of women's
work has the potential to change the current legal framework of property rights, divorce settle­

ments, collateral for credit.

3.2

A global women's human rights framework and its backlash

Important gains have been made thanks to the persistent organising and lobbying efforts of

women human rights activists and others.
A wide women's human rights framework is emerging: legitimising women's rights as human
rights in the broadest sense. Women are demanding their collective and individual right to decide
about and control their own lives and bodies both in the public and in the private arena.
Women's organisations have rallied around, exposing the violation of a diversity of women's

human rights caused by, for example, structural adjustment programmes, the testing and

dumping of nuclear arms and waste, governments imposing fundamentalist religious laws,
population control policies and practices, sexual and domestic violence, discrimination based on

sexual orientation (homosexuality), age, HIV/Aids infection, refugee status and abledness.
Women's organisations thus
reject the conventional human rights framework based on a
(hierarchical) distinction between three generations of human rights and the separation between
public and private human rights violations.9

9

Some of the achievements are:
the acknowledgement that the human rights of women and girls are an inalienable, integral
and indivisible part of human rights;
the acknowledgement that violence against women is an obstacle to the achievement of
equality, development and peace;
the acknowledgement of women's reproductive rights and a reiteration that governments
should deal with the health impact of unsafe abortion. Another achievement is the call for
the decriminalisation of women who have undergone illegal abortions. Women's right to
access to safe and affordable abortion has not yet been accepted;
the inclusion of women's sexual rights: their right to have control over and decide freely
and responsibly on matters relating to their sexuality, including sexual and reproductive
health, free of coercion, discrimination and violence. The fact that sexual orientation
(homosexuality) is being used as a ground for discrimination has not yet been explicitly
recognised;
the agreement that trafficking in women and children is a violation of human rights.

Progress at one conference has subsequently been threatened at the next one. The language and
definition of women's human rights within the Platform for Action and the Beijing Declaration
was the most contentious area prior to and during the Fourth UN World Conference on Women.
A number of countries expressed explicit reservations about the final clauses. The biggest divide
was not between the North and South, but between religiously-inspired, conservative govern­
ments and NGOs and those in favour of women's equality, who assert that religion and culture
should never be an excuse for violating women's human rights. Women's organisations in the
South have played a crucial role in the reaffirmation of women's human rights in the Platform for
Action.
Religious fundamentalism and cultural nationalism are not residues from the past. They are
today's response to fears of loss of authority and security in the context of globalisation pro­
cesses and are expressed through a reappraisal of traditional, stereotyped notions of femininity
and masculinity (at odds with the daily reality) in the name of national culture and ethnic or relig­
ious values. Cultural and ethnic nationalism, communalism and religious fundamentalism have led
to armed conflict in which women have participated on both sides.
To counteract the violation of women's rights in the name of preserving culture and national
stability, alternative notions of femininity and masculinity need to be developed. Alternative
gender-sensitive images in the media, the arts and communication at large need to confront the
ongoing stereotyped portrayal of gender roles, which negate women's (and men's) actual and
diverse lives and realities in an ever-changing world.
3.3

The women's movement as an actor in civil society

Whilst economic, political and cultural processes of globalisation have been key features of
recent history, social movements have likewise challenged the detrimental effects of globalisa­
tion at local, national and global levels. The women's movement has emerged as the most
significant social movement during the past decades, especially because it has been based on
principles of self-organisation and coalition-building around collective interests. Women's
organisations continue to emerge and expand, reaching out to all corners of the world, address­
ing an ever-growing range of social issues, and expressing a diversity of women's needs and
interests.

10

The representation of women's organisations and the debates held at the 1995 UN Conference
in Beijing and the 1995 NGO Forum in Huairou were indicative not only of the numerical strength
but also of the qualitative strength of the women's movement as a global force. The capacity to
lobby, advocate and engage with stakeholders such as the UN, the World Bank, the IMF and
national governments reflects a growing self-confidence, political maturity, leadership and pro­
fessionalism to be found in women's organisations today.
Women's organisations have played a critical role in nurturing women's leadership, at times at
the expense of their own organisation, as leaders have moved on into the arena of official
politics as politicians or into bureaucracies of mainstream institutions. In many parts of the world
women's organisations are lobbying for increased representation and participation of women
within the formal political institutions and processes. Local conditions continue to determine the
scope for women to play a role as change agents from within the (political) establishment.
The need for and relevance of institutionalisation of strategies promoting gender equality have
become more legitimate. Coalition-building and cooperation among women and women's
organisations across institutions remains critical.

(

(

11

4.

The integration of Hivos Gender, Women and Development policy into sectoral policy
priorities

The objectives of Hivos' overall development vision are laid down in the institutional document
'Full Participation: A Question of Power'. Five central themes are given priority in the support to

organisations in the South. These are: economic self-reliance, culture and development, human
rights and Aids, environment and sustainable development, and gender, women and develop­

ment. The latter has a distinct character. As gender inequality cuts across all sectors and target
groups, the GW&D policy will have to be integrated into all policies and development interven­

tions. The section below provides guidelines for this integration process and highlights the
intersection of the GW&D policy and the other sectoral policies.

4.1

Economic self-reliance

Hivos has for long prioritised the need to strengthen people's economic base as an indispensable

factor in an integral process of emancipation. Women constitute 70% of the 1.3 billion people
living in poverty. The focus of Hivos policy for economic self-reliance has to address the main

causes of women's poverty: the unequal division of labour and caring responsibilities between
women and men, discrimination against women and girls blocking access to capital, property,
education, technology, employment, equal wages and thus obstructing women's access to
independent income.

