368.pdf
Media
- extracted text
-
CPHE
J1
■
THE EVOLUTION
OF the Deliver
OF pPlMApoy
HEALTH CARF
I5
,
■n.
RIR'AN (JOHN D}
.
■
’L-
*
■
.
tSTV-'
.
’*
'
...
>
/•
CPH.E
f
£.
o
L
\
\
2, < ■
I-'A.rpALOrE
’IMMC^ITY HEALTH CEM
THE EVALUATION OF THE DELIVERY
r4
OF PRIMARY HEALTH CARE
e
The development of a methodology suitable for use ,
in Primary Health Care Units in developing'countries.
JOHN D. MARTIN)
Submitted in part fulfilment of the requirements for
I
the degree of Master of Science in Social Medicine at
the University of London.
■
<
1
I
u
1977
•
r
INDEX
Page
TITLE
SUMMARY
i
ii
INTRODUCTION
1
7
8
12
13
27
REVIEW OF THE LITERATURE
METHODOLOGY
■•1
(1)
Development of a Model
13 - 23
(2)
Adapting to Constraints
24
RESULTS AND DISCUSSION
28 - 55
CONCLUSION
56
REFERENCES
ACKNOWLEDGEMENTS
APPENDIX
Definition of Functions
Questionnaire of local population
English
Kannada
Quality of Care questionnaire
Assessment of efficiency record form
List of Mallur equipment
Maps
k
27
.i
SUMMARY
Evaluation of the delivery of health care has
been recognised as a means of improving its effectiveness
and efficiency by providing information which facilitates
improved decision making for management and planning.
The
need for such improvement is particularly urgent in
developing countries where the resources available for
the health care of the population are severely limited.
The study set out to develop and test methods
suitable for the evaluation of the delivery of health
care in Primary Health Care Units (PIICUs) in developing
countries, recognising that such evaluations should be
carried out routinely -in order to provide relevant
Since the day to day management
up to date information.
is the responsibility of local PHCU staff it was envisaged
r
that evaluation should also be the responsibility of local
staff, especially the medical officer.
It was necessary to place three important
constraints on the techniques of evaluation to be developed
so
that they would be appropriate to the circumstances
described above.
These were as follows:The methods of evaluation must be simple
(a)
(b)
The time required must be short
(c )
The resources required in terms of money,
manpower and equipment must be small.
The many varied processes involved in health
care delivery were described by .means of a conceptual
model whose components formed the framework for subsequent
analysis of information obtained in the evaluation study.
In order to obtain the information necessary
for evaluation the following techniques of data collection
■were utilised
/ •
(a)
A questionnaire survey of the local population
(b)
Examination of existing PHCU records and reports
(c)
An observational study of the activities of the
PHCU staff in order to obtain a measure of
the efficiency of health care delivery
(d)
A questionnaire study of a selected group of
recent patients in order to obtain a measure
of quality of health care activity.
The methods -were tested in three rnral PIICUs in
the State of Karmataka, South India.
The results of the study demonstrated that the
methods of evaluation,
despite necessary resource constraints,
were capable of generating information which could be used
for the improvement of management and planning of health
care delivery.
However, the merits of the study of quality
of care were not sufficiently proven and further testing of
this method was considered necessary.
IMKUDUC1 llhM
in the modern world it is generally accepted as a basic human
However,
sort should be available to everyone.
that
health
care
of
some
right
standard of care approachingoffer all her citizens a
no country can hope to
The reason
knowledge has the potential to offer.
that which modern medical
It
is Simple; insufficient resources, particularly financial resources.
■ improvement in health care
follows that in such circumstances potential for
of resources in an effective and
delivery lies in the ability to make use
efficient way (16).
is particularly
In developing countries the limitation of resources :
vital necessity for the best possible use to be made
acute and thus there is a
Since the vast majority of people in these
of the little that is available.
--1 countless small and often isolated rural
countries depends for
lor health
neaxui care
on« on
^n councxess
then it is in these where
where improved
improved health
health care
care delivery can produce
clinics
Improvement requires good management which in turn
the greatest benefit,
information about all aspects of health care delivery.
requires comprehensive
to develop and test methods for obtaining
The aim of this study has been
in the circumstances common to the
information which can be used routinely
majority of rural clinics,
developing countries is immense,
The problem of health care in
’ > also form the majority of the world’s
The majority of the population, who
of extreme poverty. (See Table 1).
population, live in conditions c.
TABLE 1
DISTRIBUTION Or WORLD POPULATION BY GROSS NATIONAL PRODUCT
PER CAPITA 1974
($)
f
Number of
Countries
Population
GNP
GNP per capita
$ millions'
%
(000)
%
5.000 or more
2,400,050
45.25
384,450
9.86
14
3,000 - 4,990
931,120
17.55
245,386
6.29
11
875,920
16.51
368,361
9.45
10
2,000 - 2,990
351,700
6.63
251,820
6.46
20
1,000 - 1,990
990
231,527
4.36
302,909
7.77
20
500 ~
490
330,310
6.23
1,045,412
26.81
24
300
290
34,940
0.66
140,450
3.60
11
200 -
190
140,730
2.65
1,064,558
27.30
24
100
90
8,200
0.16
95,590
2.45
11
50
SOURCE:
Health Planning, Management and Training Needs in Developing
Countries (1976)
capita
GRIEFITHS, A.
68% of people live’in countries .with a GNP under’ $1,000 per
These
and 60% in countries with a GNP less than $500 pci capita.
same people are
those who face the greatest health-problems resulting
TABLE 2
HEALTH STATGS INDICATORS BY GROSS HATIONAL PRODUCT
PER CAPITA 1972 '
Country
25 selected countries of over 5 million population
GNP per
?apita(S)
r
J
Life
Crude birth Crude deatk Infant
Mortality Expectancy
rate
rate
Upper Volta
70
48.5
24.9
180
39.0
Ethiopia
80
49.5
23.8
162
40.0
Indonesia
90
44.8
18.9
125
45.4
India
110
41.1
16.3
139
49.2
Tanzania
120
50.1
23.4
122
44.5
Nigeria
130
49.3
22.7
150-175
41.0
Uganda
150
46.9
.15.7
160
50.0
Viet Nam Rep.
170
41.8
23.6
100
40.5
Bolivia
200
43.7
18.0
60
46.7
Thailand
220
43.7
.10.4
23
58.6
Egypt
Ghana
240
37.8
15.0
120
50.7
300
48.8
21.9
156
43.5
Syria
320
46.9
14.4
24
53.8
Ecuador
360
41.8
9.5
87
59.6
Turkey
370
39.4
12.7
3 53
56.4
Colombia
400
40.6
8.8
81
60.9
Algeria
430
49.4
16.6
86
51.5
Iran
490
45.3
15.6
160
51.0
Peru
520
41.0
11.9
67
55.7
Brazil
530
‘ 37.1
8.8
110
61.4
Mexico
750
42.0
8.6
63
63.2
Portugal
780
18.4
10.1
50
68.0
Chile
800
25.9
8.1
71
64.3
Venezuela
1,240
36.1
7.0
52
64.7
U.S.A.
5,590
16.2
9.4
19
71.3
SOURCE:
As in Table 1
-2-
The close correlation vetween GNP per capita and expectation of
life is very clear.
For example the Life Expectancy of 71.3 years in the
richest nation, USA, is almost twice as high as that in Upper Volta, the
poorest nation, with a Life Expectancy of only 39.0 years.
