2022.pdf

Media

extracted text
EARLY CHILDHOOD MORTALITY

AND
PERINATAL PERIOD MANAGEMENT IN URBAN POOR
i.

3

Report Submitted to
THE MINISTRY OF HEALTH AND FAMILY WELFARE
(Linder USAID Child Survival Programmes), Government of India

by
The Department of Pediatrics, Maulana Azad Medical College,
New Delhi-110 002
September 1991

v

--- » MORTALITY AND PERINATAL PERIOD
BARLY CHILDHOOD
MANAGEMENT IN URBAN POOR

Investiqators
r.K.

Puri

G.R. Sethi
B. Raaji
H.P.S. Sachdev
p. choudhury

Report
Submitted to the Ministry of Health and Family Welfare
(Under USAID Child Survival Programmes), Government of India
bv the Department of Pediatrics, Maulana Azad Medical College,
New Delhi - 110 002.
September, 1991

ACN0WLEDGEMENT8

The project Investigators wish to thank the Research Officers,

field and statistical staff for the tremendous effort put in by
each one of them to ensure the successful completion of this

study.

Our special thanks are due to Dr. Onkar Mittal, Health Care
Co-ordinator, Slum Wing, Delhi Development Authority for the help

rendered in identification and selection of JJ Clusters, recruitment and training of field staff and help us out of difficult
situations arising in the field during the course of the survey.
We also gratefully acknowledge the services of Dr. Varinder

Singh, Senior Resident, Department of Pediatrics, Maulana Azad

Medical College, New Delhi for the enthusiastic help rendered in
the development of educational aids used during the project.

training of Research Officers and field staff and helping monitor
their field activities.

CONTENTS

Introduction

1

Methods

4

Observations

Part A: Mortality Survey

29

Part B: KAP Survey

59

Summary

87

Recommendations

90

References

92

INTRODUCTION

(1)

Nearly one third of the world's population lives in

substandard housing conditions in urban areas - a situation

likely to worsen with the current demographic trend of the

population doubling every thirty years.

It is estimated

that of the 830 million people in India, almost a quarter
lives in urban areas; of these over 60% live in sub human

conditions deprived of basic social and health services. In
Delhi alone, about 35,00,000 people live in slums which may

be broadly classified as unauthorized colonies, JJ resettlement colonies, walled cities or Katras and Jhuggi Jhompri (JJ) clusters (Bastees).

Of these, the last, namely JJ

clusters are the worst off as far as habitat and availability of basic services are concerned.

Estimates by the slum

wing, Delhi Development Authority (DDA) indicate that about
1.3 million people reside in these JJ clusters.
In majority of cases, the urban poor (JJ clusters) of

cosmopolitan cities are probably worse off than the rural

poor with respect to availability of basic amenities of
life (water supply and sanitation) and health care deliv-

ery.

In rural areas, there is a definite envisaged Primary

Health Centre with its network.

Conversely, in a prime

metropolitan city like Delhi, multiple authorities adminis­

ter health care services (Central Government, Municipality,
State Administration, Dispensaries, Polyclinics, Private

I
(2)

Sectors and so forth). Unfortunately, however, these services are not effectively networked for delivery of health
care to the underprivileged urban poor. Recent surveys of
the National Institute of Health and Family Welfare (1,2)

and voluntary sectors (Society for Action in Community

Health - SACK) have revealed an abysmally low utilization
of the existing health services by the residents of Delhi's

JJ Clusters.
In this setting. therefore, it would be logical to
assume a higher rate of childhood mortality. However, there
is no precise quantification of this aspect as the urban
poor are not covered by the Sample Registration System
(SRS) - the only reliable source of vital statistics on a

national scale.

Rough estimates from occasional surveys on

a limited scale indicate that early childhood mortality is

considerably higher than the projected national figures.

It is, therefore, vital to get reliable estimates of under
five mortality and its secular change amongst the urban
poor.

This would prove invaluable in prioritizing the

child survival programmes in the national context.
It is now well recognized that the peri-natal and neo­

natal periods are important components of early childhood
mortality. Baseline information on management of peri-natal

This
and neonatal period in the urban poor is scanty.

data

METHODS
I
t

i

I

I

(5)
social, sanitary
Bastees were selected as their economic.
and health conditions are considered to be worst.

obtained from the slum
According to the latest list
there are. at
wing of Delhi Development Authority (DDA),
inhabiting about 2.5 lakh
present 929 JJ Bastees in Delhi
into five zones (Table
families. These have been divided
The surveyed clusters were selected from these 929
1

areas.
TABLE H

jj

Bastee Zones as per DDA List

Zone

b.

No. of clusters

Central

93

East

123

North

227

South

282

West

204

Sample Size

conducted strictly
The childhood mortality survey was
according to the guidelines laid down by a recent UNICEF

publication from the Regional
North Africa (3) .

Office of the Middle East and

(6)

Prior to starting the survey, it was anticipated that

the proportion of children dead by the Preceding Birth
Technique would range from 0.10 to 0.15.

From the earlier

data (3) it was estimated that roughly 4 to 6 married women

in the age range 15-49 years would need to be interviewed
for one index woman of Preceding Birth Technique.

It was

decided that a precision of 1.5% would be adequate.
cluster design effect was assumed to be 2.

The

Based on these

assumptions, the maximum sample size required for by the

Preceding Birth Technique was calculated to be 18,000 evermarried women in the age range 15-49 years.

The Basic

Sampling Unit (BSU) was, therefore, defined as a household

with at least one ever married woman in the age group of
15-49 years.

C.

Cluster Selection
In the initial proposal submitted it was envisaged

that 3600 households would be surveyed in each zone by

selection of 6 clusters of 600 households in each zone.
However, later it was realized (3), that it would be more

accurate and representative to split the sample size into
as many clusters as possible.

The UNICEF booklet had suggested that the BSU number
for a cluster should not generally exceed 100 and be so

(7)
much that it could be easily covered in one day by the

recruited field workers (3).

However r these recommenda-

tions pertain to a countrywide survey.

The present study

had objectives of health education and evaluation of knowl­

edge. attitude and practices on peri-natal practices in

addition to mortality survey. Further, due to the distance

involved and need to establish confidence and contact with
the local population, it was felt that two days would be a
more realistic period required for each cluster.

It was

decided to evaluate at least 120 Basic Sampling Units in

each cluster.

The cluster number for 18000 households was,

therefore, calculated to be 150.
The 929 Jhuggi Jhompri clusters had number of house-

holds ranging from 25 to 7000.

Based on the sample size

required in each cluster, areas with less than 120 house­

holds were excluded from the randomization procedure.
The randomization procedure adopted was probability

proportional to size as described in the UNICEF booklet
(3) .

Many Bastees had a population which was larger than

the sampling interval so they were selected more than once
(2 to 5 times).

Independent samples (more than one) were

selected from such Bastees.

In this way. a list of 150

selected Bastees was prepared.

The geographic division of

these selected Bastees was 14 in Central Zone, 25 in East

FIG.1. ORGANIZATIONAL SETUP
Principal
Investigator

Overall supervision

Investigator

I nveat igator

Investigator

Investigator

Investigator

Central Zone
(30 Clusters)

East Zone
(30 Clusters)

North Zone
(30 Clusters)

South Zone
(30 Clusters)

West Zone
(30 Clusters)

Research
Officer
Zonal supervision

Research
Officer
Zonal Supervision

Research
Officer
Zonal Supervision

Research
Officer
Zonal Supervision

Research
Officer
Zonal Supervision

Field
Workers (8-10)
Data Collection

Field
Workers (8-10)
Data Collection

Field
Workers (8-10)
Data Collection

Field
Workers (8-10)
Data Collection

Field
Workers (8-10)
Data Collection

(8)

Zone, 35 in North Zone, 51 in South Zone and 25 in West
Zone.

Since the geographic North and South Zones have

larger populations as compared to other areas, the selected

clusters were more than 30.

However, for the purpose of

conducting the survey five administrative teams (Fig..1)
Investigator/
were conceived under the supervision of an
For the sake of uniformity and
Senior Research Officer.
equal division of labor, each administrative team was

entrusted with the task of surveying 30 clusters.

A few

clusters from the geographic South Zone had to be, there-

fore, transferred to administrative Central (n=16) and West
Similarly 5 clusters were transferred from
(n=5) Zones.
the geographic North Zone to administrative East Zone.

The

clusters to be transferred were decided on the basis of

geographic contiguity of the selected areas.

During the

course of the survey, one of the selected areas got burnt
out in South Zone and one cluster could not be located.
The next eligible, geographically contiguous cluster as per

the DDA list was substituted for these areas.

The final

150 clusters surveyed by the five administrative zones are
detailed in Annexure 1. These areas are also shown in a map
of Delhi in Annexure 2.

In a few instances, it was observed that the number of
households as per the DDA list was exaggerated (the lists
are prepared as per the ration cards and often the resi-

(9)
dents make fake entries).

In these clusters, the minimum

requirement of 120 BSU's could not be fulfilled (Table 2).
The minor deficiency was made up from the adjoining unnamed
clusters (if available) or the adjoining area as per the

selected cluster list.
TABLE 2:

Details of Clusters where Quota of 120 BSU*s
could not be Completed

Shortfall in 120 BSU

Zone

<5

5-10

>10

Central

5

2

0

East

2

0

0

North

1

0

0

South

0

0

0

West

4

3

1

12

5

Total

r

1

The zonal figures refer to the number of clusters

which had a shortfall in 120 BSU. The total BSU shortfall
was 80.

The zone wise distribution was central-27, East-2,

North-1 and West-50.

1

(10)
d.

gousehold Selection
Although, strictly this is not totally under the

purview of preparatory phase, it would be pertinent to
elaborate this aspect now.

The individual household selec-

tion amongst the clusters was the most crucial and tough
A comprehensive list of all the households in Jhuggi

task.

Jhompri Bastees is not available.

It was also not techni-

cally feasible to carry out a guick survey of all the
households for this purpose.

The situation was further

complicated by different geographical shapes of the various

clusters. Despite all these constraints, it was decided to

have as uniform as possible, individual household selection
procedure. The basic principle applied was that the select­

ed households should be spread as far wide as possible and
the selection should be random rather than in a haphazard

manner.

Every household would, therefore, have an equal

chance of being included in the survey.
The methodology employed for this aspect is detailed
below.

For bigger areas which were selected more than

once, each selection was treated as an individual cluster.
If blocks or natural subdivisions were available for such

subdivide for the individual
areas. these were used to

clusters. Rarely in other cases, geographic subdivisions
had to be created.

(11)

The household selection procedure for each cluster was
determined by the number of households according to the DDA

list (Table 2).

TABLE 3: Household Selection Procedure According to Number
of Households

Cluster size

Household selection

1.

Upto 165

Every successive

2.

166

176

Four out of five (Drop every fifth)

3.

177

198

Three out of four (Drop every fourth)

4.

199

264

Two out of three (Drop every third)

5.

265

396

One out of two (Drop every second)

6.

397

528

Select every third

7.

529

660

Select every fourth

8.

Above 660

Select every fifth

The selected clusters in each Zone were subdivided

into three according to the household size: (i) upto 396;
(ii) 397 to 1000;and (iii) Above 1000. A birds eye view was

obtained to get an idea of the centre and periphery of the

selected cluster.

The selected clusters were surveyed in

an alternating manner from the periphery to the centre and

vice versa in each Zone in each subdivision according to

(12)
household size referred to above.

Owing to the varying

geographic constructions of the clusters, the direction
(periphery to centre or vice versa) could not be strictly

adhered to in all cases; the general direction, however,
was maintained.
An attempt was made to select at least one natural in

road into the cluster for each worker.

If the natural

inroads exceeded the number of workers, a randomization of
the inroads was performed.

If the natural inroads were

less than the number of workers, two workers were employed
for some of these inroads with one worker sticking to the

left side and the other to the right.

If as per the randomization procedure, the worker
encountered a non existent house (without roof) or a perma-

nently locked house or an ineligible house (only men, no
ever-married women) then the next eligible household was
selected as per procedure.

In some small clusters, the eligible households finished by the randomization procedure without the worker

having completed his quota of BSU’s for the area.

In such

a case, the worker started from the middle of the first two

selected households and adopted the same procedure as

before till the required number of BSU’s was completed.

(13)

For the next worker with this eventuality, he started
from the middle of the last two houses and proceeded to the
beginning.

The envisaged randomization procedure was fairly

successful; only 3.8% of the households had to be selected
out of the randomization procedure.

A household was designated as temporarily locked if it
was eligible and the inhabitants were in town but were not

available at the time of survey.

The usual time when the

index woman could be contacted was determined from the

neighbours.

workers.

A record of these houses was kept by the field

At least two revisits were done before categoriz-

ing them as missed respondents and allowing substitutions.
A total of 324 houses were initially categorized as ’tempo­

rarily locked' and 144 finally ended up as missed respond­
ent'. Thus with perseverance 56% of such women could be

recontacted.

2.

Preparation of Questionnaires and Health Education

Material.
a. Questionnaires
All the questionnaires were finalized in the local

language (Hindi) after adequate pretesting in the field by

(14)

investigators, research officers and workers.

The details

of the various questionnaires are elaborated below.

(i) Basti Infomation Questionnaire.

In this questionnaire, basic demographic information
for the surveyed JJ Bastees was collected. (Annexture 2).
This included information on water supply, garbage dispos-

al, sanitation, presence of public toilets (Sulabh Shauchalaya), health, services available. etc. The questionnaire

was filled up by actual observation (by Research Officers)
and dialogue with the population.
(ii)

Mortality Questionnaire

In strict accordance with guidelines of the UNICEF
booklet

(3) , the mortality questionnaire was carefully

translated into Hindi.

It was confirmed that the questions

were exactly translated back into the same English by an

independent person.

Due care was exercised not to change

the order of the questions while translating.
A few additions were incorporated in the UNICEF questionnaire after pretesting.

These included an introductory

paragraph stating the purpose of the survey and instructions for the field workers to ensure proper interview

technique and filling up of the form.

The interview date

and identification number of the cluster household head

(15)

(the usual identification in Bastees) were printed in the

beginning.

A column for children's' name was also incorpo-

rated to facilitate recall.