The organisation of women wage-earners
Hivos supports the right of women workers to organise around their interests as a priority for

improving their working conditions, income and employment opportunities. Trade-unions are
absent in many of the sectors where women work, due to repression or outright prohibition
(Export Processing Zones), the isolation of women workers (domestic workers, agricultural

labourers), the seasonal, and the casual or illegal nature of employment (women migrant

workers). Moreover, where trade-union organisation has been possible, it tends to be male-led. In
general trade-unions have favoured the interests of their (majority) male membership, neglecting
the needs of women workers, which are often beyond conventional work-place issues. In the

absence of unions NGOs can play a role through research, exposing violations of (women's)

human rights, lobbying, networking, providing information on international frameworks and
standards, and promoting unionisation.
Hivos will pay attention to specific target groups among women, e.g. women employed in EPZ's
or foreign controlled and export oriented industrial plants such as the maquilas in Latin America,

the plantation sector and commercial farms, domestic service, women sex workers and women
migrant workers.
Women informal sector workers: organisation, credit, knowledge of markets and business skills
The informal sector is a growing source of income for the marginalised in the South, especially

for women. Self-organisation among women informal sector workers is a high priority for Hivos.
Support is provided to women's organisations that mobilise around issues such as police
harassment, exploitation by middle-men, municipal licenses for street trading, and sexual and
criminal violence experienced by women working and often living on the streets, and demand
recognition of union-organisation among women informal sector workers.
Access to credit for marginalised people engaged in informal sector activities and micro­
enterprises is an important feature of Hivos economic policy. Women face additional difficulties
12

in accessing credit, which limits their returns, due to lack of collateral, illiteracy, the small scope
of their activities, and the physical and social distance from credit facilities. Strategies to
overcome these obstacles will have to be incorporated into the lending policies of financial

institutions and other credit providers. Hivos cooperation with local financial institutions and
NGOs with credit schemes needs to build in targets and conditionalities to ensure that credit is

accessible to women borrowers, and has to include gender segregated monitoring of the actual

credit provision.
There are some risks in using credit as a main strategy for women's economic empowerment.

Women have become a target of small-scale credit programmes in many parts of the world,
because they are considered to be more reliable and tractable. Investing in credit for women has
become profitable, given their higher rate of repayment. The danger is that women will be used

by male relatives to access credit and income. This is even more detrimental when women have
to provide all the labour for the loan activity, without independent control over returns. It is thus

essential to combine actual access to credit with strategies that promote women's participation
in decision making regarding the loan allocation, and women's control over the income earned.

Evaluations of credit schemes for women indicate that women have greater control over benefits
when they are directly involved in the buying of inputs, selling of outputs, and accounting for the
loan.10

Hivos support for credit programmes will include specific components to provide women with
information and training programmes regarding the market and financial management.

Women and agriculture
A large part of women's invisible and unpaid work is their labour input in food production and
processing, both for subsistence and market consumption. Women lack independent access to

economic resources such as land and water which limits their control over labour returns, their
access to marketing channels, as well as their access to economic resources such as credit.
State-subsidised support services in the areas of extension, distribution of inputs and marketing
of products have been gender-biased. As these services are currently being privatised women

now face gender discrimination within commercial marketing and lending institutions.
Small farmers' organisations, such as farmers' unions and cooperatives, have shown a similar
institutional gender bias, marginalising women's interests.

Women's right to own and inherit property, in particular land, is critical. Hivos will support new
emerging initiatives in this field. In Africa a lobby for women's land rights is gaining strength.
Women's access to, and control over, natural resources is directly related to sustainable land-use
and food security (see 4.4).

In its cooperation with small farmers' organisations, Hivos interventions will focus on strengthen­

ing the gender sensitivity of the organisations and their programmes. Strategies for alternative
off-farm sources of income for rural women need to be developed.

Women and macro-economic policies
There is a lack of gender expertise in the field of macro-economics. The development of alterna­
tive visions from a Southern feminist perspective on the environment and development, popula­

tion and reproductive rights, on social exclusion, and on global and macro-economic policies has
been of value at the various UN Conferences. The challenge now is to move beyond visions
towards strategies. For this, country-specific or region-specific studies and analyses undertaken
by local institutions are required to enable strategic actions.
Research and lobbying initiatives aimed at developing gender-sensitive economic policy and

gender monitoring of global macro-economic policies will be supported.

13

4.2

Culture and Development

The arts and culture are important forms or expression of people's identity and visions. Cultural
expressions and the arts are media for communication in a process of emancipation and
development: visualising people's aspirations, dreams, values and self-respect. Hivos support to
the arts and culture sector is based on appreciation of its potential role to act as a conscience of
society, revealing contradictions and false certainties, breaking down barriers and widening
people's horizons.
The arts are a much-neglected sector in a world dominated by values of economic growth and
material development. Under these conditions local access to a diversity of cultural expressions
is threatened. Strengthening of local and indigenous art and culture is important in order to give
people access to cultural expressions of immediate relevance to their own context. But as culture
is dynamic and affected by changes in the wider social, political and economic context of a
community, a nation or the globe, the arts also contribute to an exposure to the outside world.
In Hivos view the arts can be of special importance to women wishing to challenge fundamen­
talism and conservatism. This implies a rejection of cultural notions that seek to legitimise gender
inequalities.

From a gender perspective Hivos will look into the gender biases within the arts which relate to
issues of participation and representation.
Given the universality of gender hierarchies, women's voices and perspectives are underrepresented within the arts and culture. Women's active involvement in publishing, creative
writing, the performing and (audio) visual arts needs to be promoted, as well as their access to
new communication technologies.
This can be done by supporting the products of women artists, provision of training and
exchange programmes for women artists, support to women's cultural organisations and
initiatives such as women's writing, publishing and distribution.
A second priority is the need to challenge the stereotyped gender images that are being
portrayed, not only in the arts and culture, but in communications and the media at large. Hivos
gender preoccupation with culture and development therefore extends beyond the arts in the
narrow sense. Organisations involved in developing alternative gender-sensitive images in
communications and the media will be supported. This includes initiatives such as organisation of
women journalists, women's radio programmes, women's initiatives in the print-media and elec­
tronic media. Especially for young people, female and male, challenging the gender division of
labour and caring responsibilities is important in order to break down gender stereotypes and to
provide alternative role models.
4.3

Human Rights and Aids

Respect for, and protection of, individual and collective human rights is essential in Hivos
development vision, as development also means social and political participation. Hivos supports
human rights organisations in recognition of the relevance and contribution of these
organisations to:
the exposure of the violations of human rights;
processes of emancipation and self-determination of those whose rights are being denied;
strengthening of civil society through the promotion of democratic, participatory, and
accountable relationships between the state and its citizens.