Further characteristics typical of developing countries are
illustrated in Table 3 below.
TABLE 3
CHARACTERISTICS TYPICAL OF UNDERDE\;ETjOPED countries
1
Low Gross National Product per head
2
High Birth Rate
3
High death rate
4
High population growth rate
5
Low expectation of life
6
High infant mortality rate
7
High proportion of deaths from communicable and
parasitic diseases
8
High proportion of the population aged 0-14 years old
9
Low calorie and protein intake per. head
10
Low proportion of protein intake from animal sources
11
High proportion of rural population
As in Table 1
SOURCE:
With such meagre resources available and so many important
potential drains on them e.g. industry, agriculture, education etc.,
it is obvious that expenditure on health care facilities must, of
necessity, be extremely small.
(see Table 4)
(11, 18).
TABLE 4
MINIMUM ESTIMATES* OF HEALTH EXPENDITURE IN DEVELOPING
COUNTRIES 1974
SOURCE:
As in Table 1
Assumed
GNP
Hea1th
($
millions) Expen
diture
% gnp
Population
Number
(000)
of
Countries
Income
Category
($)
Estimated Health
Expenditure
per
Total
$ million Capita S
1,000 - 1,990
20
251,820
351,700
2.0
7,034
27.9
500 -
990
20
302,909
231,527
1.0
2,315
7.6
300
490
25
1,045,412
330,310
• 1.0
3,303
3.1
200
290
11
140,450
34,940
1.0
349
2.4
100
190
24
1,064,558
140,730
0.5
703
O.G
50
90
11
95,599
8,260
0.5
41
074
111
2,900,748
1,097,467
1.25
13,745
4.7
TOTAL
-O-
The estimated expenditures in columns 6 and 7 are based on the assumed health
% GNP in column 5 and the 1974 GNPs given in column 4.
expenditure
NB the
estimates arc intended as minimum orders of magnitude.
SOURCE:
As in Table 1
These gross figures, however, do not properly reflect the
inequalities in the distribution of resources.
For example,
many countries committed to Western style training for their doctors
have constructed large medical college hospitals in their big cities
only to find that the costs required to run and maintain them have
swallowed up a large portion of the annual national health budget
(30
In addition, the services offered by such hospitals
31)
are only accessible to a tiny proportion of the population, in
particular the urban population, since in most developing countries
only 20% approximately of the total population lives in urban areas
and the remaining 80% in the rural areas.
(10,21,24,39,40,45).
To divide resources in this way is both morally and
economically wrong.
In developing countries with highly labour
intensive industries, the good health of the labour force is of
great importance in raising the potential for increased productivity.
For example, in Haiti a yaws eradication campaign was estimated to
have returned 100,000 previously incapacitated workers to their jobs.
Similarly, an antimalaria programme in the Phillipines
14 ).
(
• reduced daily absenteeism due to malaria from 35% to apprxoraately 4%.
( 14
).
From the above it is clear that two major problems face the
organisation of health care in developing countries, i. e.
1.
Severely limited resources,
2. Maldistribution of these resources.
The likelihood of (1) being solved in the foreseeable future
is highly unlikely.
(See Table 5)
TABLE 5
PROJ ECTIONS OF EXPENDITURE ON HEALTH PER CAPITA
COUNTRY
CNP per capita
(1965 U.S. S)
Expenditure on health
per capita
(1965 U.S. $)____
1963-1964
2000
1965
2000
Indonesia
99
123
.0.20
0.25
Nigeria
83
125
0.50
0.75
Thailand
126
402
0.60
1.91
Colombia
277
359
3.50
4.54
United Kingdom
1804
6530
56.00
202.26
VKK r 1 rl
nrvnn'f'
T
problem (2) can bo approached with more optimism.
Bryant
formulated it'in a positive way by regarding as a challenge the problem
of reaching "most of the people with health care using severely limited
resources in ways’ that will yield the .greatest possible benefit".
In
other words the problem is one of planning and management.
Successful planning and management, even ata local level,
requires that the functioning of health services be periodically examined,
yor too long the assessment of health care delivery has
been based on unscientific, subjective impressions gained from occasional
visits to clinics or the perusal of routine clinic reports which frequently
emphasised the attainment of certain set targets and excluded information
oh other aspects of clinic activities. ( 41 ).
With the aim of correcting this state of affairs recent
research has tended to concentrate on two broad aspects of the study of
health care delivery.components of health
(1) the problem of quantifying the many
care delivery systems,
(2)
the development of methods which can be applied to produce
reliable information which is adequate for scientific
evaluations.
Thus far much of this research has involved some lengthy and
complex procedures directed towards the analysis of different types of
health care delivery systems.
This has been of undoubted value in
aiding the future planning of suitable health care facilities for a
community but due to the relative complexity of the methods involved it
has been unable to contribute to the improvement in the day to day working
of individual PHCUs.*
The scope for improvement in the delivery of health care in
many countries, including the'u.K. and U.S.A., is illustrated by the wide
variation in both the quantity and quality of health care provided by
individual practitioners, often where there is equality of available
resources. (2
).
The need for improvement in the delivery of
health care has already been illustrated (Table 2 and Table 3) by the
continuing existence of high rates of mortality and morbidity, often due
to'diseases which are preventable, curable, and frequently both.
It is
*PHCU is an abbreviation for Primary Health Care Unit, a term used in this
s’tudy to denote all types of primary health centres which have at least
one medical officer.
-5~
accepted that improvement can bo achieved by systematic reviews of
The implication,
the working of an individual PHCU. (27,32,41)
therefore, is- that such systematic reviews must become an integral
However, in order
part of the activities of each and every PHCU.
to become a feasible objective in a situation of severely.limited
resources three important constraints must be placed on any techniques
of evaluation which are to be employed:(i)
the methods of evaluation must be simple
(ii)
the time required for evaluation must be short
(iii)
the resources required, in terms of money, equipment,
and manpower must be small.
This study set out to formulate and test methods appropriate
for the evaluation of the delivery of health care by PHCUs in developing
Thus the constraints listed above were of particular
countries.
importance, especially due to extremely limited available resources.
Further importance was given to the following aims as being
desirable in order to achieve maximum benefit from any evaluation of
health care delivery:(1)
evaluation should be a regular event in any PIICU,
e.g. annual.
(2)
the staff of the PIICU, especially the medical officer,
should be involved as much as possible in carrying out
the evaluation.(25,27,34)
Maximum staff involvement was considered desirable because
it would encourage them to see their role in the local community in much
broader terms and thus, potentially, encourage them to think about their
approach to the health problems of their community in broader terms.
However, it vzas recognised that the maximum involvement of staff would
not bo feasible in certain circumstances e.z. due to lack of staff
In these
motivation, shortage of suitably literate staff etc.
circumstances the task of evaluation could conceivably be carried out
by ’evaluation teams* from National, State or District Health Authorities.
in order to formulate appropriate methods three PHCUs were
studied.
These were the Mallur Health Co-operative and two adjacent
government PHCUs in the State of Karnataka, South India.
The Health
Co-operative is a pilot project which is exploring a new method of
financing hea-lth care in a rural community and is based in the village of
Mallur near Bangalore.
It receives' its finance from the local Dairy
-6~
Co-operative and aims tc provide total primary health care for the
4,500 people of the five member villages.
.
The organisation and
functioning of the Health Co-operative is supervised by the Department
of Preventive and Social Medicine, St. John’s Medical College, Bangalore,
at whose invitation I carried out this study.