To obviate problems during data analysis and compilation. each mortality questionnaire was given a unique

identification number (Annexure 41.

For households with

more than one eligible woman, extra pages 3 and 4 of the
mortality questionnaire were printed (Annexure 51.

There

was no printed serial number for these extra pages; the
serial number of the original mortality questionnaire was
copied by the workers on these pages. New instructions and

innovative ways of assessing age of mother were introduced in
the mortality questionnaire which are detailed later.
(iii) Perinatal Questionnaire
The KAP questionnaires for both mothers and Dais

(Traditional Birth Attendants) were prepared by the Project
Investigators in consultation with Dr. Kamla Ganesh, Pro­

fessor of Obstetrics and Gynecology at Maulana Azad Medical
Collegef New Delhi.
Mothers who had delivered a child in the 3 months

preceding the survey date were included for the KAP study
in the perinatal questionnaire.

It was estimated that with

a birth rate of approximately 30 per 1000 population, the

(16)
eligible women per cluster would be about 4. A lot of

changes were suggested in the questionnaire after pretesting in the field; leading questions were minimized and

emphasis was laid on current practices.
The • Perinatal Proforma1 was designed to assess the
KAP with respect to care during pregnancy, labor and post

partum period (Annexure 6). The questionnaire was usually
filled up by supervisors (Research Officers); and only

occasionally by the best Field Workers of the zone.

(iv)Dai Questionnaire
A Dai (Traditional Birth Attendant) was defined as a
person, invariably a woman) who was regularly utilized by

the community to assist in child birth.

This broad defini-

tion was also used for an earlier study (4) by the Urban

Basic Services (UBS) which had revealed that only 15% Dais’

were trained and most deliveries in the community were

conducted by untrained birth attendants.
The •Dai Proforma’ was filled by Supervisor (occasion-

ally best Field Worker) by interviewing the two most popular Dai's (if available) of that area.

The information

collected in this proforma (Annexure 7J_ was similar in

nature to the •Perinatal Proforma'.

(17)
b. Health Education Material

Since a large number of families were to be inter­
viewed for the mortality status and KAP Survey, the oppor-

tunity was also utilized to disseminate health education.
This health education was not restricted to the families

being interviewed or the mortality survey period.

A lot of

health education material was developed for this purpose.

(i)

Hand Outs
These health education handouts were disbursed to the

population at the end of 'Health Education Session',

The

handouts in a written and pictorial manner included basic
information on diarrheal diseases, nutrition. care during

pregnancy and immunization (Annexure 8_1_- Invariably many
other mothers in the vicinity of the selected households

were collected for health education and they also received
these handouts.

(ii) Handbook for Basti Sevika
This hand book was developed by SACK (Society for

Action in Community Health) for training of Health Workers

in the JJ Bastees.

This booklet was distributed to all the

field workers (Annexure 5L)_.

It provides information on the

concept of primary health care and important health prob­
lems.

it was designed to develop better insight in the

(18)

Field Workers so that they can effectively impart health

education to the community.

(iii)

Panch-Tantra

This small handbook giving important health informa­
tion in number of five was also utilized for health educa-

tion (Annexure 10).

This was modified and translated in

Hindi to adapt to this Project.
(iv)Video and Audio Cassettes.
Video and audio cassettes providing important health

education were also collected and utilized.
3.Recruitment and Training of Staff
a. Recruitment

The organizational set up of the project with respect

to field work has already been alluded to in Fig.1.

The

Senior Research Officer was responsible for overall field

supervision in one zone.

The five Research Officers (one

for each zone) acted as field supervisors.

supervisors were male.

Only 2 of the 5

All of them had previous experience

of working in the community (as Medical Social Worker/
Doctor) either in rural or urban areas and had the leader­

ship quality, capacity for team work and establishing good

rapport with the respondents.

(19)
With the help of Research Officers,

6 to 8 Field

Workers (Interviewers) were selected for each zone.

These

workers were mostly women (only three were males) either
from JJ Bastees or JJ resettlement colonies.

Majority of

them had previous experience of working is community level
health workers in various Health Programmes of Delhi Admin­
istration or Non Governmental Organizations (NGO's). All of

them were literate and good in establishing rapport with
respondents.
b. Training
Before embarking on the actual survey. the staff was

adequately trained for this purpose.
The Supervisors and Field Workers were motivated by
explaining the importance of data to be collected and what
would be done with it.

The essentiality of accurate

measures of mortality and the field worker's contribution

to this aspect was highlighted.

An outline of the survey plan was discussed.

The

administrative arrangements during the work such as working
hours and salary, survey schedule, transportation arrange-

ment and day to day procedures were finalized.

(20)

Extra care was taken for training of the mortality
questionnaire. Each and every sentence of the questionnaire

was explicitly discussed.

Tips on interviewing technique

and gaining the confidence of respondents were provided.
Proper emphasis was laid on avoidance of

•leading ques-

tions’, guessing age of lady, importance of completing each
assigned interview, following standardized procedures and
asking questions as exactly worded in the questionnaire.

The possible problem areas were discussed and tentative

solutions to anticipated problems were suggested.
Initially dummy interviews were conducted in the
hospital

Subsequently the interviews were monitored in the

field.

The actual field work commenced only after we were

satisfied with

respect to the quality of data entry.

The Supervisors were explained the household selection

procedure in minute details and the importance of random
selection. Written details were available with each Supervisor for ready reference.

They were instructed to monitor

at least 5% of interviews, how to check the forms for
errors and correct the problems on the spot.

A similar procedure was adopted for the Bas tee Dai and
Perinatal questionnaires and for Health Education.

(21)

4. setting yp of an Office

nd Computer Facilities

A project office was opened up in the Department of

Pediatrics, Maulana Azad Medical College with statistical
and clerical staff.
(i)

The main responsibilities were :

To keep all the questionnaires and health education materi­

al for subsequent distribution.
(ii) To keep a record of the completed proformas received from
each Zone.

(iii) To assist in manual checking of questionnaires.
(iv) To send the questionnaires for data punching and receive

them back.
(v)

To send the error forms back to field for necessary corrections after reinterview and to receive them back.

(vi)To help in final data analysis.

The office acquired a computer

(PC 386 AT) for data

Any problems from the field were referred to

analysis.

this office which acted as a sort of control room.
B.Field Work
The field work for the mortality questionnaire was

completed in a little over two months period beginning in

early June.

To minimize the chances of measurement error,

adequate stress was given to practical training and moni-

toring.

Specific duties were assigned to the Investigator,

Ch'oo

community health CEU

326. V Main, I Block

Koranic ngr.|3

Bangalore-560034

India

7

(22)

Supervisor and Field Worker as per guidelines laid down in
the UNICEF booklet (3). These guidelines were available in

written form with the Supervisors.
The selected JJ Bastee had been physically verified by

at least one of the workers a day prior to the survey date.
Before starting the survey, the Supervisor was expected to

talk to the leader (Pradhan) of the Bastee and other opinion leaders (including females) to inform them about the

purpose of the project and the methodology adapted.

Subse-

quently, from a common collection point as per the randomi-

zation scheme.
workers.

the inroads were distributed to the various

The Field Workers were instructed to contact the

Supervisor immediately in case of any problems.

The Super­

visor was asked to personally monitor at least 5% of the
interviews (one per Field Worker per cluster).

The Inves-

tigators conducted random checks on the Supervisors and
Field Workers in each cluster. The Field Workers and Super­

visors met at an appointed time (usually lunch) to review

the situation and discuss problems.

The information re-

garding households to be contacted for perinatal proformae
was passed on to the Supervisors.

At the end of the day.

the Supervisor manually checked the mortality proformae for
completion, coding errors and inconsistencies.

The incom-

plete and error forms were immediately sent back to the

(23)

respective households for correction/reverification.

On

completion of a cluster, the forms were returned to the

office on the next day between 9 to 10 a.m.

For sake of

uniformity and to prevent losses, the details of completed
forms were submitted on a proforma (Annexure 111,.
Weekly meetings were held at the central Office be­
tween the Investigators and Supervisors. The main purpose

of these was to effectively monitor the data quality and
progress in each zone.
(i)

This was achieved by:

Discussing the problems faced and offering solutions

through innovations.
(ii) Monitoring the adherence to time schedule.

(iii) Pointing out the errors (especially for a worker who would
be repeating the same mistake).

(iv)Confirming that the guidelines and instructions were being
strictly adhered to.

(v)

Returning back error forms (detected after computer and

manual check) which had escaped attention in the field for

innovations for Mortality Survey

The mortality survey was conducted in strict accord­
ance with the guidelines laid down by the UNICEF booklet
(3). Due emphasis was laid on explaining the purpose of the
survey, securing co-operation and politeness.

Efforts were

(24)

made to involve other members of the household, particularly the husband, to improve accuracy of the data, However,

a few unusual problems specific to JJ Bastees were encoun­
tered to which no obvious solutions were available.

These

^/ex*e overcome by innovations which are detailed below.

The JJ Bastees are usually constructed in a haphazard

manner and the individual households do not have specific
number or •identification mark. There is, therefore, a
possibility of two workers visiting the same house,

Also

retracing the temporarily locked houses the second or third

time is cumbersome. To obviate these problems, after visit—
ing the selected house, the workers were instructed to put

a cross (X) with a colored indelible ink pen on the wall
near the door. The worker put her individual identification

number below the cross only if the interview was completed,
the cross was encircled. With this innovation, retracing of
the required households become easy and confusion was
minimized.

It also provided a ready check for the Invest!-

gators and Supervisors, especially for the accuracy of the
randomization procedure.
It was important to collect the details of accuracy of

randomization and the extent of ineligible household

(permanently locked, only men, etc.). A uniform methodology

was evolved by asking the workers to fill in the records of

i

(25)

the households visited on a printed proforma (Annexure

2) .

The exact age recording for women constituted a major
obstacle during the pretesting. The population surveyed was

mostly illiterate and further no importance was given to
knowledge about exact age. Birth records were non-existent.
Over 95% of women did not know the date, month or year of

birth. Age was often loosely stated as a range. To minimize
this problem,the woman's age was tallied with that obtained

by splitting her age into various components (age at mar-

riage, first live birth after how much time of marriage and
age of first live born child).

If there was wide discrep­

ancy between the stated and calculated ages. the woman was
patiently asked to review her age. In addition to timing

with major events, this innovation proved particularly

useful in determining the women's age with better accuracy.
The recall problem with ages of children (particularly

the younger ones) was less.

However, the month of birth

was often stated as per local calender.

Even when the

month and year of birth were stated, the workers sometimes
committed mistakes in calculating the age of child in
completed years and months.

To minimize this problem, a

table with a sliding scale was created (Annexure

3). This

readily provided the calculated age from the local calendar

J

month and year of birth.

With this innovation the

(26)

calculation errors diminished considerably and a uniformity
was maintained in conversion of local months to English
months.

The interviewers were stressed the importance of

reconfirming the ages of children if the interval between
them was below 10 to 11 months or above 3 to 4 years.

The workers sometimes committed mistakes in coding sex
of child since the names of children were also written, the

errors could be easily rectified.
Data Compilation and Analysis

a. Compilation

The data was received from each Zone after a cluster
was finished.

A strict entry and exist record of all forms

from each Zone was maintained.

Excessive care was taken to

scrutinize and compile the mortality questionnaire forms.
During the planning stage, it was envisaged that data

entry would be done through a software supplied by the
UNICEF, Regional Office for the Middle East and North
Africa (5). However, this software proved unsuitable for

the purpose of the survey.

Data entry with this was very

time consuming and somewhat confusing. The entry format was
not planned in sequence with the mortality questionnaire.

(27)

It became obvious that in the time available, we would not
be able to finish the job with this software. A decision

was, therefore, taken to design an indigenous software.

In

this interim preparatory period of 10-12 days. only a

manual scrutiny of forms was done to detect errors.
The software designed for this survey had about 20

range and consistency checks.

Important amongst these were

age of women, number of completed forms, discrepancy in
sex, calculated age of child, calculated age at death,
interval between two children and discrepancy in live and

dead status. A separate consistency file kept a record of

all consistency errors while simultaneously showing them on
the screen.

Automatic skips were also designed in the

software depending on the number of children or required
entries. The spreadsheet for this software programme is

provided in Annexure 14. This indigenously designed software proved considerably faster.

However, our computer

time was at a premium and the data had to be punched commercially.

A strict entry and exit record for commercial

data entry was maintained.

The punched data was scrutinized manually to detect

any punching or field errors.

The punching errors were

simultaneously edited in our computer while the field error

(28)

requiring revisit were given to supervisors

forms

weekly meeting.

at

the

This process was continued actively during

the mortality survey period to provide an adequate feed
back and get the forms corrected in time.

It also provided

vital information an errors being committed by a particular
field worker or puncher.

Fig.2

depicts the process of data compilation in a

flow chart.

b.

Analysis

The childhood mortality was estimated by indirect
methods as per guidelines provided by the United Nations

(3/6).

The Q FIVE software supplied by United Nations (6)

was utilized for the Brass Technique.

FIG.2. DATA COMPILATION
FIELD
WORKER
Data Sheets

c
o
R
R
E
C
T
E
D

F
O
R
M
S

RESEARCH
OFFICER
Ver I flcat Ion

PROJECT
OFFICE
Statisticians (8)
Compi lat Ion

DATA
PUNCHING
Developed Software
Consistency Checks

PROJECT
OFFICE
Manual Re-check

E
R
R
O
R

F
O
R
M
S

OBSERVATIONS

(29)
PART A: MORTALITY SURVEY

A. CIVIC AMENITIES
The current state of important civic amenities in the
surveyed population as per the information collected in

For this purpose,

Bastee proforma is provided in Table 4.

information for a cluster was recorded only on one proforma

even though that area may have been selected more than once
for the mortality survey.