14

(

The promotion of women's human rights has been a key focus of women's organisations, as
part of the global women's movement, throughout the world.
For Hivos gender equality is a principle of human justice and human rights. The individual right to
self-determination as laid down in the broad women's human rights framework is crucial.
Support is provided to organisations that play a role as a women's human rights watch, that
provide services to women whose rights are violated, and that play a role in putting a gender
perspective into civic education and the process of democratisation of society at large. This
includes activities in the fields of:
reproductive rights, including the right of access to safe and affordable abortion;
sexual rights, including the right for lesbian women to organise on the basis of their sexual
identity;
combatting violence against women, in all spheres;
women's property and inheritance rights;
women's right to maintenance for their children;
gender-sensitive legal aid, training and civic education;
exposure of women's human rights violations in trafficking;
violation of women's rights in armed conflict and the rights of women refugees;
research and data collection for lobbying, advocacy and monitoring purposes for gender
equality (e.g. regarding national constitutions, government policies, international conven­
tions).

Aids
Since 1990 Hivos has called for specific attention to issues related to the HIV/Aids pandemic as
a threat to development and as a human rights issue. Hivos support to organisations concerned
with HIV/Aids focuses on three broad areas:
prevention strategies;
respect of the human rights of so-called "high risk" groups, the HIV-infected and people
living with Aids, and the promotion of their self-organisation;
support to emancipatory processes at the cross-roads of Aids and sexuality.

(

In the South women bear the brunt of the epidemic and are most vulnerable to infection, as HIV
is transmitted predominantly through sexual intercourse (70-80% of infections). In addition, HIV
infection due to blood transfusion is more common in women than men.
From a women's human rights perspective women's empowerment and the breaking down of
taboos concerning sexuality are of strategic importance for prevention strategies. Sexuality is a
most complex terrain, where relations of affection, desire and pleasure are intertwined with
relations of power and domination, experienced in a private and intimate context. The HIV/Aids
pandemic forces both women and men to become more open about issues of sexuality. This
could have a positive impact, in particular for the younger generations. Women's vulnerability to
HIV/Aids and Sexually Transmitted Infections relates directly to wider issues of social, economic
and cultural gender inequality.
A gender perspective in relation to the HIV/Aids pandemic for Hivos implies support for activities
which increase women's bargaining power aiming at women's empowerment. Women's right to
inherit property has emerged as a critical issue for women whose husbands have died of Aids.
Specific strategies aimed at changing male behaviour and responsibility in decisions concerning
sexuality need further attention. Women sex workers are a special target group in order to assist
in negotiating safe sex practices.

15

4.4

Environment and Sustainable Development

In Hivos view problems of environmental degradation in the South are directly linked to unequal

North-South relationships and the dominant growth oriented development model.
Environmental problems hit hardest at the local level, those most affected being those who
depend most on natural resources for their survival: above all, poor rural women in the South.

Often the victims of environmental degradation are blamed as well. This is especially so in the

perspective which sees environmental degradation as being caused by population growth. The
'population question', however cannot be separated from the need for sustainable development,
including sustainable production, consumption and redistribution of resources at local and global
levels. For Hivos women's reproductive rights and health, and women's right to self-determina­
tion over their own bodies, are the bottom-lines for any action and policy in relation to the

'population question'.
The 1989 Hivos policy on the environment needs a gender perspective. Women are only

mentioned as victims of the combined process of environmental degradation: exhaustion and

pollution. Women as active agents do not appear in the two-track environmental strategy which
aims at:
the promotion of sustainable development practices, in particular in the fields of land,

water and energy use in agriculture, livestock farming, fishing and forestry, in order to
address environmental degradation through exhaustion;

the taking of counter-measures against environmental pollution at local, national and global
levels.

Women, Environment and Sustainable Development
From a gender perspective Hivos environmental policy needs to take the debates on Women,

Environment and Sustainable Development (WED) into account.11
Hivos does not adhere to the viewpoint which attributes women a privileged position in environ­

mental management based on the assumption that women have a closer and morally superior tie

with nature. The fact that women are both victims of the environmental crisis as well as
important actors in resolving it, is due to the gendered division of labour in production, repro­
duction and distribution. For their survival poor rural women have had to rely upon the environ­

ment much more than men. Therefore women have both contributed to and prevented further
environmental degradation. Rural women's socio-economic position is related to their access and

control over natural resources. Environmental interventions will have to differentiate between

type and quality of resources and need to pay attention to the diversity among women. This will
make it possible to anticipate both conflicting as well as complementary interests between the
goals of gender equality and environmental sustainability. Environmental interventions need to
take into account the fact that poor women are unable to take risks that undermine their security

in terms of food, energy, water and fodder.

Some lessons can be drawn from the implementation of WED programmes in the South:
New labour and energy-saving technologies (solar stoves, water taps, afforestation
projects) tend to neglect the additional social or economic importance of traditional

technologies for women. A one-dimensional concern with the environment risks overlook­
ing women's multiple use of environmental resources, either for domestic purposes or for
the generation of additional income.