The 2 government PlICUs were chosen as they typified the
health care facilities available to the rural population in this part of
S.India.
All 3 PHCUs were in the same administration district and thus
In addition the
were accessible for the purpose of conducting the study.
people were culturally comparable and yet there was no overlap of the PHCU
catchment areas.
tv
/
REVIEW 01- TIIEJ.ITERATdRE
The term ’evaluation’ in the field of health care refers to
the process of measuring the effectiveness of' the processes by which the
health programme attains its objectives.
The study of health care in general has been gaining momentum
increasingly over rhe past decade as it gradually became obvious that the
demand for care was limitless and the resources to cope with the demand
Thus an urgent need has arisen for rational
increasingly limited.
planning to enable resources to.be allocated according to need and for
measurement of the effectiveness of such planning decisions. ( 26,35
).
Too often in the past the assessment of the delivery of health care has been
based on unscientific, ’subjective impressions in the absence of sound
information and frequently on an entirely intuitive basis. (26
).
The need for routine assessment of the delivery of health care
is now widely recognised as a means of aiding improvement Of services.
(12,32,41)
Similarly, the scope for improvement is apparent in many
countries including the U.K. and U.S.A, as illustrated by the wide variation
both in quantity and quality of services provided by individual practitioners,
( 2
even where there is equality of available resources. (2
).
Approaches to the problem of evaluation methods
There is general agreement that the primary objectives of
evaluation are to provide relevant information to those involved in
operating the health services at local level so that rational decisions can
be made concerning day to day manap;e*nent as well as for future plannino parp^s-
(3,12 26-23,41, 43, 44,46,47,48,50, 52-54, 56.
However, two major basic difficulties occur in developing a
successful evaluation systom:;1. Operaticn difficulties which arc similar to those facing
any developmental effort e.g. planning activities,
distribution of responsibilities, allocation of tasks to
(27)
personnel etc.
2.
Theoretica1 difficulties.
These can be separated
usefully into:~
(a)
What to measure.
(b)
How to measure.
There is unanimous agreement that the theoretical difficulties
have provided the major obstacles.
A 1967 WHO paper provides an early example of attempts to
combat these obstacles.
are as follows:~
(52)
Tho main components of this approach
(i)
(ii) ■
(iii)
(iv)
Define the objectives of the programme
Identify the component activities
Define the measurements and standards used
Define the baseline to be used or select controls
(v)
■ Describe the methodology of data collection
(vi)
Analyse the data
Later' studies have employed this objective orientated approach
)
with additional refinements. (
Considerable difficulties were encountered in implementing
some of the above steps
especially indentifying component activities,
defining suitable measurements and standard q ond developing a suitable
(27,42,50)
methodology for data collection
Recognising these difficulties Roemer, in a review of
evaluation methods, formulated four major levels of evaluation in order
of importance but also in order of measurement difficulty.(1)
(ii)
Provision of resources
Assessment of services provided and received
(iii)
Assessment of the quality of services
(iv)
Measurement of improvement in health
resulting
from the services.
(iv) is the best level of evaluation since it provides a
measure'of the ultimate objective of any health service i.e. an outcome
measured in terms of change in the health status of the catchment
population. ( 44
).
Another major type of approach began with attempts to
describe a health care system in the form of a conceptual model.
PROCESS
INPUT
(6,28,42)
OUTPUT
At its most basic it was considered that any health care
system has three components which can be classified as Inputs, Process
and Outputs where the Inputs are the health problems of a community plus
its resources; the Processes are the health care activities; and the Outputs
are the services performed which in turn affect the health status of the
community. (50)
In practice more complex models are required.
•
is that developed by Logan et al (see Figure below).
“9-
An example
FIGURE A
NEED
v
demand
i
ft
“
E-.
r.,_ y, )*
i
t *
> USE
RESOURCES
'
■■■■
COST
V
OUTCOME
in effect this model is a more precisely defined version
of the model above and thus is more useful for purposes of scientific
However, two of the components i.e. Need and Outcome,
measurement.
present major difficulties of measurement, the former because there is
no commonly accepted definition ( 13 ) and the latter because of
difficulties in clearly specifying outcome criteria other than in
relatively few medical conditions.
(7,33,38,46)
Nevertheless, measurement of Outcome in still regarded as
the best level of evaluation amd much modern research remains directed
towards defining adequate measures.
Mortality is one unequivocal
measure which will continue to play an important role in many studies but
more specific measures are required, particularly of morbidity.
(49)
Although much refinement is still required some promising results have
emerged from attempts to measure morbidity in terms of its effects on the
lives of individuals e.g. functional or social disabilities. (^
Related to the above is another important asnect of any health
care delivery system i.e. the quality of care provided^.19 ’
A recent
American study set out to explore methods of evaluating quality of primary
health care and
introduced the concept of ’indicator cases’.
An
’indicator case’ was defined as ”an example of either health or illness
care that is statistically frequent, which may be
provided by a variety
of different systems, about which there is general agreement concerning
appropriate management, and in which maintenance or improvement can be
objectively measured”. ( 15
)
Four indicator cases were chosen i.e. normal pregnancy, normal
infancy, adult onset diabetes, and congestive cardiac failure.
Numbers
of patients within these categories wore studied in order to try to measure
the care as well-as cure provided by physicians.
-10-
Although major problems of measurement remain to be
on the best sources of
solved researchers are in general agreement
the evaluation of primary health care delivery.
information for
These arer
1.
’ The catchment population
2.
Clinic records and reports
3.
Clinic staff
Similarly the following techniques of data collection are
generally recommended:-
1.
in order to
Sample surveys of the population
obtain information about their health needs
2.
Examination of clinic records and reports in
order to obtain information about demand lor
medical care and available resources
3.
Examination of clinic records and reports plus
observational studies in order to obtain information
about health care activities (process)
4.
Examination of clinic records and reports in order
to obtain information about outcome and cost.
Cost is an important
in circumstances of limited resources
Indeed, it has oeen advocated that
component of any evaluation system,
should be economic resources
the starting point for health care planning
available and not data about disease.(i)
to the costing of health
There have been two basic approaches
the total expenditure
care i.e. averaging and surveying. ■ In the former
receipt of the results of
is divided by the total number of units in
e.g. average expenditure
expenditure. Thus Ccst may be expressed in terms of
This type of
clinic attendance etc.
per head of population, or per
of good records in order to be successful
costing relies on the maintenance
(5,27,29)
method of costing but is also
Surveying is a more accurate
The need for a survey
more expensive in terms of resources required.
usually arises because of a lack of suitable data ( 8, 9, 36,37)
in practice most costing relies on a combination oi che two
methods because averaging does
does not
not provide enough detail and surveys do
not have a wide enough applicability.
-11-
I
This review of the literature describes the progress thus
far of research into the process of evaluation of the delivery of
health care and it is clear tfoat this field of research is still
Ideally since good health involves not
very much under-developed.
only physical well being but mental and social well being also any
evaluation should involve; measurement of all three components. However,
recognising the shortcomings of present methodology for evaluation it
has been concluded that in order to gain further experience in the
techniques of evaluation current research should remain concentrated on
data measurable in
the physical component of health which contains more
scientific units, a more certain literature, and the outcome of which is
more precisely measurable.
(
2
).
-12-
• METHODOLOGY
The major basic difficulties facing the development of
any
evaluation system i.e. operational and theoretical, have already
The importance of these problems was highlighted by
been discussed.
my own
own previous
previous experience in developing countries and so.the development
of a methodology began with a statement of the following aims:-
The evaluation system must be intimately associated
1.
with the decision making processes and, therefore,
with the local decision makers.