Information for a few clusters

could not be recorded in the Central Zone.
TABLE 4: Details of Important Civic Amenities in the
Population

Surveyed

Zone

Number
surveyed

No. of
house­
holds

Approx.
Populat­
ion

No. of
water
points*

No. with
Dhaloas

No.with
public
utilities

Central

14

18,885

94,425

109

10 (71)

8

(57)

East

27

25,541

1,27,705

304

15 (56)

12

(45)

North

24

33,327

1,66,635

509

8

South

25

32,136

1,60,680

238

15 (60)

11

West

25

19,927

99,635

110

10 (40)

11 (44)

Total

115

1,29,816

6,49,080

1,270

58

53

(33)

(50)

11 (46)
(44)

(46)

* Functional and safe water points, namely, India Mark II + Tap.
Figures in parentheses represent percentages of surveyed

Bastees.

(30)

(i)

Water Supply
Most residents of JJ Bastees did not have access to

The usual

sufficient water to meet their daily needs.

sources were tap, shallow hand pump or deep bore hand pump.
On an average there was 1 safe water source (tap + deep

bore India Mark II hand pump) for every 102 households or 1

safe water source for a population of 511. The distribution
of these water sources was highly heterogeneous amongst the

various Bastees and also within the Bastee particularly in

the fringe areas.

About 12% of the clusters surveyed by us

(which included 8.5% of households) did not have even a

single source of safe potable water.

At places, people had

to travel a distance of at least 1 km to procure water.

The official agencies have attempted to supply water to
some of these areas through water tanks.

This is a fairly

expensive method with a limited storage capacity.
The available water supply was generally intermittent
and of low/moderate pressure.

Due to limited storage

capacity and nature of water supply,

observed.

large queues were

Innovations were done to extract water from taps

even with low pressure and

a large number of shallow hand

pumps were installed by the Bastee people.

The official

agencies have declared these shallow hand pumps unsafe for
drinking

purposes•

(31)
(ii) Garbage clearance

Only half the Bastees had a garbage collection point

(Dhaloa) for removal by the Municipal authorities. Usually
these Dhaloas were not regularly cleaned and constituted an
environmental hazard.

(iii) Drainage

The drainage was generally defective and /or blocked

leading to accumulation of slush.
(iv) Public Utilities
Roughly half

(46%)

the Bastees had pay and use latrine

facilities (Sulabh Shauchalays and mobile latrines).

were poorly maintained with shortage of water.
the population is, therefore,

These

Majority of

forced to resort to open

defecation (road side, fields, public parks, railway lines.
etc.).

A comparison of the civic amenities documented in this
survey with a recent study conducted

by Urban Basic serv-

ices (4) and the desired norms is presented in Table 5.
The results of the two surveys are comparable and it is

obvious that the civic amenities of this population are in
a deplorable state.

Urgent remedial action is necessary to

contain these important sources of health hazard.

(32)
TABLE 5 1 Comparison of Civic Amenities*

Parameter

Present

UBS Survey

Desired

study

(1990)

Norm

1:102

1:111

1:20

53

57

NA

1:85

Safe water points

(taps + Mark II)
per households

% Bastees with Public

Utilities
(Sulabh Shauchalaya)
No. of latrine seats/

1:20

person

B. DETAILS OF HOUSEHOLDS

The pertinent details of the households surveyed are
summarized in Table 6 and Figs. 3_ and 4.
houses visited, only 3204

Of the 22,181

(14.4%) could not be surveyed.

The important reasons for these included permanently locked

houses (67.8%),

only men or ineligible women residing

(24.3%), missed respondents (4.6%) and refusals (3.3%).

The co-operation for the survey was excellent.

FIG.3. HOUSEHOLDS NOT SURVEYED
Percentage
30 -/

25 -

20
15-

10 5I

0

Central

East

North

South

ZONES

West

Total

FIG.4. HOUSEHOLDS NOT SURVEYED
(REASONS)
Percentage

7

100%
75% -

50% -

■M

25% 0%

I

Central

East

North

South

West

Total

ZONES
Men/ineligible women

Refusals

Permanently Locked

Missed Respondents

(33)

The households not surveyed were mostly located in the
West (28.9%) and South (24.1%) Zones.

Permanently locked

houses constituted almost all (94.1%) these households in

West Zone and roughly half (52-63%) in other zones except
In the North Zone, the most frequent reason

North (38%).

was absence of eligible women.
Households

Surveyed

6

Details

Middle

East

o n
North

e s
South

West

Total

3911

3786

4834

4843

3807

22,181

145
(3.7)

134
(3.5)

655
(13.5)

1169
(24.1)

1101
(28.9)

3,204
(14.4)

a) Only men,
women <15 yr
or >49 yr +

10
(6.9)

57
(42.5)

276
(42.1)

381
(32.6)

57
(5.1)

781
(24.3)

b) Refusals+

49
(33.8)

4
(3.0)

17
(2.6)

34
(2.9)

1
(0.1)

105
(3.3)

86
c) Permanently**
(59.3)
locked

70
(52.3)

247
(37.7)

735
(62.8)

1036
(94.1)

2174
(67.8)

Nil

3
(2.2)

115
(17.6)

19
(1.7)

7
(0.7)

144
(4.6)

TABLE

Z

Details

1.Households
visited

of

2. Households not surveyed
Total*

d) Missed
respondents+
*

Figures in parentheses indicate percentage of total households
visited.

+ Figures in parentheses indicate percentage of total households
not surveyed.
** Residents gone to their

native village or uninhabitable house

FIG.5. AGE DISTRIBUTION OF WOMEN
(5 Year Groups)
Frequency (Thousands)
6

5 -

4-

3 -

2 -

1 -

0

I

15-19

20-24

25-29

30-34

35-39

AGE (Years)

40-44

45-49

(34)

EVALUATION OF MORTALITY DATA QUALITY

Age Distribution o£ Women
The age distribution of 15-49 year old women by 5 yearly

groups is depicted in Table 7 and Fig*5.
TABLE 7

Age Distribution of Women by Five Yearly Groups

Age group of

women

(Years)

Number of

Percentage

women

15-19

1940

9.85

20-24

4894

24.84

25-29

5333

27.07

30-34

3639

18.47

35-39

2136

10.84

40-44

999

5.07

45-49

760

3.86

Total

19701

100.00

In this survey, women above 40 years were slightly

under represented.

The percentage of 15-19 year old girls

also appears to be somewhat lower.

In this context.

it

would be important to remember that the urban poor popula­

tion surveyed primarily comprised nuclear families which

FIG.6. AGE DISTRIBUTION OF WOMEN
FREQUENCY
2500 -i-------------------------------

2000 -

1500 -

1000 KI

500 -

■—In—
il

0

I

15

17

I

I

19

I

I

21

I

Illi

23

25

I

I

27

I

29

II

31

I

I

33

I

I

35

I

I

37

AGE(YEARS)

I

I

39

I

I

I

41

43

I

I

45

47

49

(35)
had migrated from their native places in search of work.

It is, therefore, not surprising that roughly half the
women (51.91%) were aged between 20 to 29 years and four

fifths (80.76%) between 20 to 39 years.

There is a possi-

bility that younger and older women may not have been

sought in each household, or may have been missed because
interviewers were concerned to find women who had younger

children (3) .

This possibility could be true for a mortal­

ity questionnaire designed in conjunction with an EPI
coverage module.

However t this situation appears remote

for the present survey since the questionnaire was designed

specifically to estimate early childhood mortality and due
care (including quality checks) was taken to ensure that

information was recorded for all women in the 15-49 year
age group.

The observed

age distribution.

therefore,

appears to be a true reflection of the population surveyed.

The quality of age reporting of women can be further

assessed by a plotting a histogram of women 1s ages by

single years (Fig.6).

Age heaping occurs when age is

reported more often as a digit ending in

1 o • or •S'.

Despite all the innovations and care employed to ascertain

the age of women accurately, there is evidence of age
heaping in the data.
and

This is particularly marked at 25, 30
35

years.

FIG.7. DISTRIBUTION OF REPORTED
BIRTH MONTH
FREQUENCY
2000 |--------------------------

1500 -

1000

500 -

0“
1

i;

I
2

3

4

5

6

7

8

9

REPORTED MONTHS

10

11

12

FIG.8. DISTRIBUTION OF REPORTED
DEATH MONTH
FREQUENCY
250 Y”

20071

'rrm

150 ~

z:

loo­
se -

h



zz
0 zi|I

1

2

i

T

V

3

4

5

“T”

6

7

Z“

_]

8

i

9

REPORTED MONTHS

10

“T"

n

11

12

(36)

The observed age heaping can not be entirely consid­

ered to be representative of the quality of field work.

It

is primarily a reflection of the literacy status and atti-

tude of the population surveyed.

Birth records were virtu-

ally non existent and the exact date, month and year of

birth was not remembered by over 95% of women. Age in
completed years was often stated loosely or in a range.

Patience, perseverance and innovation was required to
ascertain the age.

Strict instructions were issued not to

guess the age in any woman.
This age heaping is not a serious deficiency in com­

puting child mortality by the Brass technique since the

method requires grouping of women by five yearly intervals.
2.

Age Reporting for Children
Birth month of children was reported in only 39.55% of

cases. The distribution of reported month of birth in these
children in shown in Fig.7.

There does not appear to be

any severe deficiency with respect to this data.
Amongst the children who had expired, the death month

was reported in only 35.32% of cases.

The distribution of

reported month of death in these children is shown in
Fig.8. Once again. there does not seem to be any severe

def iciency

with

respect

to

this

data.

(37)

In roughly two thirds of the cases, the month of birth
or death was not reported.

This is primarily a reflection

of the literacy status, culture and attitude of the popula-

tion surveyed.

These deficiencies do not affect

the Brass

estimates or the Preceding Birth Technique, which do not
Live tables.

rely on dates at all.

however,

are very

sensitive to such misreporting.

3.

Sex Ratios of Children

Table 8 summarizes the sex ratios of children ever born and
those who died by five year age groups of women.

TABLE 8 j_ Sex Ratios by Five Year Age Groups of Women

Age group

Male CEB/
Female CEB

Total males dead/
total females dead

15

19

1.02

1.00

20

24

1.09

1.19

25

29

1.09

1.02

30

34

1.12

1.06

35

39

1.13

1.11

40

44

1.18

1.11

45

49

1.20

1.20

1.12

1.09

Total

CEB = Children ever born.

(38)
The sex ratios for children ever born should usually
fall within the range 1.02 to 1.07 and should not vary

greatly by age of mother, but they may vary substantially

for small samples (3) .
There was marginal under-reporting of female children

in the 20-24 and 25-29 years age group which increased
slightly till 35-39 years age group and was substantial for

40-44 and 45-49 year age groups.

The sex ratios for deaths

were mostly lower than the sex ratios of children ever born

except for the 20-24 year age group.

The sex ratios for

dead children were above 1.07 in the 20-24, 35-39, 40-44
and 45-49 year age groups.

There is thus a possibility of slight under-reporting
of female births, probably those who died.

This is partic­

ularly true for the 20-24 and 45-49 year age groups. Thus
we may assume that our calculations derived from these age
groups may be slight underestimates.

The possibility of a

relatively smaller sample size for the older age groups

(>40 years) can not, however, be totally discounted.
4.

Average Parities by

e of Mother

The average parities by five yearly age groups for all
children and separately for males and females are depicted

in

Table

2..

(39)
The average parities for male and female children and
both sexes combined increased with maternal age and there
was no deviation from the expected pattern. The present
data is, therefore, reliable in this context.

TABLE 9 £. Average Parities by Age Group of Women
Samples
)
in
(All
Women
in.
C22:
Male

Parities
Average
Both Sexes
Female

Age group

N

15-19

1940

0.17

0.17

0.34

20-24

4894

0.84

0.77

1.61

25-29

5333

1.58

1.46

3.04

30-34

3639

2.22

2.00

4.22

35-39

2136

2.61

2.33

4.94

40-44

999

2.94

2.49

5.44

45-49

760

3.17

2.65

5.82

Total

19701

1.62

1.45

3.07

The average parities for the oldest age group of

women, who have completed, or nearly completed their child

bearing, show an approximate average completed family size
which indicates a fairly high fertility.
5.

Proportion Dead of Children Ever Born

Table 10 summarizes the proportion dead of children

ever born (males, females and both sexes) by five year age
groups of women.

(40)

TABLE 10: Proportions Pead of Total Children Ever Born
dead
Both sexes

Age group
(years)

All
women

Male

Proportions
Female

15-19

1940

0.154

0.157

0.156

20-24

4894

0.146

0.133

0.140

25-29

5333

0.140

0.150

0.145

30-34

3639

0.161

0.170

0.165

35-39

2136

0.180

0.183

0.181

40-44

999

0.201

0.214

0.206

45-49

760

0.215

0.214

0.219

Total

19701

0.165

0.169

0.167

These proportions increased steadily with age (for

both sexes) after the 20-24 year age group.

However, for

males, the proportion was lower in 25-29 year age group in

comparison to the earlier two age groups.

A higher child­

hood mortality in the younger age group women is well
documented (3) and is the probable explanation for these

observations.

There does not appear to be a marked rise in

proportions after the 25-29 year age groups; data, therefore, appears to be fairly reliable in this context.
D.

ESTIMATES OF CHILDHOOD MORTALITY
Considering the data quality discussed above. the

indirect methods.

namely.

the Brass Technique and the

(41)
Preceding Birth Technique would be the most appropriate for

this situation.

1.Brass Technique
Tables 11 to

13

depict the mortality estimates from

the Q FIVE computer software supplied by the United Nations
(6). The population surveyed is best represented by the

South Model in the Coale-Demeny/Trussel equations and South
Asian in the United Nations Palloni-Heligman equation. The

under five mortality rate, i.e.f q(5) represents the number
of deaths in under five children per thousand live births
whereas infant mortality rate, i.e^, q(l) refers to deaths
below one year of age per thousand live births.

The proba-

bility of dying between 1 to 5 years of age is the differ-

ence of these two parameters.

The estimates of childhood

mortality by the two equations are not markedly variant.
Since a default value of mean age at maternity was utilized

for the Palloni-Helignman equations and also for the sake

of simplicity, only the South model of Coale Demeny version

will be referred to for the mortality estimates.

The under

five mortality rate (q 5) shows the least variation amongst
This
different models and is most stable to errors (3) .

should, therefore, be the preferred estimate of mortality.