16

Strategies that address women's material needs can conflict with environmental concerns.
Improved access to drinking water through new wells or taps reduces women's workload
but can result in increased water consumption, as more people and cattle will be attracted,
over-exploiting resources and threatening the environmental sustainability of vulnerable

semi-arid areas.
A conflict of interests can also emerge in land-registration programmes which are propa­
gated as part of sustainable land-use policies. Land-registration is promoted, given the
direct relationship between access and control over natural resources, including land, and
people's interest in investing (e.g labour) in improving the quality of these resources. In
many societies women do not own land in the formal sense, but traditional usufructuary
land rights secured a degree of women's control and decision-taking in land-use. These
usufructuary rights are now being threatened due to commercialisation, land scarcity, and
environmental degradation, with the result that women's usufructuary land rights are
limited to smaller, more marginal and less productive pieces of land. Formal registration of
land rights tends to neglect women's usufructuary rights.

(

Hivos environmental policy encourages all its counterpart organisations to take environmental
concerns into account. Given the ambiguous relationships between environmental sustainability
and gender equality, Hivos now encourages organisations to recognise the relevance of:
a socio-economic and gender analysis of the use of natural resources, at micro-, meso- and
macro-level (including marriage and inheritance laws) prior to environmental interventions,
and women's participation in this analysis;
locally specific interventions, given the diversity of gender inequalities in production
systems, land-use, ownership and control over natural resources and other resources such
as credit, new technologies, etc.;
strengthening the participation of women in identification and decision-making concerning

environmental interventions;
revaluing women's knowledge of indigenous natural resources ( e.g. bio-diversity) and
techniques as well as increasing the possibilities for women to benefit from the application
of their knowledge and expertise;
promoting women's access to, and knowledge, of new sustainable technologies and
resources;
women wishing to establish separate women's structures, e.g. departments, unions or
cooperatives in order to further their interests.
Hivos support to environmental campaigns by action groups encourages the integration of a
gender dimension into their campaigns against environmental degradation and pollution and into
North-South debates on unequal production and consumption patterns. Support is provided to
campaigns which highlight the danger to women's (reproductive) health resulting from the use of
pesticides and the testing and dumping of nuclear arms and waste.

17

5.

Operationalisation

5.1

Strategic choices

The previous chapters have outlined the rationale for, and the vision behind Hivos Gender,
Women and Development policy. Changing global conditions have been taken into account in
drawing attention to a critical intersection of the GW&D policy with Hivos other sectoral policies,
This chapter describes how Hivos intends to implement its GW&D policy. Implementation iis a

combination of general and context-specific strategic choices in selecting and relatingI to
counterpart organisations in the South. The GW&D policy includes general criteria for the
selection of counterpart organisations and strategies for Hivos interaction with the counterpart
organisations, which are valid for all regional programmes. The specific implementation is

determined by the geographical context and is therefore beyond the scope of the overall GW&D

policy. The regional policy documents and annual plans outline which gender issues are most
urgent in relation to Hivos sectoral policies.

At the general level Hivos will continue with its two-track operationalisation within all regional
programmes: support to women's organisations and support for the process of integrating
gender into mixed counterpart organisations and their programmes. Given the progress made in

terms of agenda-setting of gender issues,

Hivos instruments of support, including non-financial

support, will prioritise processes of organisational development, organisational change, institu­
tional development and gender capacity building.

Hivos proposes a three-fold strategy:

support for capacity-building and organisational development of women's organisations
aiming at strengthening their organisational performance and institutional impact;
support for the process of integrating gender into mixed organisations and their programme
activities, with the emphasis on tailor-made interventions aiming at organisational change
and development;

support for increasing both female and male expertise and capacity in the field of gender
and organisational development, organisational change and institutional development.

These three strategies are further elaborated below. It is important to bear in mind that Hivos
does not engage with all its counterpart organisations with the same intensity. Neither is Hivos

of equal importance to all counterpart organisations. Hivos is committed to a qualitative improve­

ment and deepening of the gender capacity and performance of its counterpart organisations.
This, however, requires an active relationship. The priorities for Hivos interventions as identified
below are based on lessons drawn from such an active engagement in the past. These lessons

are of relevance to new Hivos counterpart organisations as well, but Hivos first preoccupation
will be with counterpart organisations which receive institutional support for more than 2 years.

Hivos programme staff and the individual counterpart organisation will jointly decide which
organisational gender priorities will be addressed during the next four years and how. This is
expected to be reflected in the organisation's annual plans and programme proposals.

5.2

Women's only counterpart organisations

Hivos interest in supporting women's organisations is based on:
the importance of self-organisation in a process of emancipation and empowerment, given
the individualised and personalised nature of women's marginalisation;

18

the

contribution

of

women's

organisations

to

confidence-building,

assertiveness,

organisational skills and leadership among women;
the need for the services, training and information provided by intermediary women's

organisations, given their specific gender-sensitive skills in a particular field;
the contribution of women's organisations in strengthening civil society, influencing public

opinion, playing an advocacy and lobbying role promoting gender-equal policy, representa­
tion and participation in relevant institutions, organisations and sectors.

In order to qualify for Hivos support women's organisations need to be committed to the broad
women's human rights framework, to principles of emancipation, democratic organisational
functioning, accountable leadership, and have delivery capacity in their area of work. Hivos

supports

women's interest organisations

at both grassroots/community

level and

at the

intermediary level.
Women's organisations tend to experience particular organisational issues which relate to their
relative youth, their search for an alternative organisational culture and structure, and their
relationship with the state and interaction with other organisations. Many women's organisations
are now moving from the pioneer-stage towards a next stage in their organisational develop­

ment. Hivos support to its institutional (more than 2 years Hivos support) women's counterpart
(

organisations therefore prioritises specific interventions aimed at strengthening organisational
performance, institutional impact and professionalisation of women's organisations. Resources

and/or (external) and professional expertise will be made available to assist with:
strategic planning, defining priorities and delineating programme boundaries based on the

expertise and strength of the organisation;
assessment of the economic feasibility and viability of those women's organisations which
aim at economic empowerment of women, and especially to assist these organisations in
separating and delineating economic activities from other activities aimed at empowerment
in the social and political sense;
design of strategies and programme activities targeting men, aiming to change male

behaviour and clarifying their role in a process of change towards gender equality;
building of leadership and management capacity, including the training of a second
generation of leadership;
issues of organisational structure and processes of decision making, in relation to the

division and delegation of authority, tasks and responsibilities, and issues of downward
accountability towards the constituency and beneficiaries;
processes of internal organisational learning and reflection, ensuring a capacity to monitor

and evaluate the impact of the work being done;
the building of alliances and strategic cooperation with other organisations based on

(

shared interests and needs.
With regard to new Hivos counterpart organisations, programme staff will encourage networking
with already more established women's organisations within their region.