In particular
the .involvement of the PIICU medical officer'was
considered imperative.
2.
The evaluation system must be capable of producing
reliable information which is relevant to the needs of
the local decision makers.
3.
in view of the problem of severely limited resources
facing all developing countries the evaluation system
I
must have the following limitations in order to be a
feasible, practical tool;a)
b)
the methods must be simple and easily applicable.
impl.^■mentation should require the minimum possible
time.
c)
the resources required in terms of money, manpower
and equipment must be as small as possible.
Development of a conceptual model
The delivery of health care is a complex process involving
a large number of factors.
It is essential for an evaluation system
to provide an
an analytical framework which permits an understanding of
<
the components which go to make up such a process.
One possibility is to condense the performance of a system
into a few broad categories in a similar way that GNP is used to describe
a nation’s performance,
•z Alternatively every single activity can be
examined and evaluated ( 27 ).
Neither of these extremes is acceptable since the first is
too crude and insensitive and the second would require extensive data
processing facilities which would be totally prohibitive in circumstances
of'limited resources.
In order to conform to the aims outlined above this study
Figure B
CONCEPTUAL MODEL
CAjRE
' HEALTH
’—I
activities!
I
\
\
MEDICAL RELIEF
MATERNAL AND CHILD HEALTH SERVICES
FAMILY PIANNING
COMMUNICABLE DISEASE CONTROL
• ENVIRONMENTAL SANITATION
LIAISON
ADMINISTRATION
HEALTH FUNCTIONS
OF A
P. H. C. U.
I
HEALTH
NEEDS
V
r
INPUT
PROCESS
DEMAND
PHCU ACTIVITIES
IN TERMS OF QUANTITY
RESOURCES
QUALITY, EFFICIENCY
OUTCOME
MEASUREMENT
OF---OUTPUT HAVING
ASSUMED --------- . HEALTH BENEFITS
- >
MEASUREMENT
OF CHANGE IN
HEALTH STATUS IN
TERMS OF MORTALITY
go MORBIDITY
I
COST
• •••
• ••••
• •••
•••••••••
• •••
sought to formulate a modcl which would
(a)
adequately describe a health care delivery system
so as to facilitate analysis, (6,23).
and (b)
would act as a teaching model to those local health
personnel, especially the medical officer, approaching
the subject of evaluation for the first time.
The model chosen is described in Figure 'B*.
of this model is the well
The backbone
accepted concept of health care delivery
consisting of three components, i .e.
—> ■ PROCESS
INPUT •-—
-------
OUTCOME
a community plus
input refers to the health problems of
to-deal with these problems,
the resources of a health care delivery system
involved in the delivery of health
Process refers to the many activities
health,care activities
Outcome refers. to the end product of the
care.
health status of the population.
i.e. a measure of the change in the ]
Building upon this basic: framework the components of the
model described in Figure ’A* were adopted as they appeared to meet the
criteria of adequately describing a health care delivery system as well
as being suited to teaching.
The components were defined as follows;-
This refers to the felt needs of the local
Health Needs.
population for health care,
There is obvious scope for a lot of value
judge-rent in the assessment of need and thus objective measurement is
extremely difficult.
This is closely related to need
Demand for health care.
and represents those people who not only have a felt need for health care
PHCU in order to receive care.
but actually present themselves at a .
> made by consulting attendance
Measurement is much easier and can be
records.
and equipment which is
Resources i.e. the money t manpower
allocated to a PHCU.
Process.
This refers to the many activities performed by
For purposes of evaluation it can
PHCU staff in coping with demand.
be separated into three categories i. e. quantity, quality and efficiency.
-14-
The importance of this component has already
Outcome.
been discussed in the Review of Literature as well as the difficulties
Ideally outcome is measured in terms of change in
of measurement.
health status of the people using a PHCU i.e. in terms of change in
However} measurement in these terms is
mortality and morbidity.
not possible without well established data collection procedures which
are non existent in many real life PHCUs.
,
in order to deal with these shortcomings an intermediate
component, termed output, has been introduced.
Output will be
measured in terms of proportions of the catchment population who have
received services from the PHCU which have assumed automatic beneficial
effects on health status.
Thus, for example, the proportion of
the population receiving vaccination coverage, proportion of deliveries
conducted by trained staff, porportion of drinking wells routinely
chlorinated, were all considered to have automatic beneficial effect on
health status and thus were regarded as legitimate, if second best,
measures of outcome of delivery of health care.
Cost.
In general the most stable feature of a health
care delivery system is that of available resources which can be
measured in. terms of cost both cheaply and easily. ( -7
This
).
total cost can then be allocated e.g. according to each type of service
providcd or per patient contact, thus providing information of economic
interest and in addition providing another crude measure of system
performance.
The simple costing procedure outlined above was considered
too crude principally because it makes no allowance for the major
division of PHCU work into that carried out in a clinic building and
that carried out in the community.
For this reason the study adopted
the system developed at the Rural Health Research Centre, Narangwal,
Punjab. ( 5
).
In effect this system divides costs not only
between the major sectors of clinic and community, but also within each
sector costs are allocated to broad divisions of health care activities.
(Sec next section for detailed description).
Having decided upon the major components of the model the
conceptual problem remained of fitting into this model the many health
care activities involved in the provision of services.
Examples
of such activities range from the examination, diagnosis and treatment
of patients’ ailments to the filling out of administrative forms etc.
The problem was alleviated by grouping the health care activities into
broad divisions or FUNCTION'S, similar to those mentioned under cost above.
-15-
Seven functions were used in the study as follov/s:Medical Relief
Maternal and Child Health Services
Family Planning
Communicable Disease Control
Environmental Sanitation
Liaison with other Health Care Services
Administration
The introduction of Health Functions had the following
advantages;-
(i)
They acted as a bridge between health care
activities and the components of the model
(ii)
They reduced the number of variables to be
considered
(iii)
They aided objective analysis since they lay
outside traditional administrative divisions.
(For definitions of health Functions -see Appendix)
Sources of Information
In order to obtain adequate information for the analysis
of all the components of the model the following sources of information
were utilisod;1.
The catchment population of the PHCU
2.
PHCU records and reports
3.
The PHCU staff
Methods of Data Collection
It was necessary to use multiple techniques of data
These may be summarised as follows;-
collection.
1.
Information bn health Needs was obtained from a
questionnaire survey of a sample of the catchment
population.
2.
Information on Demand for health care was obtained
from examination of the existing clinic records and
reports.
3.
The component of Process was analysed in terms of
quantity, quality and efficiency.
-16-
(a)
information for the analysis of quantity
was obtained from existing clinic records and
reports.
•(b)
Information for the analysis of quality of
health care was obtained from a questionnaire
survey of selected recent patients.
(c)
Information for the analysis of efficiency of
delivery- of health care was obtained from an
observational study of PIICU staff activities.
4.
Information on available Resources in terms of staff,
equipment and finance was obtained from examination of
existing clinic records and reports.
5.
Information for the measurement of Outcome was obtained
from existing clinic records and reports as well as from
the questionnaire survey of the catchment population.
6.
Information for the analysis of Cost was obtained from
existing clinic records as well as from the observational
study of the activities of the staff.
DETAILED METIIODOWGY
For the purposes of developing and testing the methodology
All three were
the kind co-operation of throe PHCUs was received.
situated in Kolar District, State of Karnataka, South India.