(42)
table

11: Input Data for Delhi Bluns (Both Sexes)

- Enumeration Date : July, 1991

Age Group

Number

Number of

Number of

of

of

Children

Children

Women

Women

Ever Born

Dead

15-19

1940

665

104

20-24

4894

7872

1099

25-29

5333

16224

2348

30-34

3639

15357

2540

35-39

2136

10541

1913

40-44

999

5436

1118

45-49

760

4421

968

A default value of 27.0 for mean age at maternity was not
utilized for Palloni Helligman Equations.

(43)
TABLE 12 : Indirect Estimation of ChilAood Mortality for
Both Sexes (Coale - Deaency Models)
Enuaeration Date: July, 1991

I

KSKZZZZSZZZZZZZZZSZZSCZZZZXZSZSZZZZZXZZZZXZZZZZZZSZZXZSZKZZZZZ3ZZXZKZZZZZZKZZKZZZZCZXZZZZZZZZZSZZEZZZZZZZZZZZZ

Age of
Woman

Pro­
portion
Dead

Average No.
Children
Born Surviving

Coale-Demeny Models
(Trussell Equations)

Age

q(x)

15-19
20-24
25-29
30-34
35-39
40-44
45-49

.343
1.609

3.042

2.602

4.220

3.522
4.039

4.935
5.441
5.817

156
140
145
165
181
206
219

.289
1.384

4.322
4.543

South

North

x

t(x)

.147 ( 1.2)
.132 ( 2.6)
.134 ( 4.5)
.160 ( 6.7)
.188 ( 9.1)
.211 (11.7)
.220 (14.5)

1
2
3
5
10
15
20

q(x)

t(x)

.141
.139
.144
.167
.187
.208
.219

( 1.2)
( 2.6)
( 4.6)
( 6.9)
( 9.5)
(12.3)
(15.4)

East
q(x)
t(x)

.156

.141

.143

.165
.185
.207

.218

( 1.2)
( 2.7)
( 4.7)
( 7.1)
( 9.7)
(12.6)
(15.8)

q(x)

West
t(x)

152 ( 1.2)
140 ( 2.6)
142 ( 4.7)
165 ( 7.0)
185 ( 9.5)
207 (12.2)
218 (15.1)

:z:

Coale-Demeny:

Age of
Woman

North

South

Reference

Reference

East

West

Reference

Reference

q

Date

q

Date

q

Date

q

.147
.104
.095
.098
.099
.103
.100

1990.3
1988.9
1986.9
1984.6
1982.0
1979.2
1976.1

.141

.120

1990.3
1988.9
1986.8
1984.4
1981.8
1978.9
1975.8

.156
.121
.116
.125
.130
.139
.139

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.5

.152
.113
.106
.113
.116
.122
.119

1990.3
1988.9
1986.9
1984.6
1982.0
1979.2
1976.1

.104
.062
.056
.064
.068
.076
.075

1990.3
1988.9
1986.8
1984.4
1981.8
1978.9
1975.8

.063
.044
.041
.046
.049
.054
.053

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.5

.089
.058
.053
.058
.060
.065
.063

1990.3
1988.9
1986.9
1984.6
1982.0
1979.2
1976.1

.231
.164
.154
.167
.175
.187
.186

1990.3
1988.9
1986.8
1984.4
1981.8
1978.9
1975.8

.209
.159
.152
.165
.173
.185
.185

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.5

.227
.165
.154
.165
.169
.179

Date

Infant Mortality Rate : q(1)
15-19
20-24

25-29

30-34
35-39
40-44
45-49

1990.3
1989.0
1987.1
1984.9
1982.4
1979.8
1977.0

Probability of Dying Between Ages 1 and 5:
15-19

20-24

25-29

30-34
35-39
40-44
45-49

1990.3
1989.0
1987.1
1984.9
1982.4
1979.8
1977.0

.118
.076

.065

.069
.070
.074
.071

.109

.104
.110
.114

.120

q
4 1

Probability of Dying by Age 5: q(5)

15-19
20-24
25-29
30-34
35-39
40-44
45-49

1990.3
1989.0
1987.1
1984.9
1982.4
1979.8
1977.0

.248
.172

.154
.160
.163
.170
.164

Note: A q value of .999 Denotes value below A Level 1 Model Life Table
«
.000

Above A Level 25
"

.175

(44)
Table 13 £ Indirect e»ti■ation of Chilcbood Mortality for Both Sexes (Palloni - Neligaan Equations)
^XX^^XZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

Ageof
Woman

Pro­
portion Age
Dead x

Average No.
ChiIdren
Born Surviving

United Nations Models
(Palloni-Heligman Equations)

15-19
20-24
25-29
30-34
35-39
40-44
45-49

.343
1.609

.289
1.384

3.042
4.220
4.935
5.441

2.602
3.522
4.039
4.322

5.817

4.543

.156
.140
.145
.165

.181
.206
.219

Mean Age at Maternity = 27.00

1
2
3
5
10
15
20

.150
.142
.145
.168
.188

.204
.219

( 1.1)
( 2.5)
( 4.3)
( 6.4)
( 8.8)
(11.6)
(15.2)

So.Asian

Chilean

Latin An.
t(x)

q(x)

q(x)

t(x)

.167

( 1.3)
( 2.7)
( 4.5)
( 6.7)
( 9.2)
(12.0)
(15.6)

-147
.147
.167
.184

.205
.218

Far East

q(x)

t(x)

q(x)

t(x)

.149

( 1.1)
( 2.5)
( 4.4)
( 6.6)
( 9.0)
(11.9)
(15.8)

.153
.142

( 1.2)
( 2.6)
( 4.4)
( 6.5)
( 8.9)
(11.5)
(14.8)

.143
.146
.169
.188

.210
.220

.143
.165
.184
.204
.218

General
q(x)
t(x)
.151

.142
.144
.166
.187

.204
.219

( .1)
( 2.5)
( 4.3)
( 6.5)
( 8.9)
(11.6)
(15.1)

(Default value used)

— s-=-===5-XX= = XZS=XXXXSS = Z=== = ZZ=ZZZZZZ=ZZZZZZZ==Z=Z=Z===ZZZ = = === = ===Z= = ZS = === = = ===== ===: ======== = === = = = = = === = = =

Chi lean

United Nations : Latin Am.

Age of
Woman

Infant Mortality Rate:

q(1)

1990.5
1989.1
1987.3
1985.1
1982.7
1979.9
1976.4

.150

15-19
20-24

25-29
30-34
35-39
40-44

45-49

Reference
Date
q

Reference
Date
q

1990.2
1988.9
1987.0
1984.8
1982.3
1979.5
1976.0

.110
.102

.105
.107
.111

.112

So.Asian

Far East

General

Reference
Date
q

Reference
Date
q

Reference
Date
. q

.167
.130
.124
.132
.137
.145
.144

1990.5
1989.0
1987.2
1985.0
1982.5
1979.6
1975.8

.149

.062

1990.5
1989.0
1987.2
1985.0
1982.5
1979.6
1975.8

.113
.070

.112
.104
.110

.113
.120
.122

.151
.114
.105

.110
.113
.109

1990.4
1989.0
1987.2
1985.1
1982.7
1979.9
1976.4

1990.3
1988.9
1987.2
1985.0
1982.7
1980.0
1976.8

.109
.067
.059
.062
.063
.066
.061

1990.4
1989.0
1987.2
1985.1
1982.7
1979.9
1976.4

.110
.068
.060

1990.3
1988.9
1987.2
1985.0
1982.7
1980.0
1976.8

.245
.173
.158

1990.4
1989.0
1987.2
1985.1
1982.7
1979.9
1976.4

.245

1990.3
1988.9
1987.2
1985.0
1982.7
1980.0
1976.8

.153
.114
.106
.109

.109

.111

.115
.114

Probability of Dying Between Ages 1 and 5:
15-19
20-24
25-29
30-34
35-39
40-44
45-49

1990.5
1989.1
1987.3
1985.1
1982.7
1979.9
1976.4

q
4 1
1990.2
1988.9
1987.0
1984.8
1982.3
1979.5
1976.0

.124
.075
.065
.070

.072
.075
.077

Probability of Dying by Age 5:
.255
1990.5
15-19
.177
20-24
1989.1
.160
25-29
1987.3
30-34
1985.1
.168
35-39
1982.7
.171
.178
1979.9
40-44
.180
1976.4
45-49

.040
.037
.041

.043
.048
.048

.062
.067

.071
.079

.080

1990.2
1988.9
1967.0
1984.8
1982.3
1979.5
1976.0

.218
.165
.156
.167

.174
.186

.185

1990.5
1989.0
1987.2
1985.0
1982.5
1979.6
1975.8

.245

.174
.160
.169
.175
.189
.192

iSXSSSSSZSSZSZZSSSSSSSSSSSSSSSS

Note: A q value of .999 Denotes value from Table with life expectancy less than 35

.000

.066
.069

.069

q(5)

Z2ZZZZZZZZZZZZSZZZZZ2ZZZZZZZZ1

"

.064

"

Greater than 75

.165
.166
.171
.163

.174
.159
.166
.170
.176
.175

(45)
Figs. 9 to 11 depict the estimated infant mortality.

under 5 mortality and 1 to 5 year mortality over the years.

In this context, it is important to remember that many of
these JJ Bastees came into existence within the last 10 to

15 years. The pre 1980 estimates should,

therefore, be

viewed as those of this population and not the urban slum

area.
The values for the two youngest age groups depart from

the observed trend due to the known biases associated with
the estimates from these age groups (3) :

(i)

Children of the youngest women, 15-19, are known to experi-

ence higher majority of their children than are those of
older women.

(ii) The children of the youngest women are more likely to be
first births, who are also likely to have higher mortality
than all births in a population.

(iii)This bias is especially strong in later-marrying popula­
tions where the women bearing children at younger ages are
more likely to be a selected group, for instance women who
left school earlier than usual, women from poorer groups in

the

population,

etc .

For all these reasons, estimates based on the mortali-

ty experience of these women is unlikely to reflect the
overall mortality experience of the population.

The

FIG.9.ESTIMATED INFANT MORTALITY

150

140
CO
X
X

S 130
o
o

s

120

E

110

100

>■

1975

1

I

I

1979

1983

1987

YEAR

I

I

1991

FIG. 1 ©.ESTIMATED UNDER 5 MORTALITY

240

220
(0
X
X

2 200
o
o

s

180

(E

160

140 ■—“■

1975

I

I

I

1979

1983

1987

YEAR

1991

FIG. 11.ESTIMATED 1-5 YR MORTALITY

no
100
(0
X
F
X

90

o
o

80

s

70

2

<E

60

50 ■—L
1975

I

1

I

1979

1983

1987

YEAR

1

,

I

1991

(46)
indirect estimation methods use models of fertility and

mortality to distribute these births and deaths in time and
by age.

Even though no information on age of these chil-

dren is used in the estimation procedure, by this method.

I

the two youngest age groups give us the estimated probabil­

ity of dying by age one (infant mortality rate)
probability of dying by age two.

and the

They are also the esti-

mates which relate to the closest time periods to the

survey (around 1-5 and 3-4 years prior) because, on average. mothers from these age groups gave birth more recently

than older mothers (3).
For the reasons stated above these estimates of mor­
tality do not accurately reflect the probabilities of death

by ages one and two for the whole population. Older mothers

in the population have children who are likely to have
higher probabilities of surviving to these ages, yet, their

mortality experience does not figure in the estimates.

It

is best. therefore, to disregard the estimates from the 1519 and 20-40 year old women. and to look for other methods

of obtaining mortality estimates for the more recent period
(3).
This does not necessarily contradict the advice given
above for translating estimates from older mothers into one
mortality

index

for

each

group.

It

only

means

that

(47)
whatever index we use, we should disregard the estimates

obtained for the mortality levels of the two youngest age
groups (3).

On the graph (Figs. 9.

to 12)

it may appear as though

mortality has risen recently, when it only means that these
indirect estimates for the time closest to the survey are

not representative of levels of childhood mortality in the
population as a whole (3).
One of the assumptions implicit in the Brass method is

that the age pattern of fertility has been constant in the
recent past.

In effect, we assume that the age pattern of

current fertility is the same as that earlier. Hence,one

possible explanation of the irregular pattern observed in
some surveys is that the ratio of average parities is a

poor description of the prevailing age pattern of fertility.

Another quite different possibility is age error.

A

third possibility is that the survey missed certain women,
which biased the estimates (3). There are no such obvious
irregular patterns observable in the present survey.

A method which produces an estimate nearer to the
survey date without reliance on death dates is the Preced-

ing Birth Technique.

(48)

It would be useful to examine the data separately for boys
and girls. Tables 14 to 17 summarize the childhood mortali-

ty for boys and girls, separately.

The under five mortali-

ty rate (q 5) and infant mortality

rate (q 1) from the

South model is compared for both sexes in Figs. 12

and 13.

The infant mortality for males is marginally higher

than females and shows a downtrend till the last two age
groups.

The downtrend in female ql starts from 40 to 44

years age group (reference period 1979.1) and continues

till 20-24 year age group (reference period 1989) after
which the expected upsurge is obvious. These findings are

probably due to under-reporting of female deaths in the 4549 and 20-24 year age groups, the male to female sex ratio

of dead children in these two age groups is high (1.19 and
1.20, respectively).

The infant mortality estimates refer-

ring to these two age periods should, therefore, be considered as underestimates.
On considering under five mortality, a gradual decline for

males is obvious till the youngest two age groups. For
females, as for ql. the decline for q5 begins from 40-44

year age group (reference period 1979.1) and lasts till 20-

24 year age group (reference period 1989).

This once again

suggests that the mortality estimates for these two period
are

underestimates.