5.3

Integration of gender into mixed counterpart organisations

The susceptibility of organisations to the need to integrate gender issues varies and obviously
does not depend on Hivos interventions alone. A visible presence of women's organisations in

the institutional environment of an organisation and the presence of a kernel of gender advocates
within the organisation are critical enabling factors. The impact of Hivos interventions is
influenced by the relative importance of Hivos financial support to the organisation and the
extent to which other donor agencies take gender issues seriously as well.

19

Hivos concludes that interventions by donor agencies can and should contribute to the gender

agenda-setting process within mixed organisations. The challenge is now to contribute to a
process of operationalisation and implementation of gender-equal policies within organisations
and their programmes in a more systematic and consistent manner. This requires both legitimacy
and scope for an active Hivos engagement, which depends on factors such as:

the relative importance of Hivos in an organisation's engagement with donor agencies and
the duration of the relationship with Hivos;

the type of organisation

and

its

needs:

e.g.

grassroots/community-based,

interest,

advocacy, intermediary development, service rendering, network;

the organisation's scale and size of operations as well as its history and age;

the organisation's capacity in terms of human and financial resources (quantity and quality)
in relation to needs and type of organisation;
the external environment (constraints and opportunities for gender equality, e.g. legal

context, availability of gender expertise, strength of women's organisations, etc.).
Hivos experiences in the past point to a number of areas which deserve further attention by its
mixed counterpart organisations. Hivos programme staff will incorporate these into the routine of

their engagement with those organisations which receive institutional support for more than two

years. Within each continent a small number qualify for a more intensified approach, as Hivos
interventions to date highlight both the critical need or demand and the scope for a pro-active

strategy for support over a number of years.
Priorities for all mixed counterpart organisations
For the improvement of the gender capacity and performance of mixed counterpart organisations

implementation of the following strategies is recommended:
strategies to increase the number of female staff working in the organisation, and in
particular to increase their number in senior management positions. A minimum guideline is

the objective of 30% female staff (excluding administrative and logistical support staff).

Commitment to gender equality also needs translation into organisational policies that
address career opportunities, training, decision-making, harassment and the scope for
women to act as an organised group within the organisation.
commitment to employ staff with specific gender expertise and with a mandate to utilise
this expertise for the benefit of the organisation. A combination of strategies seems to
produce the best results: a more centrally located unit/person with authority and access to
resources, with decentralised gender responsibilities throughout the organisation. Promo­
tion of gender awareness among all staff of the organisation is encouraged.
allocation of financial resources for programme activities aimed at gender equality.

providing data on the gender composition and the respective positions of the staff and the

board of the organisation and on the gender impact, qualitative and quantitative, of
programme activities.
generating knowledge about socio-economic, political and cultural gender relations in the
programme area and methodologies which enhance the participation of women benefi­
ciaries in identification, design and decision-making concerning planned activities.
commitment by smaller organisations or organisations incapable of generating gender

expertise and gender-sensitive goals, and ways and means to achieve these, to cooperate
with women's organisations and other gender experts for this purpose.
Hivos is aware that implementing these recommendations is a process. Therefore in consultation
with its current counterpart organisations which receive institutional support for more than two
years, a realistic time framework and priorities for implementation will be decided upon. If
20

(

necessary, Hivos will make resources and/or external and professional expertise available for this
process. After four years Hivos programme staff and the counterpart organisations should be

able to assess whether progress has been achieved. If they fail to reach an agreement on the
direction of change and the possible future need for change, Hivos will reconsider future
cooperation. In engaging with new counterpart organisations Hivos will from the start emphasise

that the above issues need attention.
Intensified process of gender integration
With regard to a small number of counterpart organisations, Hivos is of the opinion that they

qualify for a more intensified process of gender integration. Critical reflection on past perform­

ance, Hivos interventions in the past and a growing internal demand for change, have widened
the scope for gender integration within a number of strategic Hivos counterpart organisations. In
order to move ahead, a three to four year ongoing process of tailor-made support aimed at

institutionalising gender equality is envisaged.
Hivos Continental Bureaus will select three to five counterpart organisations for this process.
Organisations will be selected on the basis of criteria such as commitment to the process,
enabling conditions such as internal and external allies which can play a role in the change

process, the strategic importance of organisations in respect of critical gender issues within the
context of a country or region, the exemplary role and potential to draw lessons from the

organisation's engagement with the process.
Hivos and the participating organisations will draw up a contractual agreement specifying the
objectives and an activity plan for the process of organisational change, including the nature of
support required. This will be stipulated in a contract. Hivos will make resources available for the

process which will be facilitated by local gender experts. These will assist organisations in

analysing the state of affairs, identifying objectives for the change process and development of
an operational plan. This plan will combine gender objectives regarding programme activities and
internal organisational aspects. Monitoring and evaluation of the process are built in to facilitate

organisational learning for both Hivos and the participating organisations. These experiences will
enable Hivos to distinguish better between generally applicable interventions and organisation
specific interventions.