Two were
run by the State Government and the third, the Mallur Health Co-operative
was an independent pilot project (Sec Introduction).
(A)
Questionnaire Survey of Catchment Populations
There has been general agreement amongst resoachers on the
usefulness of surveys of local populations to obtain information nou
otherwise available. (26-41^O)
However, it would neither be practical
nor scientifically acceptable to attempt to elicit all desirable information
Thus the subject matter of questionnaires must
frorn a single survey,
be decided according to local priorities.
in this study the questions
were concerned mainly with Maternal and Child Health, a topic of major
importance in all developing countries.
-17-
I
Target Population:
samples of member villages of the Mallur Health
Co-operative Society and those of the catchment areas of the two
Government PHCUs.
Data Collected; the questionnaire contained 63 questions designed
to gather information about the local population.
education,
(a) General information i.e. age, sex,
occupation and income of the inhabitants of
each househld.
(b)
information about Environmental Sanitation, i.e.
number of residents in each household, presence or
absence of a sanitary latrine, source of drinking
water.
(c)
Information about Maternal and Child Health.
As
a subject of major importance in all developing
countries this category was the one most thoroughly
examined by the questionnaire.
The information
sought was as follows:Number of live children, previous still births,
(i)
previous miscarriages, previous child deaths.
This information contributed towards an estimation
of need for Maternal and Child Health services
in each of the catchment populations studied.
(ii)
Uptake of PHCU services as assessed from present
and recent pregnancy practices regarding antenatal
examinations, care at the time of child birth and
attendances for post natal examinations.
(iii) Information about each individual child undei
5 years old i.e. number and type of vaccinations
received, number, type and duration of illnesses
in the past month and whether or not a doctor was
consulted.
This provided information about
Need for care for the under fives as well as an
estimate of Outcome in the case of vaccinations
and uptake in the example of recent illnesses.
(iv)
Felt need of the population for health care was
assessed by enquiring into knowledge and attitude^
towards certain common diseases and conditions.
These examples were otitis media, diarrhoea in under
i
I
/
fives, worm infestation, and care during pregnancy.
An assessment of the need .for Family Planning•services
was obtained from information about family size, birth
interval and family planni.ng knowledge and practice.
The information about family planning practice facilitated
an assessment of Outcome.
Pilot Study to test Questionnaire
Thirty village mothers co-operated in the testing of the
questionnaire.
Assessment of’the validity and repeatability presented
problems but these were diminished by the fact that the questionnaire
aimed to characterise groups of the population rather than individuals.
The information about previous deaths, still births, and miscarriages
could not be validated and the accuracy of information received from,
each respondent was obviously suspect due to the problem of recall.
However, since the purpose of the questions was merely to help in
assessing birth interval amongst village mothers this section of questions
v/as retained.
' . ’
The problem of good repeatability was approached by making
maximum use of-closed ended questions, many with a simple ’yes’ or ’no
reply.
in addition each pilot interview was timed and showed a mean
time for completion of 21 minutes with SD-4.5 minutes, thus indicating
a fairly uniform ability of respondents to answer the questions.
Unit of enquiry:-
Mothers with children aged under 5 years
residing in villages served by the three
PHCUs.
Sampling Unit:-
village household.
Sampling Frame;-
Lists of village households maintained
by each PHCU.
Sample Design;-
The sample of households was selected
using a two stage technique (55)
In the case of the government PHCUs only
villages within a three mile radius of the
PHCU were chosen because previous Indian
studies have shown that some 75% of
consultations are from people living within
Since all the villages
a three mile radius.
of the Health Co-operative were within a three
mile radius of the PHCU no exclusions were
necessary.
In order to obtain a population roughly equivalent to that
of the Health Co-operative villages the total number of villages in
each of the two government catchment areas was weighted according to
Thus six
the average number of households per village in the area,
villages were required in one government PHCU area and eight in the
The total number
other, the villages then being selected at random.
as follows
of households thus selected in each first stage sample were
MALLUR 638 HOUSEHOLDS
GOVERNMENT
PIICU
(B) 855 HOUSEHOLDS
GOVERNMENT
PIICU
(J) 822 HOUSEHOLDS
150 second stage units (households) were then selected from
t'ne sampling frame in each PHCU, the number in each village being weighted
according.to the number of houses and the
households then being selected
using Random Number Tables.
Since the object was to interview a total of 100 mothers with
children aged under 5 years in each area a total of 150 households were
selected to allow for the fact that some households would contain no under
fives.
This was unavoidable since the sampling frames did not accurately
document the ages of the members of each he usehold.
(B) Examination of Clinic Records andReports
The clinic, records and reports for the past complete year i .e.
information about Demand for medical
1975, were examined in order to obtain
Cost
care, available Resources, Process involved in coping with the Demand,
and Outcome of the delivery of health care.
The information was categorised
by Function.
(C) Assessment of the quality of medical care
The assessment made use of the concept of Indicator Cases,
an indicator case being defined as "an example of either health or illness
care that is statistically frequent, which may be provided by a variety of
different systems, about which there is general agreement concerning appropriate
management, and in which maintenance for 'improvement of health status of the
’ individual studied can be objectively measured".
indicator cases were used;-
1.
Anaemia of pregnancy
2.
Diarrhoea in under fives
3.
Skin abscess
-20-
In this study the following
/
A register was maintained at each PIICU in order to ‘record
the names and addresses of patients whoso diagnoses fell within these
categories.
A questionnaire wass then put to these patients in order
to assess the quality of care which they had received,
The categories
for the assessment of quality of delivery of health care wore as follows:
Physician performance i.e. type of history elicited, extent
of examination carried out, method of diagnosis employed,
details of treatment, and management of the patient’s disease.
Physician communication i.e. communication by the doctor
of the diagnosis and details of treatment etc. to be given
to each patient.
Patient compliance i.e. patient’s actions in carrying out
the instructions given by the medical officer.
Cost and convenience to the patient i.e. in terms of wages
lost and travelling cost plus travelling time and time spent
waiting at the clinic for the doctor’s attention.
Referral or re~appointment i.e. for further diagnostic tests
or treatment.
Patient satisfaction
!'
i . e. patient’s opinion of the result
of this treatment.
(For example of above questionnaire see Appendix)
(D) Assessment of Efficiency of Health Care Delivery
Efficiency•of the delivery of health care was assessed by
measuring the amount of time spent by the members of both clinic and field
staff on activities which were categorised by Function.
The Functions were
those already mentioned i.e. Medical Relief, Maternal and Child Health sei vices,
Family Planning, Communicable Disease Control, Environmental Sanitation,
Liaison and Administration.
In addition categories were included to allow
for time spent travelling on clinic business, time spent on personal activities
and a final miscellaneous category to allow for activities such as assisting
Medical Relief e.g. helping another staff member perform a minor operation,
put on a dressing etc., as well as cleaning the clinic premises.
The measurement of time spent on each category was carried out
by means of an observational study,
An observer was posted to each clinic
for two weeks i.e. 10 working days,
During that time each observer spent
approximately the same number of working days with clinic and field staff,
the division depending on the numbers and responsibilities of the field staff.
The programme of attendance by the observer either at the clinic or with field
staff was arranged using random number tables.
In order to record the time spent by each member of staff
the observer was provided with a form (See Appendix) on which the
activities were listed by the functions mentioned above.
The functions
were then subdivided into more'specific activities e.g. medical relief
was subdivided into the eliciting of a history1, examination of the patient,
prescribing of drugs, dispensing of drugs, putting on a dressing, giving an
injection.