FIG.12.SEX & UNDER 5 MORTALITY

240

220

co

oc 200
ffi
o
180
AC

160

140 *—L
1975

1

1

I

1979

1983

1987

YEAR

J
1991

v Q5F
• Q5M

FIG. 13.SEX & INFANT MORTALITY

150 r

140
CO
X
F
tt

2 130
o
o
120
AC

110

100 L-■1975

I

1079

1983
YEAR

1987

I

1991

v Q1F
• Q1M

kzzsccz:

Age of
Woman

Average No.
ChiIdren
Born Surviving

Pro­
portion
Dead

Coale-Demeny Models

(Trussell Equations)

Age
x

15-19
20-24

.174
.838

.147

25-29
30-34

1.585

1.363
1.864

2.223

.154
.146
.140
.161
.180
.201
.215

.715

35-39
40-44

2.615

2.939

2.144
2.347

45-49

3.170

2.488

.147
.139
.130
.156
.186
.206
.216

1
2
3
5
10
15
20

q(x)

t(x)

qCx)

t(x)

Wast
q(x)
tCx)

.141
.146
.139
.163
.185
.203
.215

( 1.2)
( 2.6)
( 4.6)
( 7.0)
( 9.6)
(12.4)
(15.5)

.155
.148
.138
.161
.184
.202
.214

( 1.2)
( 2.6)
( 4.7)
( 7.1)
( 9.8)
(12.7)
(15.9)

.151 ( 1.2)
.147 ( 2.6)
.137 ( 4.7)
.161 ( 7.0)
.183 ( 9.6)
.202 (12.3)
.214 (15.2)

East

South

North
t(x)

q(x)

( 1.2)
( 2.6)
( 4.5)
( 6.7)
( 9.1)
(11.8)
(14.6)

:s=zz=zzzszzzz=zzzz=zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz====zz

«XXZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ:

Coale-Demeny:

Month

Age of
Woman

Reference
Date
q

15-19
20-24
25-29
30-34

35-39
40-44
45-49

15-19
20-24
25-29
30-34

35-39
40-44
45-49

q(1)

.147
.111
.094
.099
.102
.105
.102

1990.4
1989.0
1987.0
1984.6
1982.0
1979.1
1976.0

1990.4
1989.0
1987.1
1984.9
1982.4
1979.8
1976.9

20-24
25-29
30-34
35-39

40-44
45-49

.141
.115
.104
.111
.116
.121
.121

Probability of Dying Between Ages 1 and 5:
4 1
.095
1990.4
1990.4
.111
.063
1989.0
.076
1989.0
1987.0
.050
.060
1987.1
.058
1984.6
.064
1984.9
.065
1982.0
.067
1982.4
.071
1979.1
.070
1979.8
.071
1976.0
.067
1976.9

Probability of Dying by Age 5:
15-19

1990.4
1989.0
1987.1
1984.9
1982.4
1979.8
1976.9

.241
.179
.148
.156
.161
.167
.162

Reference
Date
q

Reference
Date
q

Infant Mortality Rate:

1990.4
1989.0
1987.0
1984.6
1982.0
1979.1
1976.0

West

East

South

Reference
Date
q

1990.3
1988.9
1986.8
1984.4
1981.7
1978.9
1975.7

.155
.129
.114
.125
.133
.140
.140

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.4

.151
.121
.106
.114
.119
.124
.122

1990.3
1988.9
1986.8
1984.4
1981.7
1978.9
1975.7

.056
.043
.036
.041
.045
.049
.049

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.4

.080
.058
.047
.053
.056
.060
.058

1990.3
1988.9
1986.8
1984.4
1981.7
1978.9
1975.7

.203
.166
.146
.161
.172
.182
.182

1990.3
1988.9
1986.9
1984.5
1982.0
1979.3
1976.4

.220
.172
.148
.161
.168
.177
.174

q

q (5)
.223
.171
.148
.163
.174
.183
.184

zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzxzzzzz

Note: A q Value of .999 Denotes Value Below A Level 1 Model Life Table
"
.000
"
Above A Level 25
"

(50)
ition
of
Chil<t»ood
Mortality
ffir palea (Pallgni. z Neligman Fquatiom)
TABLE J5 1 Indirect Eatii

tXXBXXXBXXXXXXXXXXXXXBXXXXBXXXXXXXBXXXXBXXBXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX:-:l:==S:r=s=x==x:xx:=:==:==:=====

KB9

Average No. ProAge of
Children
portion Age
Woman Born Surviving Dead
x
X

United Nations Models
Latin Am.

q(x)

t(x)

(PalIoni-HeIigman Equations)
Far East
Chilean
So Asian
tCx)
t(x) q(x)
q(x)
t(x)
q(x)

( 1.1) .165
1 .148
.154
.147
.174
15-19
.149
( 2.5) .154
.146
2
.715
.838
20-24
.140
( 4.3) .143
3
.140
1.363
25-29 1.585
.164
( 6.4) .163
5
1.864
.161
30-34 2.223
( 8.9) .182
.186
10
.180
35-39 2.615 2.144
(11.7) .201
.200
15
.201
40-44 2.939 2.347
(15.3) .214
.215
20
.215
45-49 3.170 2.488
(Default
Value
used)
Mean Age at Maternity s 27.00

( 1.3)
( 2.6)
( 4.5)
( 6.8)
( 9.3)
(12.1)
(15.7)

.148

( 1.D .151

.150
.141
.165
.187

( 2.5)
( 4.4)
( 6.6)
( 9.1)
(12.0)
(15.9)

.205
.216

.149
.139
.161
.182
.200
.214

( 1.2)
( 2.6)
( 4.4)
( 6.5)
( 8.9)
(11.6)
(14.9)

General
q(x)
t(x)
.150
.149
.139
.162
.185
.200
.215

( 1.1)
( 2.5)
( 4.3)
( 6.5)
( 8.9)
(11.7)
(15.2)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx:IXXXXXXXXXXXXXBXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

United Nations: Latin Am.

Age of
Woman

Reference
Date
q

Infant Mortality Rate:

q(1)

1990.5
1989.1
1987.3
1985.1
1982.6
1979.9
1976.3

.148

15-19
20-24
25-29
30-34
35-39
40-44
45-49

.120
.103
.109
.114
.116
.117

Chilean

So Asian

Far East

General

Reference
Date
q

Reference
Date
q

Reference
Date
q

Reference
Date
q

.151
.122
.105
.110
.112
.114
.111

1990.4
1989.0
1987.2
1985.1
1982.6
1979.9
1976.4

.150

.121
.123

1990.3
1988.9
1987.2
1985.0
1982.6
1980.0
1976.7

.103
.072
.057
.063
.068
.074
.076

1990.3
1988.9
1987.2
1985.0
1982.6
1980.0
1976.7

.095
.067
.053
.057
.059
.060
.057

1990.4
1989.0
1987.2
1985.1
1982.6
1979.9
1976.4

.089

.236

1990.3
1988.9
1987.2
1985.0
1982.6
1980.0
1976.7

.232
.181

1990.4
1989.0
1987.2
1985.1
1982.6
1979.9
1976.4

.225

.148

.147
.147

1990.5
1989.0
1987.2
1985.0
1982.5
1979.5
1975.7

q
4 1
.051
1990.2
P988.9
.038
1987.0
.031
.035
1984.8
.039
1982.3
.042
1979.4
.042
1975.9

1990.5
1989.0
1987.2
1985.0
1982.5
1979.5
1975.7

1990.2
1988.9
1987.0
1984.8
1982.3
1979.4
1975.9

.165

.139
.123
.133

.140

.119
.103

.109

.114

.123

.107
.113

.117

.120
.120

Probability of Dying Between Ages 1 and 5:

15-19
20-24
25-29
30-34
35-39
40-44
45-49

1990.5
1989.1
1987.3
1985.1
1982.6
1979.9
1976.3

.100
.070
.056
.061

.065
.067
.068

.064
.051
.056

.059

.061
.062

Probability of Dying by Age 5: q(5)
15-19

20-24
25-29
30-34
35-39

40-44
45-49

1990.5
1989.1
1987.3
1985.1
1982.6
1979.9
1976.3

.233
.181

.153
.164
.171
.175
.178

1990.2
1988.9
1987.0
1984.8
1982.3
1979.4
1975.9

.208
.171
.150
.163
.173
.183
.183

1990.5
1989.0
1987.2
1985.0
1982.5
1979.5
1975.7

.182
.154
.165
.174
.186
.190

.153

.161
.164

.167
.162

.179
.152
1162

.169
.174
.175

sx=-==-s---s---=s-x=«====xxx===x===xxsxx==«==xe===xx==x=x===s==s=========:========s:====;s=:ss:===:s==:s=ss==========

Note: A q Value of .999 Denotes Value from Table with Life Expectancy Less Than 35

,ooo
"
Greater Than 75

(51)
TABLE 16 i Indirect Eatimation of Early Age Mortality for Females (Coele Demeny Models)
— ——— —

—— — w—XWXZZZXZZXZZZZZZZZZZZZZZXZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

Age of
Woman

Average No.
ChiIdren
Born Surviving

15-19
20-24

.171
.771

25-29
30-34

1.455
1.994

35-39
40-44

2.318

1.893

2.490
2.647

2.082

45-49

Pro­
portion Age
Dead
x

Coale-Demeny Models
(Trussell Equations)
South
North
q(x)
t(x)
t(x)
q(x)

.157
.133
.150
.170
.183
.214
.214

.143
.125
.139
.166
.191
.220
.216

.144
.668
1.237

1.654

1.958

1
2
3
5
10
15
20

( 1.3)
( 2.6)
( 4.5)
( 6.7)
( 9.0)
(11.6)
(14.4)

.137
.131
.149
.172
.190
.217
.214

East
t(x)
q(x)

.153
.134
.148
.170
.188
.216
.214

( 1.3)
( 2.6)
( 4.6)
( 6.9)
( 9.5)
(12.2)
(15.3)

( 1.3)
( 2.7)
( 4.7)
( 7.1)
( 9.7)
(12.5)
(15.7)

:---SXXXXXZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

15-19
25-29

30-34
35-39

40-44
45-49

West

Reference
Date
q

Reference
Date
q

Reference
Date
q

Reference
Date
q

q(1)

1990.3
1988.9
1987.0
1984.9
1982.5
1979.9
1977.1

.143
.097
.095
.097
.096
.102
.094

1990.3
1988.9
1986.9
1984.6
1982.1
1979.3
1976.2

.137
.102
.104
.110
.112
.120
.114

1990.3
1988.8
1986.8
1984.4
1981.9
1979.0
1975.9

.153
.113
.117
.125
.127
.139
.130

1990.3
1988.8
1986.9
1984.6
1982.1
1979.4
1976.6

.148
.105
.107
.112
.112
.120
.111

1990.3
1988.8
1986.8

1990.3
1988.8

.095
.060

1986.9

.061

1984.6
1982.1
1979.4

1975.9

.070
.045
.047
.052
.053
.061
.055

1976.6

.066
.066
.073
.065

1990.3
1988.8
1986.8
1984.4
1981.9
1979.0
1975.9

.212
.153
.159
.170
.174
.191
.179

1990.3
1988.8
1986.9
1984.6
1982.1
1979.4
1976.6

.229
.159
.161
.170
.170
.184
.168

q
4 1

15-19

1990.3

.124

1990.3

20-24

1988.9

.075

1988.9

25-29

1987.0
1984.9
1982.5
1979.9
1977.1

.073

1986.9
1984.6
1982.1
1979.3
1976.2

.110
.059
.063
.070
.073
.084
.076

1990.3
1988.9
1986.9
1984.6
1982.1
1979.3
1976.2

.232
.156
.161
.172
.177
.193
.181

35-39
40-44
45-49

.076

.075
.081
.072

Probability of Dying by Age 5:

15-19
20-24
25-29
30-34
35-39
40-44
45-49

XXSSZXSSSZZKKSZ

East

Probability of Dying between Ages 1 and 5:

30-34

( 1.3)
( 2.7)
( 4.7)
( 7.0)
( 9.5)
(12.1)
(15.0)

South

Infant Mortality Rate:

20-24

148
132
147
170
187
216
214

North

Coale-Demney:

Age of
Woman

West
t(x)
q(x)

1990.3
1988.9
1987.0
1984.9
1982.5
1979.9
1977.1

1984.4

1981.9
1979.0

q(5)

.249
.165
.161
.166
.164
.175
.159

zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzszzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz

Note: A q Value of .999 Denotes Value Below A Level 1 Model Life Table
"
.000

Above A Level 25

COMMUNITY
HEALTH CELL
32b, V Main,
I Block
Koranic ng ia
®angalore-56Q034
India

(52)
TABLE 17 i Indirect gatimtion of Early Age Mortality for Fe—lies (Palloni Heli flman fguatiml
Bas-E-sscs-SXSSX==XXXSSSES=2XaX= SX8XXBSXSXSXX = SS = XXXXS = SSS:S = XXX*XX== = S= = S=«=s = :=Ss::=: = = = S = = = S:SS= = = = S = C==S = =

Age of
Woman

Pro­
Average No.
portion Age
Children
x
Born Surviving Dead

United Nations Models
(Palloni-Heligman Equations)

Latin Am
t(x)
q(x)

So Asian

Chilean

Far East
t(x)

q(x)

t(x)

q(x)

t(x)

165
.147 ( 1.1)
1
.144 .157
.171
140
.135
( 2.5)
2
.668 .133
.771
153
.150
( 4.3)
3
1.237 .150
1.455
173
( 6.4)
.173
5
1.654 .170
1.994
186
.190
( 8.8)
10
1.893 .183
2.318
214
(11.5)
15
.212
1.958 .214
2.490
213
(15.0)
.214
20
2.082 .214
2.647
45-49
Mean Age at Maternity = 27.00 (Default Value Used)

( 1.3)
( 2.7)

.147
.136
.151
.174
.191
.218
.215

( 1.1) .151

15-19

20-24
25-29
30-34
35-39
40-44

United Nations: Latin Am

Age of
Woman

Reference
Date
q

Infant Mortality Rate:
15-19
20-24
25-29
30-34
35-39
40-44
45-49

1990.5
1989.0
1987.2
1985.1
1982.8
1980.1
1976.5

1990.5
1989.0
1987.2
1985.1
1982.8
1980.1
1976.5

( 9.1)
(11.9)

(15.4)

( 2.6) .134
( 4.4) .149
( 6.5) .170
( 8.9) .186
(11.8) .212
(15.6) .212

General
q(x)
t(x)

( 1.3) .150
( 2.7) .135

( 4.4) .149
( 6.5) .171
( 8.8) .189
(11.4) .212
(14.6) .214

( 1.2)
( 2.6)
( 4.3)
( 6.5)
( 8.8)
(11.5)
(14.9)