5.4

Gender capacity in organisational development, organisational change and institutional

development

Professional expertise in the field of gender and organisational development and organisational
change (OD & OC) is required for Hivos intervention strategies vis-a-vis both women's and mixed

counterpart organisations
One strategy for Hivos support is to contribute towards increasing the number of women in the

South with expertise in the areas of OD & OC. Some work as individual consultants or trainers,
while others are based within women's organisations or within organisational development &
training institutions. Building up national or regional networks among these women is important
in order to enable exchanges of experience and strategies and to strengthen their capacity and

expertise.
A second strategy for Hivos is to support processes which increase gender-awareness and
gender-specific OD & OC capacity within and among its OD counterpart organisations and pool
of organisational consultants which Hivos draws upon. Gender-sensitivity of male OD & OC
practitioners is of strategic importance in particular for reaching out to a male audience.

21

A third strategy is to support institutional development through promotion of cross-fertilisation
between strategically located women's organisations and other NGOs, through facilitation of
joint meetings, workshops, exchange programmes, consultations, information sharing and
networking. Support to regional and international women's organisations and networks will be
continued. Issues of downward accountability and the relationship between members and the
coordinating structure will be a focus of Hivos attention.

22

6.

Hivos institutional strategies

6.1

Policy, programme and project work

Two core processes are central in the organisation of Hivos work. The first relates to the area of

policy, the second to the implementation of policy, the project cycle, as expressed in Hivos

support to individual organisations and their programmes. Both policy and project work are
cyclical processes, cross-fertilising one another on the basis of Hivos organisational matrix­

model. Both cycles involve specific instruments and procedures, as laid down in the Hivos inter­
nal procedures file. The consequences of the GW&D policy for both cycles are described below.
GW&D policy cycle

There are a number of stages in the development of policy. For GW&D policy this implies the
following:

1.

Agenda-setting
Agenda-setting of gender issues remains a continuous process, as new insights and issues

will continue to emerge, from within and outside Hivos.

(

2.

Policy development
The 1996 GW&D policy is the second, thus renewed, Hivos policy aiming at women's
emancipation and empowerment. Its implementation will point to the need for revision or

updating in five years' time.

3.

Operationalisation
Detailed operationalisation of GW&D policy is and will be reflected in the continental

policies. General operationalisation of GW&D policy includes:
- Continuation of the current level of financial support to women's organisations (16% in
-

1995).
The process of integrating gender into mixed counterpart organisations will have to be

reflected in an increase in organisations whose programmes pay attention to issues of
gender inequality. By the year 2000 at least 50% of the organisations with which Hivos

has an institutional relationship for 2 years or more will express a commitment to
implement the recommendations listed on page 20 in their programme proposals and
-

annual reports.
Incorporation of GW&D policy in other Hivos sectoral policies, which should result in an
increasing number of counterpart organisations which can be classified under both
GW&D and another sectoral policy.

(

4.

Implementation of policy
This is the actual project cycle and is discussed in more detail below.

5.

Monitoring and evaluation of policy
This involves monitoring and evaluation beyond the level of individual organisations and
programmes.
For monitoring, the instrument of quality audit is used through the half-yearly mid-term

review based on the targets laid down in the continental annual plans. This is the responsi­

bility of the continental GW&D policy officers using the data base of organisational
assessments of counterpart organisations, evaluations, travel reports, consultancy and

advisory missions.
For evaluation, the instrument of the summative evaluation is used. There are two types of
summative evaluation:

23

(i)

meta-anaiysis, based on a GW&D evaluation of the total programme within a particu­

(ii)

inter-project analysis, based on a comparison between a number of evaluations

lar region (country or continent) based on external evaluations already undertaken,
carried out in the field from the GW&D policy perspective.

The strategy of intensified gender integration will be subject to a meta-analysis. In each
continent one inter-project analysis will be carried out.

GW&D policy and the project cycle

Implementation of GW&D policy implies a review of the instruments and decision-making

processes which are part and parcel of the project cycle.
1.
Identification

When deemed necessary during the stage of programme formulation, Hivos will stimulate
organisations to make use of locally available gender expertise and knowledge, and will

provide them with support for programme development embracing the needs and interests
of women.
2.

Assessment

The

Organisational Assessment (OA)

is the internal instrument assessing

both the

organisation and its programme activities. It prescribes a gender assessment of the

organisation with regard to its objectives, strategies/policy, activities, internal organisation

and the institutional context. In future Hivos programme staff will have to provide refer­

ences

to the five dimensions

of empowerment:

well-being,

access,

ideology/value,

participation/agency and control. Assessments of mixed organisations need to refer to the
recommendations as listed on page 20.
Weaknesses identified in the assessment will have to be dealt with in subsequent gender

intervention strategies. When the assessment indicates a total lack of attention and inter­
est as regards gender issues, and no feasibility or scope for implementing gender inter­

vention strategies, Hivos will in principle turn down the funding application. Otherwise

3.
i-

explicit motivation and justification is required.
Approval

The contractual relationship between Hivos and the counterpart organisation will include

gender conditions relating to the gender intervention strategy when this has been identified

to be of critical and strategic relevance in the Organisational Assessment.
4.

Monitoring
The monitoring process involves a mix of different instruments such as discussions and

correspondence between Hivos and the organisation, field visits, the organisation's annual
reports and plans.
The minimum standards for reporting, as laid down in the General Conditions, will have to

be adapted in order to include more detailed gender disaggregated reporting requirements.
A challenge for both Hivos and the counterpart organisation is to analyse performance

against the five dimensions of empowerment, bearing in mind that empowerment is a
process of social change which is difficult to capture in straightforward quantitative and

qualitative indicators. Improvement of the gender data base within Hivos internal adminis­
trative system is in progress.
5.

Evaluations
The standard Terms of Reference for external evaluations prescribe a gender assessment.
Adaptation is required to ensure that the gender analysis will be undertaken at all levels of

the evaluation: organisational purpose as regards a particular target group/constituency;

24

6.

internal organisation; and the institutional context. The team of evaluators should always
include gender expertise.
Renewal
Both Hivos and the counterpart organisation will have to incorporate into the next phase of
programme planning and design, the insights generated during implementation and
monitoring and evaluation of the organisation's performance. If no agreement can be
achieved in terms of a shared commitment to strategies aiming at gender equality, Hivos
will reconsider continuation of its engagement with the counterpart organisation.