Beginning at the clinic opening time in the morning the
observer then recorded (by placing a tick in the appropriate category)
every two minutes throughout the working day, the activities of each staff
member.
In the case of field staff the observer accompanied one staff
member during his working day in the villages and maintained the observation
form as above.
Thus, for example, during a six hour working day 180
observations were made of the activities of one staff member, permitting
the building up of a profile of the day to day working.
In order to check that the correct number of observations had
been made the working hours of each PIICU were recorded each day of the ten
day study and then the number of recordings on each form checked against
these.
A previous study by Feldstein et al (20), measured the time
spent by staff members on health care activities.
However, this consisted
of a time and motion study of each task performed by a staff members.
Given the important constraints on methods used in this study such a time
consuming exercise was rejected in favour of the above.
(E) Analysis of Cost
A simple method of cost analysis was envisaged whereby the
PHCU resources could be allocated by the functions already used in the
other techniques.
This involved the summing up of annual expenditure,
both capital and recurrent, and then allocating these to Functions using
the data from the assessment of efficiency study in section (D) above.
This system was originally developed at the Rural Research Centre,
Narangwal, Punjab.
'
Expenditures were derived according to the following
procedures:Capital expenditures:
Capital expenditures for buildings
and equipment were obtained from various PHCU records and
reports.
The total capital expenditures were then depreciated .
using an arbitrary 1% annual rate of depreciation.
The 1%
figure was considered appropriate in the circumstances of this
study since the PHCU buildings and equipment were very basi
and
required and received very little in the way of maintenance etc.
-22-
Maintonance Exoendj turc: The annual maintenance expcnairuru
for buildings and equipment was assessed from existing clinic
records and reports.
Drug Expenditures:
The annual expenditures for the purchase
of drugs and other expendable items wore readily available
from PHCU records and reports.
Salaries: Annual expenditures on salaries and allowances
for PHCU workers were similarity obtained from existing
records and reports.
Allocation of Expenditures by Function
The procedure for costing PI1CU activities on a functional
basis involved allocating all the PIICU expenditures among the specified
functions.
The allocations were made as follows:.
Capital expenditures:
The 1% annual capital depreciation
figure was allocated in the first place either .to PIICU
or the Field and was based on an estimate of the principle
location of the staff activities.
Subsequent allocation
of expenditures to the functional categories was based on
the results of the Assessment of Efficiency study.(See
section (D) above).
Recurring Expenditures:
(i)
Maintenanco Expenditures
Building maintenance expenditures were distributed
in the same manner as the capital depreciation
expenditures mentioned above.
Equipment maintenance
expenditures, where they existed, were distributed
according to whore the particular item of equipment
was most utilised e.g. the Mallur medical officer’s
motor cycle was utilised entirely in the field for
activities within the functions of Medical Relief
and Maternal and Child Health.
(it)
Drugs
The total expenditures on drugs were allocated by
examining the PHCU records and separating out drugs
used in activities related to specific locations and
functions.
Expenditures on personnel Salaries and Allowances:
The salaries and staff allowances were allocated among the
various functional categories using the data from the Assessment of
Efficiency study (as in section (D).
-23-
AD/WING TO CONSTRAINTS
Mention has already been made of the importance of the
methodology conforming to the following constraints:(a)
methods should be simple and easily applicable
(b)
implementation should require a minimum of man hours.
(c)
resources required in terms of money, manpower and
equipment should bo as small as possible.
The following techniques were affected by these constraints:~
■ i.
Questionnaire survey of samples of the population
ii.
The ’quality of care* questionnaire
iii. The assessment of efficiency study.
(i)
Questionnaire Survey
It was recognised from the outset that this technique would
present the most difficult problems in terms of time, manpower and
The important factors for consideration
organisational requirements'.
were as follows:Length of questionnaire i.e. the number of questions,
number of respondents required,
number of interviewers
required,
the quality of interviewers required i.e. trained or
untrained,
access to respondents.
Initially an arbitrary limit of one month was imposed as the
requirement for completion of the questionnaire survey in all three centres.
One month was chosen as being a likely upper limit of time for which a
P1LCU could possibly spare a member of staff for participation without serious
disruption of services.
The subject matter of the questionnaire which has already been
outlined consisted of 63 questions.
A pilot study using 30 respondents
found that the average time required for completion of each was 21.3 minutes,
SD~ 4.5 minutes.
The choice of the number of respondents was chiefly dependent
on the following factors i.e. the arbitrary limit of one month for completion
of the survey and the number of questionnaires which could reasonably be expeetc
to bo completed in one day,
The latter in turn depended on the accessibility
of the respondents and the number of interviewers employed.
Number of Interviewcrs
It was considered desirable to employ a total of three interviewers
but for practical reasons it was possible only to employ two, one of whom had
previous experience and training in survey techniques and the other who had
neither training nor experience.
-24-
//
•
I
Accessibility
t
*
•
•'
•
Having taken into account the terrain and public transport
facilities available in the three study areas it was concluded that between
5 and 10 questionnaires could be completed daily without undue effort.
Thus in one month i.e. approximately 20 working days, the range of completed
questionnaires would, with two interviewers, be between 200 and 400 in all
three centres.
*
From these rather crude calculations it was decided to
e. a final total of
make 100 respondents the target for each study centre i.e.
300.
Although the information for each questionnaire was gathered
from one respondent i. e. the mother of each household, it must be stressed
that the.information covered an entire family which had an average membership of 6 (from a previous report of the Mallur Health Co-operative),
This meant, in effect, that 600 people were covered in each of the three
study centres.
(ii)’Quality of care1* questionnaire
The concept of indicator cases and the six criteria employed for
the assessment of ’quality of care’ have been described previously,
It was considered that the best way of obtaining adequate
information whilst ensuring simple methods and minimum resource requirements
was to incorporate the criteria into a questionnaire (see Appendix) which
would ‘then be put to patients who had recently attended the PHCUs for
treatment for each of the 3 indicator cases, i.e. anaemia in pregnancy,
diarrhoea in under fives and skin abscess.
It was decided that the assessment of quality of care provided
by a PHCU could most usefully; be made by judging whether or not certain basic
procedures, which could reasonably be expected to be carried out, were in
fact carried out in the course of treating a.patient.
In addition, and
■ less importantly, consideration was given to cost and convenience of treatment
to the patient.
In effect quality of care was judged on the answers received
to the following questions:Was a history taken?
Was an examination carried out?
Was a diagnosis made purely on clinical grounds
or were aids to diagnosis employed e.g. Hb test?
Was the patient informed of the diagnosis?
V.’hat treatment was prescribed?
Were appropriate instructions or advice given to
the patient concerning management of the condition?
Was the treatment successful in the -patient's
opinion?
The remainder of the questions dealt
with cost and convenience
of treatment to the patient.
The three indicator diseases required different management
However, due
different questionnaires were necessary.
so that three
for variation
to the general. nature of the questions and. the narrow scope
in management
the differences between the questionnaires were very minor,
For practical reasons it was possible only to carry out this
A
Mallur Health Co-operative.
part of the study in one PHCU i.e. the
'and addresses of 10
register was maintain.od in the clinic and the names
collected.
When
patients suffering from each of the three diseases were
homes and completed
the register was completed an interviewer visited their
the questionnaires.
This part of the methodology was
unsatisfactory foi' the following
reasons:(a)
no.attempts were made to test the validity and
repeatability of the questionnaires,
(b)
the numbeT involved were too small.