Chi lean

So Asian

Far East

General

Reference
Date
q

Reference
Date
q

Reference
Date
q

Reference
Date
q

q(1)

.147
.101
.099
.101
.101
.106
.102

1990.2
1988.8
1987.0
1984.8
1982.4
1979.7
1976.1

Probability of Dying Between Ages 1 and 5:

15-19
20-24
25-29
30-34
35-39
40-44
45-49

( 4.6)
( 6.7)

q(x)

.152
.079
.077

.079
.079
.086

.080

1990.2
1988.8
1987.0
1984.8
1982.4
1979.7
1976.1

.165
.122
.125
.132
.134
.145
.137
q
4 1
.071

.040
.042

.047
.048
.056
.050

1990.5
1989.0
1987.1
1985.0
1982.6
1979.8
1976.0

.147
.106
.106
.110
.112
.121
.116

1990.3
1988.9
1987.1
1985.0
1982.7
1980.2
1976.9

.999

1990.5
1989.0
1987.1
1985.0
1982.6
1979.8
1976.0

.120
.067

1990.3
1988.9
1987.1
1985.0
1982.7
1980.2
1976.9

.999

.067

.072
.074
.085
.079

.107
.107
.109

.108
.113
.103

.066

.065
.068

.067
.073
.062

1990.4
1989.0
1987.2
1985.1
1982.8
1980.1
1976.6

.150
.105
.104
.106
.105
.111
.105

1990.4
1989.0
1987.2
1985.1
1982.8
1980.1
1976.6

.135
.071
.070
.073
.072
.079
.072

Probability of Dying by Age 5:q(5)
15-19

20-24
25-29
30-34
35-39
40-44
45-49

1990.5
1989.0
1987.2
1985.1
1982.8
1980.1
1976.5

.277
.172

.168
.173

.172
.183
.174

1990.2
1988.8
1987.0
1984.8
1982.4
1979.7
1976.1

.224
.157
.162
.173
.176
.193
.180

1990.5
1989.0
1987.1
1985.0
1982.6
1979.8
1976.0

.249

.165

.166
.174
.177
.196
.186

xx=xx==xxxxxxxxx==xxx=xxx=xxxx==x=x=x==sx==xxx=xxxxxxx=xsxxxxxxxx=x:

1990.3
1988.9
1987.1
1985.0
1982.7
1980.2
1976.9

.999

.166
.165
.170
.168
.179
.159

1990.4
1989.0
1987.2
1985.1
1982.8
1980.1
1976.6

.264
.169

.167
.171
.170
.181
.169

SSSSXSXSSSSSSSXSSSSSB&XBSBSXSSXZSS

Note: A q Value of .999 Denotes Value from Table with Life Expectancy Less Than 35

.000
*
Greater Than 75

(53)
In contrast to infant mortality, the under five mor-

tality is higher for females.

This should not lead us to

question the reliability of the date.

A more logical

explanation is that the under five mortality (particularly

1 to 5 year mortality) is actually higher for females.
This is compatible with the well known preference for male
children, particularly in this situation.
2.

Preceding Birth Technique

With this technique we can derive reliable estimates

of childhood mortality closer to the date of survey.

This

method, is quite robust to error (3).

There were a total of 6485 women with the youngest
live born child below 2 years age (only those children were

considered whose age was known with certainty) and a pre-

ceding birth also.

There were 878 deaths amongst the

preceding births, giving the proportion dead as 0.135.

The mean age of all last births (10.4 months) was

calculated from 5834 cases in whom the age of child in

I

years and months was known with certainty.

The mean birth interval length for the interval bet-

ween the last and preceding birth was 32.85 months.

This

was calculated from 4295 cases in whom the ages of both

I

i

(54)
children (youngest child and preceding birth) in years and

months were known with certainty.
The estimate of 135 deaths per thousand live births,
therefore, refers to 36.7 months or roughly three years (q3

or under three mortality rate).

This refers to a time

period 28.8 months ago (roughly 2 years 5 months ago) or to
the year 1989.2.
Assuming a design effect of 2.5 for cluster sampling.
the 95% confidence intervals for the estimated q3

(under

three mortality) are 122 to 148.
To make comparisons with the Brass Technique meaningful, the q3 requires conversion to common indices q5 and

ql)

The converted q5 is 146 and ql is 100.
From the section on Brass Technique we are aware that

the estimates from the two youngest age groups should be
disregarded as overestimates.

We can substitute those

estimates by the converted under five (q5) and infant (ql)
mortality rates derived from the Preceding Birth Technique.
Figs. 14 and 15 graphically depict the estimated secular
change for both sexes combined for q5 and ql.

Only the

last value is derived from the Preceding Birth Technique

the

earlier

ones

being

the Brass estimates.

The values

FIG.14.UNDER 5 MORTALITY CHANGE

190

180
<0

z

IfiC

2 170
o
o

s

160

c
150

140 L-L
1975

I

I

I

1979

1

1

I

1983
YEAR

I

I

I

1987

,

I

1991

FIG.15.INFANT MORTALITY CHANGE

130

(0

120

z
z
2
o 110

s
AC

100
••

90
1975

■X—1

I

1979

1

1

1

1983

1987

YEAR

I

I

I

1991

(55)
derived

from the Preceding Birth Technique are

compatible

with the Brass estimates and indicate a gradual declining
trend in both under five and infant mortality rates.

In

thirteen years (from 1976.1 to 1989.2), the under five
mortality has declined by 16.7% and infant mortality by

21.5%.
It should, however, be cautioned that the estimates
for childhood mortality should be considered as underestimates be

cause of some evidence of under-reporting of

females deaths.

Ei.

COMPARISON WITH AVAILABLE DATA

There appears to be no precise quantification of
childhood mortality in urban slums of India.

However, it

would be useful to compare the results of this survey with

the projected national figures and available data from

micro-studies in a similar environment (Tables 18 and 19) .
In this context, it is once again pointed out that the pre

1980 estimates of this survey may not necessarily represent
those of urban slums since many JJ Bastees were built after

this period.

However, they do provide vital information on

mortality experience of this population.

(56)

TABLE

8: Comparison of Infant Mortality Rate (ql)

Year

1989.2

1986.9

1984.6

1982

1979.2

1976.1

1. Present
survey
SRS
India
2.
Urban (7)

100*

104

110

114

120

120

62

62

66.1

65.2

72.2

80

3. SRS India
rural (7)

102

104.6

113.3

113.7

129.7

139

4. SRS India
combined (7)

94

96.4

104

104.8

120

120

5. Delhi (8)

37.9

42.2

44.4

47.6

41.7

55.2

6. Urban
Slum (9)

123

7. Delhi
slum (10)

112

* Estimates based on preceding birth technique.

The urban slum ql estimates from the present survey
are comparable with Sample Registration System (SRS) fig-

ures for rural India (7) .

The SRS urban estimates are only

two thirds of the urban slums.

Unfortunately, data from

the SRS is not published for Delhi since the sample size is
too low.

The infant mortality in slums is about 2 1/2

times the projected figures for Delhi by the Bureau of
Economics and Statistics (8) .

(57)
TABLE 19 j_ Comparison of Under Five Mortality Rate (q5)

Year

1989.2

1986.9

1984.6

1982

1979.2

1976.1

1. Present
survey

146*

154

167

175

187

186

2. SRS India
urban (7)

85.3

92.5

96.0

92.4

102.7

128.5

3. SRS India
rural (7)

146.7

156.3

171.7

165.8

198.6

215.9

139.5
4. SRS India
combined (7)

144.8

156.8

152.4

182.2

199.4

* Estimates based on preceding birth technique.

Infant mortality rate figures for urban slums have

been quoted in various reports (9,10). Comparisons with

these figures are difficult since the methodology has not
been described and these appear to be based on micro-stud-

ies.

In slums,

estimates have ranged from 26 to 208,

averaging 86 in one study.

In another study in slums of

Delhi, ql averaged 112 but went upto 200 in one slum (10).
On comparing the under five mortality rate, essentially similar conclusions emerge (Table 19)•

The SRS esti-

mates for rural India are roughly comparable whereas those

for urban areas are considerably lower (about 60%) •

Unfor-

tunately, similar estimates, even from micro-studies are

scare for q5 in urban slums.

(58)

From the above comparison it can be seen that child­
hood mortality in urban slums is considerably higher than

the projected estimates for that area.

The slum areas are

not included for estimating the vital events in reliable
surveys like the Sample Registration System.

These areas

comprise a significant proportion of the urban population.

It is, therefore, obvious that the projected figures are
underestimates. A further implication of this finding is
the need to focus urgent remedial action in these ’’hidden

areas” in the context of Child Survival Programmes.

(59)

PART B 1 KAP SURVEY

A

Knowledge, Attitudes and Practices of Mothers with
Regard to Pregnancy, Childbirth and Neonatal Care

Six hundred and thirty two mothers were interviewed
from 150 JJ clusters on various aspects related to KAP with
regard to pregnancy, childbirth and neonatal care.

The

pertinent details are summari zed be1ow.

1. Age and Parity

Two thirds (66%) of the mothers were aged between 2030 years while the rest were equally distributed between

15-19 years and 31-40 years.

Almost half (48%) the women

had a parity of more than 3 and 18% were primiparous.
2. Birth Spacing

Almost 91% mothers considered that an

adequate birth

interval was essential for health of mother and child.
the mothers answering in the affirmative,

Of

63.2% felt a

birth interval of 1-3 years was adequate for maternal and
child’s health, while 36% felt that more than 3 years

interval between births was essential. A total of 59% women

stated their awareness of birth control methods and a
little over half (53%) felt that continued breast feeding
was an effective method of contraception.

(60)
3. Perinatal Outcome

Nearly all

(94%)

of the women had experienced an

uneventful delivery during the last childbirth.

Only 2.6%

of the multiparous women interviewed had experienced in­

trauterine death in the delivery preceding the last birth.
Of the 618 mothers who

live births, 4.6% had experienced

death of their child under 3 months of age - 55% dying in

the neonatal period and the rest between 1-3 months of age.
Only half the mothers were aware of their

child’s cause of

death.

4. Antenatal care
Nearly two thirds 62% women felt that antenatal care

was essential, but only 55% could get antenatal care during
their last pregnancy. Almost all (98%)

of these women

availed of hospital based antenatal services.
TABLE 20 J. Distribution of Women by Antenatal Visit

Visit No.

% women

One

13.6

Two

31.7

Three

27.4

Four or more

27.3

0
CD
Z
O

z
D
O
CD
Q_
CQ
CD

CD
CO --

B

XI

m

(Z)

o

(Z)

CD

O
ID

o

z

o

O
DO I
CD -k
CD DO
2- GO CD

o
ro ®
co

z

0
O

(61)

Only a third of mothers had recorded a minimum of four
antenatal visits (Table £0)

which is considered adequate

for satisfactory perinatal outcome (11).

In a recently

published report by the Indian Council of Medical Research

(12)

it was reported that in urban slums in three major

metropolises of the country only 36% women received four or

more antenatal checks during their pregnancy and 34% wanted
these antenatal checks during an illness.

Nearly half (45%)

of the women did not receive any

antenatal care.

The reasons for these are shown in Table

21 and Fig.16.

Two-thirds (67.1%) of the women were igno-

rant of the need for antenatal care.
Table 21

Reasons for not Receiving Antenatal Care

Reasons

%

1. No knowledge

67.1

2. No time

6.9

3. No MCH facility in neighbourhood

12.5

4. No faith

2.3

5. Miscellaneous

11.2

(61a)
5. Nutrition During Pregnancy

We assessed the woman’s concept of dietary intake in preg­
nancy, which is presented in Table 22.

Table 22 1 Dietary Intake in Pregnancy

Quantum of dietary
intake

i Respondents
Awareness
Actual intake

1. Normal (Pre-pregnant)

41.1

33.5

2. More than normal

24.8

22.6

3. Less than normal

28.1

31.8

4. Uncertain

6.0

12.1

It is evident that only a quarter of the mothers
interviewed felt the need for an increased dietary intake

during pregnancy. The data with regard to actual intake
compared well with their existing knowledge.

This suggests

that awareness of dietary needs during pregnancy was poor

and economic constraints probably had a lesser influence

than maternal education and awareness on their actual
dietary intake. The ICMR study in urban slums also observed
that only 37.3% women were aware that there is need to

consume a larger dietary intake during pregnancy (12), an
observation similar to that of the present study.

(62)

Interestingly 93% of women believed that adolescent

girls must get adequate nutrition and almost a similar

number stated that they got sufficient food to eat
their adolescence. However,

during

what these women considered as

adequate intake is open to question in view of the above

observations on dietary intake during pregnancy.

6. Smoking and Alcoholism
About 12% of women stated that they smoked both before

and during pregnancy, while only 2% women admitted to

alcohol consumption during pregnancy.

However the study

did not attempt to quantify the extent of smoking or alcohoi intake.
Interestingly the recently published ICMR study (12)

observed that in the urban slums only 2% women smoked and
there was no alcohol consumption.

These differences might

have been due a much larger number of urban slum clusters

represented in the present study (150 clusters) compared to
the ICMR study (21 clusters).
7. Iron and Folic Acid Intake

Nearly half (55%) of the women were aware of the need

for iron and folic acid during pregnancy for prevention of
anemia, and an almost similar number(60%) were aware that
these tablets were available free of cost at MCH Centres.

(63)

When asked regarding consumption of iron and folic acid
tablets during their last pregnancy, 51% answered in the
affirmative.

About 20% women stated that they had consumed

some form of tonics during their pregnancy.

8. Tetanus Immunisation
Nearly three-fourths (78%) of the respondents were

aware that some injections were administered to pregnant
women.

Only two-thirds (63%) stated that they had received

tetanus immunization during pregnancy. Nearly half (52%)

women had received at least 2 doses of tetanus toxoid

during the antenatal period, 9% had received only one dose,
the rest either had not received the toxoid or were uncer­
tain about the doses received.

9.

Work and

Rest During Pregnancy

Of all respondents, about 25% stated that they were
normally doing heavy work, including lifting of weights and

about 26% stated that even during pregnancy they continued
to do heavy work/labor.
About 72% women felt that pregnant women should get

additional rest and about 76% of the respondents stated
that they got sufficient rest during their last pregnancy.