(

4

25

Annex I Hivos Institutional State of the Art

1.

Institutional framework

Hivos does not foresee a change in the institutional set-up regarding its internal GW&D capacity,
given the recent restructuring of the organisation. This resulted in a quantitative increase in
GW&D capacity and more emphasis on sectoral policies within the Programmes and Projects
Department.

The GW&D working group has the organisational mandate to contribute to the GW&D policy
development process, monitoring and evaluation of its implementation, and to promote gender
expertise among staff. The GW&D working group is institutionally based within the Programmes
and Projects Department. It consists of three GW&D policy officers, one in each Continental
Bureau, of which one has additional institution wide gender responsibilities. The GW&D Senior
Policy Officer as part of the team of five Senior Policy Officers has to guide the process of inte­
grating gender into the cycles of overall and sectoral areas of policy development. Cooperation
between the GW&D working group and other departments in the organisation embraces:
Bureau Monitoring, Evaluation and Control: regarding the integration of gender into the
various instruments of the policy and project cycle, into the administrative system of data
collection, and.into the summative evaluation processes.
Bureau Internal Affairs: regarding affirmative action strategies (recruitment, training,
promotion to management positions) and other internal organisational policies (e.g.
regarding sexual harassment).
Bureau Communication & Marketing: regarding the integration of gender into the informa­
tion, campaigning and fund-raising strategies.
Hivos Board: a GW&D Board Advisory Committee has been established for consultations
regarding policy development. GW&D experts are included among members of the Board.

2.

Hivos external relationships

The GW&D Senior Policy Officer represents Hivos in relevant external fora. Of particular
relevance are the Gender Group of the four Co-Financing Agencies (GOM Vrouwenoverleg),
Women's Coalition Development Institutions (Vrouwenberaad Ontwikkelingssamenwerking), and
the Eurostep Gender Working Group.
In the field of gender Hivos maintains contacts with a wider range of organisations in the
Netherlands and abroad, beyond the field of development. This includes women's studies
institutions, women's organisations and women's units within mixed organisations concerned
with Dutch emancipation policies, and a new emerging post-Beijing initiative.
As a follow-up to its choice to engage with the women's movement in the South, Hivos also
wishes to strengthen its cooperation with women's organisations and the wider women's
movement in the North, especially in the Netherlands. In the near future Hivos aims to
strengthen its institutional ties and relationships in this respect, in recognition of the need for
alliances and cross-fertilisation of emancipation and gender integration strategies.

26

3.

Hivos gendermap; board and staff (RO's included), as per 31-12-1996 (*)

Board
1991

1988

Executive
Board

1996

male

female

total

male

female

total

male

female

total

9

4

13

16

5

21

10

7

17

69%

31%

100%

76%

24%

100%

59%

41%

100
%

Staff

management/
senior staff

executive
staff

supporting
staff

total

1996

1991

1988

7

2

9

12

3

15

11

6

17

78%

22%

100%

80%

20%

100%

65%

35%

100
%

9

6

15

16

16

32

16

21

37

60%

40%

100%

50%

50%

100%

43%

57%

100
%

2

11

13

1

14

15

1

20

21

15%

85%

100%

7%

93%

100%

5%

95%

100
%

18

19

37

29

33

62

28

47

75

49%

51%

100%

47%

53%

100%

37%

63%

100
%

(
(*) Number of boardmembers and employees according to definition Annual Reports

27

Annex II Statistical overview (1991 -1995)

Table 1:
Disbursements based on Hivos policy priorities (*)
Africa

Latin
America

Economic self-reliance
33
Culture
5
Gender, Women & Development
11
Environment & Sustainable Development 8
Human Rights & Aids
18
Total
75

30
3
20
16
9
78

9
4
24
28
22
87

8
6
11
66

28
4
16
15
15
78

Other sectors

25

22

13

34

22

100%

100%

100%

Asia
Europe/
International

41

Total

100% 100%
('

(*) figures taken from the annual report 1995

Table 2:
Hivos policy priority: gender, women & development per continent (*)

number of
partners

%

expenditure
(x Dfl. 1000)

%

Africa
Asia
Latin America
Europe/lnternational

24
34
25
4

13%
20%
15%
14%

2.421
2.700
4.611
318

11%
23%
20%
8%

Total

87

16%

10.050

16%

(*) figures taken from the annual report 1995

Table 3:
Target group projects - based on gender (total programme) (*)

Women ( > 80%)
Women & Men
Men ( > 80%)

Africa

Latin
America

Asia

Total
1995

Total
1994

18
77
5

17
79
4

24
74
2

19
77
4

17
80
3

100% 100%

100%

100%

100%
(•) figures taken from the annual report 1995

28

Table 4:
Counterpart organisations and expenditure Gender, women and development per continent (1991 - 1995)

1991

1992

1993

1995 (*)

number
of
organi­
sations

%

Dfl (x
1.000)

%

number
of
organi­
sations

%

Dfl (x
1.000)

%

number
of
organi­
sations

%

Dfl (x
1.000)

%

num­
ber of
organi­
sations

%

Dfl (x
1.000)

%

Africa

18

16

1.055

6

22

16

1.483

9

28

17

2.652

15

24

13

2.421

11

Asia &
CIS

18

15

683

10

21

14

1.304

11

34

23

1.901

19

34

20

2.700

20

Latin
America
&
Caribbean

42

23

5.033

21

31

17

4.552

16

33

19

4.304

18

25

15

4.611

24

4

24

148

10

3

17

105

8

4

14

318

8

78

16

7.487

13

98

19

8.962

17

87

16

10.050

16

Intern.
Total

78

21

6.771

13

(•) figures for 1995 are not comparable with those for 1991 -1993 due to a change in classification