(iii) Assessment of efficiency study
The methods described previously were considered both suitably
simple and easily applicable so that only the constraints of time and
minimum resources were of importance.suitable
Two working weeks i.e. 10 days, were chosen as a
duration, with 5 days allocated to observation of the activities of the
This ensured that any
Clinic staff and 5 days to the Field staff,
or
variations in the PHCU weekly programme would be included e.g. a day
For
part of a day when only antenatal examinations were performed.
a PHCU with an average working day of 8 hours, a total of 2,400 observations
Although this was considered, sufficient for the construction
were recorded.
of an accurate profile of PHCU activities it was recognised that the
observations could be biased due to the study coinciding with a period when
staff activities wore
diverted to special programmes, e.g. during an epidemic,
and.thus interpretation of the results could only be made after consultation
with the Medical Officer to ensure that no such special circumstances existed,
observer
The chief component of resources utilised was manpower, one
being allocated to each PHCU.
However, since the methods were
simple the only essential attri bute, apart from average intelligence, was
-26-
x
I
l
I
an ability to read and write and thus it was possible to employ three young
men from local villages, thereby reducing to a minimum maintenance and
travel expenses during the 10 days of the study.
The equipment required was
also obtained cheaply and consisted of the observation forms, pens and
watches.
Watches were owned by two of the observers and in the third
case one was borrowed from a PIICU staff member.
■
‘‘
I
\
i
r
’
/
-27“
11 .J
KJ J M
C »>
x --•*
The results are presented as an illustration of the competence
of the methodology in obtaining information and thus they have been combined
into a single chapter together with the discussion.
The data from the Mallur Health Co-op is presented as the main
topic foi' discussion but data from the other. PHCUs is referred to when
comparisons are relevant.
The components of the conceptual model (See Fig.B) were used as the
basic framework for the analysis of data and the results have been presented
in sections corresponding’to the 7 health Functions.
Each Function has been
analysed and presented in terms of Demand, Process, Outcome, Resources and
Cost where appropriate.
The results concerning Need for health care have been
presented separately due to the controversy over its definition and the.
consequent difficulties of measurement.
However, as
an important preliminary procedure the boundaries of
as an
responsibility of each PHCU have been determined by measurement of each catchment
I
population (48).
Catchment Population
Measurement of the catchment population is a very important
first step in the study of any health care delivery system,
It sets tie
boundary of responsibility for health care and as such is a basis denominator
which is essential for the calculation of many parameters of health care
delivery.
For example, knowledge of population size, even as a single number,
permits a crude estimate of potential demand for health care and this in turn
gives a crude indication of staff and other resource requirements.
If the
age and sex structure of the population is also known then the potential tor
If, for example, the size of the under
more refined planning is much enhanced.
5 population is known then it is' possible to estimate the numbers of staff
required to carry out an immunisation and developmental screening programme.
Where the numbers of staff is fixed due to financial constraints then the
information can be used to organise the most efficient programme which is
possible under the circumstances.
Knowledge of the geographical distribution
of the population is of obvious importance in planning such a programme since
the time required for staff to travel from the PHCU to the homes of their
patients must be included within their working hours.
The time, effort and cost required for patients to reach the PHCU
has important implications especially since studies in developing countries
have shown that “patients within the theoretical catchment area but living more
than approximately 3 miles from the PHCU tend not to attend for treatment (14,41)
-28~
TABLE (>
MALLUR AGE ALP SEX DISTRIBUTION AS IN
OCTOBER 1975
AGE GROUP (IN YEARS)
VILLAGE
0
5
5
15
15
30
30
50
50
60
M
F
M
F
M
F
M
F
M
F
M
F
MALLUR
169
210
242
319
333
310
228
210
43
60
57
89
2270
KA CI IABALLI
24
25
37
44
44
44
33
43
30
20
13
10
367
MUTHUR
102
90
156
158
128
102
142
151
37
46
16
17
1145
BIIATRENAHALLI
16
25
38
30
24
15
36
33
3
6
5
3
23'4
A llurunagenaha LLI
13
8
13
21
5
8
15
16
2
3
1
3
108
324
358
486
572
534
479
454
453
115
135
92
122
4124
12.95
11.61
11.00
10.98
2.79
3.27
2.23
2.96
TOTAL
60+
I
% TOTAL
7.86
8.68
11.78
13.87
SOURCE: CENSUS STUDY CARRIED OUT BY INTERNS OF ST.JOHN’S MEDICAL COLLEGE, BANGALORE.
I
J_____
TABLE 7
POPULATION
(Age/Sex distribution (%)
age groups
)
(YEARS)
GO +
0
5
5
15
15
30
30
50
50
60
M
F
M
F
M
F
M
F
M
F
M
F
MALLUR
7.86
8.68
11.78
13.87
12.95
11.61
11.00
10.98
2.79
3.27
2.23
2.96
PHCU (B)
9.76
8.56
15.02
13.96
8.26
13.36
11.71
7.36
2.10
2.10
4.35
3.45
PHCU (J)
9.62
10.86
12.21
13.89
10.23
16.18
11.29
7.02
1.07
1.37
2.90
3.36
NOTE: Distributions of Age/Sex for PHCUs B and J obtained from
sample survey.
Thus the potential demand estimated solely from knowledge of the numbers
of the entire catchment population may be much more limited in practice.
The availability of data about the catchment population is
variable, a fact which was confirmed in the study.
The Mallur Health Co-op consists of the population of 5 villages
situated close to one another.
house
Thus it had been possible for a house to
census to be carried out without difficulty (See Table 6).
Such
detailed information was not available for the catchment populations of the
2 other PHCUs because the catchment areas were much larger and the populations
larger and widely distributed in small villages,
However, maps were available
showing the locations of villages and, in addition the following information
was known to PIICU staff:-
BASETTTHALLI PHCU (B)
Total population
11,875
Number of villages •
42
For each village the number of houses and the number of
inhabitants per house were known.
JANGAMKOTE PHCU (J)
Total population
15,402
Number of villages
%
•• •
30
Similarly the numbers of houses and inhabitants were known.
Additional information from Questionnaire Survey
The prime purpose of the Questionnaire survey was to fill in
important information gaps.
Thus an estimate of the Age/Sex distributions
of the catchment populations of PHCUs (B) and (j) was obtained and these
are included in Table 7.
Need
The fundamental problem for any health care delivery system is
to cope with the health care needs of the catchment population.
For this
reason Need has been incorporated as a component of the conceptual model
adopted in this study (See Figure B).
However, the dilficuities
difficulties oi
of quantifying need are very considerable
mainly because there is no accepted definition of the term.
Thus, for example
in a given situation the need for care as assessed by a health professional
may differ markedly from the assessment of a lay person.-
The scope for value .
judgement is enormous and this is of particular importance in developing countries
where the concepts of modern Western health care may clash with traditional
customs and beliefs. (51).
For this reason it is essential that local attitudes
-29"
/
_
should be investigated and an attempt made to assess the Felt Need (13)
of the population for health care in common clinical situations.
If
this
is not done then confidence in local health personnel may be undermined and
the potential of a PHCU to improve the health status of the population
seriously diminished.
In the study a section of the Questionnaire Survey was devoted
to the assessment of Felt Need of village mothers for health care in certain
given situations.
These included common local illnesses of childhood as well
as Felt Need for care during pregnancy and delivery, and for family planning
services.