However,

the

study did not attempt to assess mother’s

(64)
knowledge as to what constituted adequate rest during
pregnancy.

10. CoMplications During Pregnancy
The respondents were asked to state if the selected

symptoms in pregnancy pose a risk for mother/fetus/both.
Table 23 summarizes the answers obtained.

TABLE 23 2, Awareness of Risk of Obstetrical Complications

Symptoms
Yes(%)

Risk present
No(%)

Uncertain(%)

1. Vaginal bleeding

80.8

8.2

11.0

2. Edema feet

56.0

30.8

13.2

3. Headache, dizziness,
visual blurring

45.2

40.9

13.9

4. Burning micturition

48.3

29.7

22.0

A quarter (24%) of women stated that they had suffered
from one or more of the above symptoms during their last
pregnancy.

When asked where they sought medical advice for

their problems, 26.7% stated that no help was sought. 35.2%

sought help from Government health facilities. 16.3% from
private

doctors

and

14.3%

from

Dais.

(65)
11. Place of Delivery

A vast majority (82.5%) of

women had their delivery

at home, 15.8% at Government hospitals/centres and the rest
availed of private medical facilities. In contrast the ICMR

study had observed that 43% of urban slum dwellers deliv-

ered their babies at the hospital (12). The reasons for
opting for the place of delivery are listed in Table 24.

TABLE 24

Reason for Choosing Place of Delivery

Place of Delivery
Home Delivery

Reasons
1. By choice

%

86.2

2. No means to reach
hospital

1.9

3. Apathy in government
hospitals

1.9

4. Others

9.9

Institutional

1. By choice

93.8

Delivery

2. Obstetrical problems

3.1

(n=98)

3. Non availability of Dai

3.1

(n=522)

In case of home deliveriest the persons who conducted

the births were Dai (TBA) 85.4%, relatives(6.1%), Nurse/ANM

(5.1%),doctor(2.4%) and Others 1.0%. The ICMR collaborative
f

study (12) had also revealed that 61% of the home deliver­

ies were conducted by TBAs, most of whom were untrained.

FIG.17. TYPE OF FIRST FEED

Ghutti

25.1
Breast Milk

17.1

Honey

19.3
Others

15.4
Formula Milk

15.9

Glucose Water
7.2

(66)

12. corO Car*
The instruments used to cut the cord as stated by the women
are listed in Table 25.-

Table

1_

instruments

for

Cord

Cutting

%

Instrument

1.

Blade

70.1

2.

Knife

2.0

3.

Scissor

17.5

4.

Others

1.1

5.

Not known

9.3

Nearly three fourths

(73%) of the women stated that

the cord cutting instrument was boiled for at least 10

minutes and 67% stated that the thread for tying the cord

was also boiled. In the only other large multicentric study
on KAP with regards to childbirth in urban slums conducted

by the ICMR (12) it was observed that while 97% women were
aware that the instrument for cord cutting was to be steri-

lized, only 75% actually sterilized the instrument.
13.

Neonatal Feeding

(a) First feed : The type of first feed offered to the baby
after

birth

is

depicted

in

Table

26

and

Fig.17 .

(67)
29 ± TVP» Qt Zlzit ZMfl

t

Type of feed

1.

Breast milk

17.1

2.

Ghutti

25.1

3.

Honey

19.3

4.

Formula milk

15.9

5.

Glucose water

7.2

6.

Others

15.4

Ghutti was the first feed in 17%, honey in 10% and

formula feeds in 12% of cases in the 1990 ICMR multicentric

study in urban slums (12).
to our

(b)

These findings were comparable

observations.

Colostrum : When asked the time at which breast feeding was
initiated, most women did not clearly remember and so the
data was not analysed further.

When asked regarding the feeding of colostrum,

42%

No
women stated that they had fed colostrum to their baby.
satisfactory reasons were elicited from those who had not

fed colostrum.

(68)

c)

Top Feeds» About 15% mothers were either exclusively or

partially top feeding their babies at the time of the
interview. Almost 90% of these mothers were diluting the

milk with water. Sixty per cent of the mothers were using a
bottle to feed their baby and the rest a cup and spoon.
Surprisingly almost half of these mothers had initiated top

feeding under one month of age. The reasons given by the

mothers for initiating top feeds are provided in Table 27.TABLE 27

Beasons

or Initiating Top Feeds

1

Reasons

1. Breast milk insufficient

50.0

2. Maternal illness

12.5

3. Doctor’s advice

6.3

4. Breast milk didn’t suit

6.3

24.9

5. Others

14. Immunization

The mother’s knowledge with regard to immunization

schedule and the diseases against which protection is
offered was abyssimally low. However,

49% mothers stated

that their children had received some form of immunization
upto

the

t ime

of

this

study.

(69)

B, Knowl+Qq?* Attitudesi Practices o£ Pai? with R?q?rfl 12

Perinatal and waonatal Cara
A total of 180

Dais were interviewed to assess their

KAP with regard to pregnancy,childbirth and neonatal care.

In 40% of the clusters selected for the survey, no Dais
were available.

The largest proportion of clusters without

any Dais were located in the West and South Zones.

1. Background Information
(a)

Age : Two-thirds (68%) of the Dais interviewed were between

30 to 50 years of age. Only 5% were under 30 years of age.
the rest being more than 50 years.

(b)

Caste Status: Forty-four per cent of Dais belonged to the

Scheduled Castes, 21% were from the upper castes (Thakurs
and Brahmins), 7% were non-Hindus and 18% were from other
Backward castes. These observations are contrary to popular

belief that most Dais are essentially from the Scheduled
Castes.

(c)

Training and Experience : Only 16% of the Dais had received
any formal training. Of the

remaining,

44% had acquired

the skills from relatives, 16% were self taught and 15% had
learnt it from other Dais. Fifty per cent of the Dais had a

traditional background of this occupation and a third (32%)
were into the trade for the first time.

(70)

Almost a third(28%) were in this profession for under

5 years and a similar number were in it for 6-10 years.
Thirty per cent had been in this profession for between 11-

20 years, the rest having more than 20 years experience.

(d)

some income from
Income and Workload: Most Dais could earn
Only 72 Dais responded to our
being in this profession.
query on their monthly income. About 43% earned Rs.150.00
27.7%
or less per month, 10% between Rs.150-300 per month,

between Rs.300-500 per month and the rest claimed to earn
more than Rs.500 per month. Eighty six Dais their earnings

The charges levied by them Dais

per delivery attended.

ranged from a few rupees to over Rs.250 per delivery. About
40% earned Rs.100.00 or less and an equal proportion earned

between Rs.101-250 per delivery attended.
that they

were paid

in

kind

About 6% stated

by the family for their

services.
Most Dais conducted less than 5 deliveries per month.

The more popular Dais conducted more than 10 deliveries per
month and they constituted about 10% of the Dais inter—

viewed.
2. Confirmation of Pregnancy
Nearly half (54%) of Dais stated that they could confirm a
pregnancy between 1 to 3 months.

Early confirmation of

(71)

pregnancy was in most cases possibly an intelligent guess

from missed menstrual cycle or symptoms of morning sickness. Only 36% Dais were aware of the method of calculating
the 'due date* of delivery from the mother’s last menstrual

period. As many as 20% had no idea of estimating the

due

dates of delivery’.
3. Nutrition in Pregnancy
Nearly three fourths (76%) of the Dais thought nutriRoughly half (53%) of

tion influenced pregnancy outcome.

the Dais felt that women should eat less during pregnancy

so that smaller babies are produced leading to easy childbirth. Thirty-three per cent,

should eat more food

however,

during pregnancy.

felt that they

The rest felt that

a normal dietary intake during pregnancy was good enough or

they

had

no

idea

of what the nutritional intake in

pregnancy should be.

4.

Anemia
A majority (84%) of the Dais felt that women in our

country were anemic and that this was aggravated during

pregnancy.

Sixty per cent felt that iron and folic acid

should be taken during pregnancy and an almost similar

proportion (63%) were aware that these tablets were avail­

able

free

of

cost

at

MCH

centres.

(72)

cliniTable 28 summarizes the Dais ability to assess
The Dais however, had little
cal features of anemia.

concept of iron rich foods which mothers could consume

during pregnancy.
TABLE

1_

Anemia

Pais

bx

Detsct ion
% awareness

Symptoms
*

Pallor of eyes, face and
nails

58.0

*

Edema feet

3.0

*

Dizziness

4.0

*

Tiredness/breathlessness

2.0

*

No knowledge

33.0

5.

Rest and Work in Pregnancy

Table 29 gives the attitudes of Dais with regard to
rest and work during pregnancy.
Table 29

:

Attitude to Rest and Work

in Pregnancy

No(%)

Don't know(%)

Attitude

Yes(%)

a. Needs extra rest

71.1

7.8

21.1

b. Avoid strenuous work,
but continue household
chores

70.5

10.7

18.8

c. Rest only during last
trimester

66.6

16.7

16.7

(73)

Roughly 60%

Dais felt that women in JJ clusters got

enough rest during pregnancy . Those answering to the connot getting sufficient
trary gave a variety of reasons for
excessive houserest- poverty and therefore need to work,
hold chores, family problems,etc.
6. Tetanus Immunisation

About 80% of the Dais had heard of the disease called

tetanus. Forty-one per cent said that they had seen cases
of neonatal tetanus in the community.

About 73% Dais were

aware that tetanus immunization was necessary for pregnant
mothers but only 10% were aware that tetanus immunization

to the mother protects her offspring.

In fact 82% thought

that tetanus immunization in pregnancy protected the mother

against this infection.

Only 17% Dais were aware that at

least two doses of tetanus immunization were required

dur ing pregnancy. Only a third of Dais (31%) were aware
that tetanus toxoid had to be administered between 6-9

months of pregnancy.

Most were unaware of the actual

number of doses that a woman must receive and at what
interval. In fact, 9% Dais reported that a baby delivered

by them had developed neonatal tetanus.

(74)

7.

Advise on Antenatal Visits
Almost three-fourths (77%) of the Dais were advising

mothers to attend antenatal clinics at MCH centres, In fact

50% of the Dais claimed that they accompanied the mothers
to the MCH centers on one or more visits.

a

About 25% Dais felt that getting to the MCH Centre was
problem and/or the staff at the health centre were apa-

thetic and failed to provide satisfactory advice.

8.

High Risk Factors

When asked to state conditions likely to adversely influ-

ence pregnancy outcome,14% stated fetal malpresentation as
an important risk factor.

Five per cent stated that de­

creased fetal movements was a risk factorr while anemia and

vaginal bleeding were stated to be risk factors only 3% of

the Dais.

Their knowledge about other risk factors was

poor. Their awareness with regard to situations precluding

safe domiciliary birth is summarized in Table 30.
DELIVERY PRACTICES

9. Attendance During Childbirth
Eighty-nine per cent Dais stated that she was called

at the onset of labor pains and 83% stated that they were
present throughout labor. Only 10% were called to cut the
cord and deliver the placenta.

(75)
Precluding Safe Domiciliary
TABLE 30 i Maternal Conditions
Birth

Maternal conditions

Risk for domiciliary birth
yes
no
Don’t Know

1. Short stature

64.4

30.1

5.5

2. Weight < 40 Kg

62.2

32.3

5.5

3 . Age < 18 years

61.1

33.4

5.5

4 . Age >= 35 years

55.5

49.0

5.5

5. Parity > 4

58.3

36.2

5.5

6. Fetal malpresentation

86.1

8.4

5.5

7. Previous Cesarian birth

87.7

6.8

5.5

10. Vaginal Examination
A total of 69% Dais stated that they carried out
Only 44%
between 1-3 internal examinations during labor.
When
washed their hands before the examination procedure.
asked to state reasons

for internal examination,

small number of Dais responded (Table 31).

TABLE 31 1 Reasons Stated by Daig for
Internal Examination
Maternal condition

1. Fetal malpresentation
2. No labor

pains

%

17.0

11.0

3. Excessive bleeding

11.0

4. To assess cervical dilatation

6.0

only a

(76)

11. Delivery posture

The delivery posture preferred by the Dais is summarized in
Table 32.

TABLE 32 j. Preference for Delivery Posture
% preference

Posture
1.

Squatting

14.4

2.

Lying only

25.5

3.

Either

squatting

or

52.2

lying

7.9

4. No reply

12. Cleaning and Washing
(a)

Hands : A majority

(91%) of Dais stated that they washed

their hands before conducting a delivery. For this prac­

tice, 89% used soap and 6% used only water for hand wash­
ing. In contrast the ICMR study (12) observed that only 55%
women washed their hands before attending to a delivery.
(b)

Mother :

The mothers perineum was cleaned before delivery

by 84% of the Dais.

While 62% used a wet cloth to clean

the area, only 10% used soap and water for cleaning of
perineum.
(c)

House s
birth

The Bastee was said to be cleaned before a childby

87%

of

Dais .

(77)

13. Labor Managaaant

When asked as to what the Dais considered as prolonged
labor, the answers ranged from one hour to more than 24
hours. This clearly indicates their lack of knowledge with
regard to what constitutes prolonged labor. Nearly half

(48%) of the Dais took the help of a nurse or local medical

practitioner to give some intramuscular injections to the

pregnant women to increase labor pains( a risky practice
which should be strongly discouraged).

14. Delivery of Placenta
The replies

obtained with regard to their practices on the

delivery of the placenta are listed in Table 33..
TABLE 33

Practices for Delivery of the Placenta

Practice

%

a. Nothing (allow spontaneous
delivery of placenta)

13.3

b. Internal examination

13.8

c. Pull the cord

17.7

d. Abdominal massage

46.1

e. No response

9.1

Sixty one per cent Dais stated that in the event of

retained placenta, the woman was referred to the hospital.

(78)

However,

the stated duration beyond which

referral

than 3 hours).
was needed was highly variable (0.5 to more
Three per cent Dais had never faced the situation of re-

tained placenta.
15. Maternal Daath
Eight per cent Dais reported
the cases delivered by them.

a maternal death amongst

However, most could not give

the possible reasons for this death. Some of the causes
alluded to included retained placenta, twin delivery, still
birth and anemia in mother.