29

Table 5:

Women's counterpart organisations as per 1-1-1996, in number of organisations and % per
sector and continent (*)

Second sector &
Continent

Africa

Asia
& CIS

Latin America
& Caribbean

Inter­
national

Total

Multi-sectoral
women's
organisations

7
25%

7
21%

12
48%

3
50%

29
31%

Economic
self-reliance

5
18%

5
15%

2
8%

12
13 %

Culture &
Development

2
7%

4
12%

1
4%

7
8%

(

Human rights &
Aids

14
50%

Environment &
Sustainable
Development
Total

28
100%

16
47%

9
36%

2
6%

1
4%

34
100%

25
100%

3
50%

42
45%

3
3%

6
100%

93
100%

(*) the number of women's organisations is slightly higher than in the annual report 1995, because in this table mixed
organisations whose programmes focus entirely on gender equality have been included

(

30

Table 6:
Overview of mixed counterpart organisations with and without gender focus, shown per sector,

as per 1-1-1996, in number of organisations and %

Africa

Asia &

Latin

Internatio­

CIS

America

nal

Total

& Carib­
bean

Economic

+ gender focus

self-reliance
no gender focus

+ gender focus

11

31

10%

5
4%

8%

7%

34
22%

8
6%

25

4

71

17%

17%

16%

15

3

2%

Culture &
Development

Human rights

no gender focus

+ gender focus

& Aids

no gender focus

Environment &
Sustainable

Development

No Sector

+ gender focus

no gender focus

+ gender focus

no gender focus

Total

+ gender focus

no gender focus

Total of
organisations!*)

4

7

3%

2%

17

1
4%

4%

48
11%

14%

5%

8
6%

21
13%

12
9%

8%

4
17%

15
10%

32
24%

15
10%

1
4%

6
4%

10
7%

4
3%

10
6%

25
19%

17
12%

15
10%

8
6%

17

3

12%

13%

29
19%

39
27%

6

101

20%

26%

22%

60
38%

35

47

7

26%

33%

30%

149
33%

100

96
67%

16
70%

308
67%

143
100%

23
100%

457
100%

8

96
62%

156
100%

(*) excluding women's counterpart organisations

31

27

74%
135

100%

11

63

20
4%
4
17%

56

12%

43
9%

Notes
1.

In: Stella and Frank Chipasula (eds.) (1995): The Heinemann Book of African Women's Poetry,
pp. 16, Heinemann, Oxford.

2.

'Vrouwenbeleid onder de loep', Hivos 1991.

Evaluation Hivos and WID: 'Report Women in Development - Central America', July 1993;

'Report Women in Development - Asia', September 1994, 'Report Women in Develop­
ment - Africa , January 1995, 'The 1994 Evaluation of Hivos' Incorporation of Gender

Issues in Theory and Practice', July 1994.
'Hivos' Gender Policy in Indonesia: An Assessment', October 1995.

3.

'Full Participation: A Question of Power, An institutional approach'. Policy Document Hivos in
the Third World, The Hague, 1988.

4.

UNDP (1995): 'Human Development Report 1995', Oxford University Press, New York/Oxford.
No improvement has been recorded in the subsequent 1996 Human Development Report.

5.

Based on a projection of the male-female composition of populations in which women are

treated more equally, there is a shortfall of women in particular regions (North Africa, South,

South-East and West Asia). This is caused by discriminatory practices such as female infantici­
de and abortions, violence against women, and higher death rates amongst girls than boys at

a young age due to discrimination in the provision of health care and nutrition.

6.

This a revision of the empowerment framework as developed by Sarah Hlupekile Longwe, 'A

framework for understanding women's empowerment', paper for UNICEF Workshop,
Gaberone, 1992.

7.

Slogan of the NGO Forum on Women, Huairou, 30 August - 8 September 1995

8.

The impact of the process of globalisation on gender relations has began to feature within the
work undertaken by a number of international women's organisations. Much analysis is still

required. Economic globalisation features in:

DAWN (1995): Markers on the Way: The DAWN Debates on Alternative Development,
DAWN's Platform for the Fourth World Conference on Women, Beijing, September 1995'.

Natacha David (1996): Worlds Apart Women and the Global Economy, ICFTU, Brussels.

9.

A human rights framework has been laid down in a number of international declarations such

as the Universal Declaration of Human Rights (1948), various UN Covenants and Conventions
such as those on Civil and Political Rights (1966), on Economic, Social and Cultural Rights
(1966), on the Elimination of All Forms of Discrimination Against Women (CEDAW, 1979), and

on the Rights of the Child (1989). The distinction between three generations of human rights

has become a much used framework:
First-generation human rights, which relate to individual civil and political rights;

Second-generation human rights, which include economic, social and cultural rights;

Third-generation human rights, which include collective rights such as the right to develop­
ment, a liveable environment, peace and security.

10.

Brooke A. Ackerly (1995): 'Testing the Tools of Development: Credit Programmes, Loan
Involvement, and Women's Empowerment' in: /DS Bulletin, Getting Institutions Right for

Women in Development, Vol. 26, No. 3, pp. 56-68.

32

11.

For a discussion on the emergence of Women, Environment and Development (WED) as a
theme within the development debate, see the following two publications, which provide a

useful insight into the diversity of positions.
Rosi Braidotti, Ewa Charkiewicz, Sabine Hausler, Saskia Wieringa (eds.) (1994): 'Women,

the Environment and Sustainable Development: Emergence of the Theme and Different

Views', in: Women, the Environment and Sustainable Development, pp. 77-106, Zed
Books, London.

Heleen van den Hombergh (1993): Gender, Environment and Development, a guide to the

literature, International Books/INDRA, Utrecht/Amsterdam.

(

33

Not viewed