16. Hospital Referral in Labor
Twenty-two per cent Dais stated that they had an
occasion to refer women in labor to hospital.

Only a

quarter stated reasons for doing so, which included prolonged labor,

maternal fatigue,

excessive bleeding and

malpresentations. These statements do not, however, reflect
whether the referrals made were early(at an optimal time)

or what was the condition of the mother at the time of
referral. Most Dais (85%) accompanied the referred mothers

to the hospital and were satisfied with regards the treatment provided.

T

0

0
CD
CD

I

o

03

1
I

O

CD

hO o
CD

O

o
Z)

TJ
O
03

00

o
o
u
"U

d
CD

o

CD CD
Z

I
I

I
I
I
1
I
I
I

o

o
DO
O Z

K) O

D
CO

CD

o
z

CD

(79)

yonatal Cara

(a)

Care

Cord

17 .

Instrument; Eighty-four per cent Dais used a blade for cut-

ting the cord, 10% used scissors and the rest a variety of
instruments.

(b)

Boiling: Ninety per cent Dais stated that the instrument

for cord cutting was boiled for at least 10 minutes.

(c) Cord applications: A variety of substances were applied on
the cord by the Dais. The responses obtained are listed in

Table 34 and Fig.18.
Table

Cord

34

Substance

%

Substance

Applications
%

1. Nothing

20.0

5. Dettol

8.0

2. Medications

21.0

6. Powder

8.0

3. Haldi

14.0

7. Ash

6.0

4. Ghee

11.0

8. Others

12.0

In contrast to our observations an earlier study by
the ICMR (12) observed that 31% women applied nothing to

the cord while 62% applied medicated creams. This differ-

ence might be explained by the substantially larger number
of slum women delivering in a hospital in the ICMR study as
compared to the present one, where only 15-20% deliveries

were institutional.

(80)

18. Clearin? of Mouth at Birth
A vast majority (93%) of Dais felt that clearing of

the mouth of the baby at birth was necessary and in fact

carried out this practice routinely.

19. Asphyxia Management
About 80% Dais were able to identify an asphyxiated

baby by virtue of it being apneic or blue at birth. The
Dais were then asked regarding their practices when faced
with an asphyxiated

baby at birth. Their responses are

listed in Table 35.

TABLE

35

Asphyxia

Intervention

Management

%

1. Pour warm water

38.0

2. Blow into baby's mouth

58.0

3. Others(camphor,onion,baby
hung upside down,massage cord)

4.0

4. No response

8.0

20. Assessing size (Weight) at Birth
Twenty per cent Dais stated that they guessed the possible

birth weight. Four per cent responded that they could
manage to weigh the baby(none possessed a weighing scale),

(81)
the rest made no assessment of baby’s size at birth.
21. Bathing the Baby at Birth

Ninety-two per cent of Dais stated that they bathed
the baby after birth.

( We believe that this practice is

potentially harmful as the baby runs the risk of becoming

hypothermic).

The prevalence of this potentially harmful practice in
urban slums was also observed in the ICMR study (12) where
97% of babies were bathed before 24 hours of age (majority

within 8 hours of birth).
2? .

Feeding

(a)

First Feed: Only 6% of Dais allowed breast milk as the
first feed. The other responses are listed in Table 36.

Table 36 j. Type of First Feed

Type of Feed

%

1. Breast milk

6.0

2. Ghutti

41.1

3. Honey

36.6

4• Glucose water

7.8

5. Top milk

1.6

6. Others

6.9

I;
(82)

(b)

Colostrum: Only 36% of Dais stated that they advised the

(c)

colostrum

of

feeding

to

the

baby.

Initiation of Breast Feeds: Only half the Dais responded to

our query regarding when breast feed was initiated after

birth (Table 37).

TABLE 37

Initiation of Breast Feeds

Time to First Feed

%

* At birth

5.5

* On first

day

8.3

* On third day

22.2

* On second

*

(d)

8.3

day

55.7

Non responders

Feeding Methods : The Dais knowledge with regards to feeding methods in babies not able to suckle at the breast was

also assessed (Table 38).

More than half the Dais resorted

to cotton or spoon feeding.
TABLE 38 1 Feeding Methods Propagated

Feeding Method
1.
2.
3.
4.
5.

Cotton
Spoon
Bottle
Dropper
Non responders

%

33.3
28.8
12.2
4.4
21.3

(83)

(e)

Lactagogues: The Dais attitudes with regard to lactagogues
was evaluated but the responses were varied and interesting

(Table 19).

TABLE 39

Attitude with Regard to Lactagogues

Lactagogues

* Milk and Ghee
*

Green leafy vegetables
and fruit juices

%

30.0

33.0

* Meat,eggs,fish

11.0

* Pulses,Cereals

14.0

* "Ajvain'and 'Jeera1

14.0

23. Care of LBW Babies
We were interested in evaluating how the Dais viewed
Low Birth Weight (LBW) babies and the care they considered
appropriate.

Less than 10% Dais were conversant with the

concept of special care for LBW babies.The responses of
these Dais alone is presented in Table 40.
Hardly any Dai gave advice on maintenance of baby

temperature which is essential to prevent hypothermia in

this group of babies. In the ICMR study (12) too less than

1%

health functionaries advised mothers on clothing and

environment for their babies.

(84)

TABLE 40 1 Care of

iBW Babies

Advice

%

1. Adequate milk feeding

7.0

2. Ghutti/Tonic

3.0

3. Baby massage

3.0

4. Seek medical help

7.0

24. High Risk Referral

Almost half (46%) the Dais stated that they had expe­

rienced situations where the neonate had to be referred for
medical help. However, they were unable to clearly state

the risk situations that demanded the referral.
25. Infant Immunization

Three fourths (75%) Dais advised mothers on their infant’s

immunization. However, their awareness with regards to

vaccine preventable diseases and what immunizations pre­

vented

them

was

very

dismal

TABLE 41 j. Awareness of Infant Immunization

Vaccines

* BCG
* Tetanus
* Poliomyelitis
* Pertussis

%

10.0
11.0
8.0
3.0

(Table

4i) .

(85)

26. Family Waifare Services
About 46% Dais were aware of birth control methods.

Only 37% stated that they accompanied women to MCH centres
for Copper-T insertion or tubectomy. About 18% Dais admitted to carrying out abortions in the community.
COMMENTS

The awareness and utilization of MCH services by
mothers in the J J Bastees was poor.

Their awareness and

practices with regards to antenatal and neonatal care was
largely a reflection of prevalent socio-cultural attitudes.
Most mothers chose to deliver at home and locally available

Dais (most of whom were untrained) were the prime caretak­

ers during pregnancy, childbirth and possibly also of their

neonates.

However, the knowledge, attitudes and practices

of these birth attendants were no better than that of the
mother’s.

Their knowledge with regard to nutrition in

pregnancy, rest and maternal tetanus immunization were far
from satisfactory.

Some of their practices with regard to

labor management, its augmentation (particularly the use of
oxytocin), delivery of placenta and neonatal care during

the first few hours of life were potentially harmful.
'high risk’ concept,

referral

was

The

its identification and subsequent

almost

non-existent.

(86)
This state of affairs continues to prevail even several years after the Report of Task Force on Minimum Perina-

tai Care, Government of India (1982)

(11),

Task Force

Report on Maternal Mortality and Morbidity, Government of

India (1987)

(13) and the adoption of the Safe Motherhood

Initiative (Nairobi, 1987)(13).
There is need to launch a campaign to increase aware-

ness of women in the reproductive age group on perinatal
and neonatal care practices.

Simultaneously efforts to

train birth attendants (even new willing entrants) has to

be dovetailed with other efforts to improve perinatal and
neonatal outcome. Several workers have shown that "Delivery

huts or booths’ in the community where mothers can have
safe childbirths is a workable proposition.

These efforts

can only succeed if adequate linkages are established
between the community and existing health facilities in

urban areas.

The creation of a cadre of "Link Worker*

(or

Bastee Sewika) will help fill this void in the referral

chain.
The urgent need of such efforts is strengthened by the

observation that barring 3-4 States, in the rest of the
country urban neonatal and perinatal mortalities have shown

no decline over the last decade, unlike the rural region
which has shown a 4% annual decline in neonatal mortality

rates (7).

SUMMARY

(87)

In this survey, 150 Jhuggi Jhompri clusters of Delhi
were studied with respect to early childhood mortality and
knowledge, attitude

and practices (KAP) of mothers and Dais

for perinatal and neonatal care.

A total of 22,181 house-

holds were visited for this purpose.

The current state of civic amenities in the surveyed

population was deplorable.

There was one safe water point

for 102 households (desired norm is 1 in 20) and only 53% of

the Bastees had Public Utilities.

The most recent estimates of childhood mortality ob­
tained by preceding birth technique relate to the year

1989.2.

These indicate an infant mortality rate of 100 and

under five mortality rate of 146. With Brass technique there

is evidence of secular decline in both infant and under five
mortality.

The infant mortality rate is marginally higher

for boys whereas the under five mortality rate is higher for

girls.

The latter is

probably a reflection of well known

preference for the male child.

The estimates from the present survey are comparable
with Sample Registration System figures from rural India and

substantially higher than urban India.

The infant mortality

in slums is about two and a half times the projected figures
for Delhi by the Bureau of Economics and Statistics.

(88)
It is apparent that childhood mortality in urban slums

is considerably higher than the projected estimates for that

area.

The slum areas are not included for estimating the

vital events in reliable surveys like the Sample Registration System.

These areas comprise a significant proportion

of urban population.

It is, therefore, obvious that the

projected figures are underestimates.

A further implication

of this finding is the need to focus urgent remedial action
in these "hidden areas" in the context of Child Survival
Programmes.

The KAP study of mothers and Dai’s in JJ Bastees with
regard to perinatal and neonatal care practices underscores
the glaring inadequacies in the MCH services to this popula­

tion.

Almost half (45%) the mothers did not avail of antenatal services and only about 16% women had received the
optimal four ante-natal checks during pregnancy.

Knowledge

with regard to nutrition. rest and complications during

pregnancy was poor.

As high as 12% of women continued to

smoke even during pregnancy.

However, the satisfying find­

ings were that 63% mothers had been immunized with tetanus

toxoid and 51% were receiving iron and folic acid tablets
during pregnancy.

It is strongly felt that intense health

(89)
educational

efforts

would

improve

the care of

pregnant

mothers and their offsprings.

Eighty-two per cent women chose to deliver at home and
most deliveries were conducted by Dais (most of whom were
untrained).

The attitude and practices of Dais with regard

to pregnancy and childbirth were far from satisfactory.

In

fact some of the practices relating to labor management and
neonatal care were potentially harmful and need to be ac-

tively discouraged.

The 'high risk' concept and referral

were unfamiliar to most Dais, but what may offer a glimmer

of hope is their willingness to be trained if given an
opportunity.

RECOMMENDATIONS

(90)

On the basis of this survey, the following recommendations

emerge :

1.

The currently projected urban childhood mortality figures

should be considered underestimates.

The urbanites residing in

slums must be included for estimating the realistic vital sta­

tistics .

2.

The vital statistics for urban areas should have a separate

category for urban poor.

3.

In formulating Child Survival Programmes for urban areas,

attention should be primarily directed to slum dwellers.

4.

The

high

chiIdhood

mortality

despite

widespread

availability of health services is a cause for concern.

Efforts

should be directed urgently to ascertain the specific causes for

formulating effective remedies.

5.

The areas which require immediate attention in this context

include :

(i)

Improvement

in the availability and quality of

civic

amenities to match the recommended norms.
(ii) Intense and sustained Maternal and Child

Health

(91)

related educational inputs to improve the knowledge,

attitude and practices in this context.
(iii) Traditional birth attendants (Dais) must be identi-

fied and trained to ensure safe motherhood.
link

worker

(Bastee

of

a community level

akin

to community health

volunteer

in

rural

areas merits consideration.

This will

facili-

tate

delivery

(iv) Creation

Sewika)

poor.

of primary health care for

the

the

urban

REFERENCES

(92)

1.

Bhatnagar S, Dosajh U, Kapoor SD. Availability and utiliza­
tion of health care facilities in urban slums of Delhi. Part

III. Health Popn-Perspect Issues 1988, 11:162-170

2.

Bhatnagar S. Health services development in urban areas.
Health Popn-Perspect Issues 1989, 12:71-93.

3.

Bisharat L,David P, Hill A. Measuring Childhood mortality: A
Handbook for Simple Surveys.

Amman, Jordan. UNICEF Regional

Office for the Middle East and North Africa, 1989.
4.

Report on Situational Analysis in JJ Clusters, Delhi Urban
Basic Services (UBS), 1990.

5.

Data Management Software for Rapid Mortality Surveys.
United Nations International Children’s Emergency Fund,

Regional Office for the Middle East and North Africa, 1990.

6.

Step by Step Guide to the Estimation of Child Mortality.

United Nations Department of International Economic and
Social Affairs, Population Studies No.107, New York, 1990.

7.

Puri RK,Sachdev HPS.Secular Trends and Determinants of Under

Five Mortality Components in India - Implications for Child
Survival Strategies.

Report Submitted to the Ministry of

Health and Family Welfare (Under USAID Child Survival Pro­
grammes), Government of India, September 1991.

(93)
8.

Delhi Statistical Hand Book.

Bureau of Economics and Sta-

tistics, Delhi Administration, Delhi, 1990.

9.

A Programme tor Children and Women in India.

Plan of Opera­

tion 1991-1995 Government of India and UNICEF, P 61.

10.

The

Invisible Child.

United Nations

International

Children’s Emergency Fund, 1990, pp 3-11.
11.

Report of the Task Force on Minimum Perinatal Care.

Minis-

try of Health and Family Welfare,Government of India, 1982.
12.

A National Collaborative Study of Identification of High
Risk Families, Mothers and Outcome of their Offsprings with
Particular Reference to the Problem of Maternal Nutrition,
Low Birth Weight, Perinatal and Infant Morbidity and Mortal-

ity in Rural and Urban Slum Communities.

Indian Council of

Medical Research, New Delhi, 1990.
13.

Report of Task Force on Maternal Mortality and Morbidity,
Ministry of Health and Family Welfare, Government of India,

1987.

Position: 1036 (5 views)