Table 8 presents the results of investigation into the Felt
Need of mothers concerning childhood diarrhoea, antenatal care and family
planning.
On each of these topics the responses of the Mallur mothers was
significantly different statistically from the responses in the other PHCUs.
The reasons for these differences in attitude could only be discovered by a much
deeper sociological investigation but the fact that they occur, even within
such a small geographical area, illustrates the importance of investigating
local attitudes.
-30“ %
TABLE 8 Felt Need
Is watery diarrhoea serious in young children?
Q.
M (N=100)
B (N=100)
•
J (N-^100)
YES
97
61
69
NO
1
39
30
DON’T KNOW
2
0
1
100
100
100
TOTAL
Differences between Mallur and PIICUs B + J are highly significant
(P< 0.001)
Q.
Is antenatal care important?
M (N=100)
B (N=100)
J (N=100)
YES
44
85
23
NO
39
14
67
DON’T KNOW
17
1
10
TOTAL
100
100
100
Compared with Maf ur the positive response at PHCU (B) was
significantly greater ( P< 0.001) and the positive response at PHCU (J)
was significantly lower (P < 0.005).
5L_.
Do you think Family Planning is good?
M (N=100)
GOOD
B (N=100)
J (N=100)
58
90
BAD
1
3 '
4
DON’T KNOW
1
39
6
100
100
100
TOTAL
Compared with Mallur the positive response was significantly greater
than PHCU (B) (P < 0.001) and than
PHCU (J) (P < 0.005)
Medical Relief i.e. the health care activities involved in the
management of the sick, is the most important of the 7 Health Functions
in terrasof consumption of PHCU resources.
It takes up a large proportion of
staff time and effort as well as consuming a major part of the financial resources
expended on drugs and equipment.
However, the effectiveness of Medical Relief
activities in improving the health status of the population is very much in doubt,
and for these reasons careful evaluation of all aspects of these activities is
warranted.
DEMAND is measured in terms of Outpatient attendances at the PHCU.
The amount of information recorded about each attendance may vary enormously in
different PHCUs.
However, even a record of the total numbers of weekly or monthly
attendances is useful for planning purposes.
These figures will yield information
about the average volume of attendances as well as showing variations which
occur eg. throughout a year.
may
This is important for estimating staff requirements
and possibly the extent of PHCU hours of availability to the public?.
As an example Table 9 shows the attendance figures for the year 1975
in each of the 3 PHCUs which were studied.
TABLE 9
DEMAND (Total) FOR MEDICAL RELIEF 1975
M
B
J
JANUARY
2,089
2,739
1,298
FEBRUARY
1,757
2,170
1,339
MARCH
1,674
2,431
1,491
APRIL
2,109
2,283
1,547
MAY
1,880
2,465
1,840
JUNE
1,847
2,897
1,485
JULY
1,438
2,866
1,493
AUGUST
1,801
3,114
1,532
SEPTEMBER
1,889
2,581
1,193
OCTOBER
1,738
3,327
2,025
NOVEMBER
1,631
2,656 .
1,607
DECEMBER
1,751
2,567
1,616
TOTAL
21,604
32,096
18,466
CATCHMENT POPULATION
M
4,124
B
=_-= 11,875
T
1 R 4H9
Ch
X
t-r-H-i
o
o
XL
f--r
- L.
•
I ■
(11117 7
T
w
w
I
oo
O
4^
O
I
o
_______
XL
r?
O
dZXl
4143
aX
/1-I111-.
XUU-k
|
’
____ . : ;
. 1...-J!-i—j—| 'j-i't'-f
X4-xX4--H4--44H
\.... k
zjnz-i
TEV
I
| =
7 7
i: S’-in
7X41 .: 4yrfc
a-
-j-x ■ i-xj' 4-■-
±^4—t—
X-X -U-X.u__ :
..L4_-pU.--- L 4—;--- u-t ±44-1
1;±-E1113211
•n"H—Hi rT':
heIitrinTi
ex
lii-jV1"'
1177
___ _ __ 1 l. .. .—4------
.:_.p
u—
±EtE
=§ .1.-.:
EEESffix
J_
§L_1L
H-Tyny ’
4
‘
-J
i-i. 141-1;
s: EX
st
g
r-
jXE 7
I
/
7
-k
I
i
I -"-4
I----
X
—f
TZt
y
r
4 -
\
- .. ty.i... 4
3>0
Ilici
—ii-^
I
■t
14.1
i
r -
2
7
g ■
xa-
4 1\
k
F!_____
ETE
i
Eifi
!
r
Ei
k<-
11
!
■■JX
• I
-
.
i
.
i EE?
- it:
i
T“T:
1
- i
• i
—- -
4
X4
..:c
- T -■
-I
/■
[ - •
■
'
/
■
iiiii .i: :i 7
■
5 - i5
ULiij::::::
i ■ •;-
7-8-
]
d——
n
-H-+... 3j~H - _■
E
Q
I
SiE:E-i'gii 4
- cy-- -----
7:771 7
1
iX
7:1
1S
w
,tp±i±tt
L-L H
-T-:7 -
■
ii.m-i
iihb Tjy
xi:’ 7!
k 4
;
iQ
The above figures were plotted on a graph (see Figure C) in order
to demonstrate the variations in attendance more vividly.
Examination of the
attendances at PHCU (B) shows that the period of lowest demand occurred from
February to May whilst from June to August there was a period of high demand.
For planning purposes this is evidence that the early part of the year has the
lowest staff requirements and thus, for example, may be the best time for staff
leave.
However, there may be important reasons for low demand e.g. there may be
a period of very time consuming agricultural activity such as planting seeds or
harvesting so that the local population cannot afford time to travel to a clinic.
In such circumstances it might be better if PHCU staff visited the villages to
investigate any current illness.
Another reason for low attendances may be that PHCU staff were on
leave so that the reduced numbers reflected reduced availability of services
and not low demand.
It is important, therefore, that possible reasons for
variations in demand should be considered.
Variations in the monthly attendances were further explored by
constructing the 3 frequency distributions.
All 3 conformed roughly to the
theoretical Normal distribution although the inadequacy of 12 readings for each
curve must be admitted.
However, on the assumption that monthly attendances were normally
distributed the mean and standard deviation were calculated for each of the
3 PHCUs.
M
1800.3,
S.D. -
185.6
B
2674.7
S.D
334.8
J
1538.8
S.D
+
226.2
Taking Mallur-(M) as an example it could be assumed with confidence
that it would be most unlikely for the monthly attendances to exceed
i.e. 2171, or to fall below
2 S.D. i.e. 1429.
+ 2 S.D.
These figures can then be
employed in calculating the numbers of staff and the clinic time required to
provide services.
As already mentioned variation in demand may be due to many different
reasons and it is important that the possibilities should be considered.
This is
especially important when a period of unusually high attendance is noted because
this may indicate the occurrence of an epidemic.
If any preventive action is to
be initiated in such an event then demand must be monitored continuously in
order that the staff may be alerted at the earliest possible time.
The routine recording.of diagnosis is important not only in the early
detection of an epidemic but also as an aid to constructing a profile of local
morbidity.
Tables 10 and 11 show the monthly attendances at Mallur in 1975
-32-
TABLE 10
O.P. ATTENDANCE BY DIAGNOSIS (MALLUR 1975)
1--------
Smallpox Whoc
Cou£
’Flu
Measles
19
360
0
0
0
38
32
290
0
o
0
9
32
21
182
0
0
0
102
10
33
31
18
218
b
o
o
31
192
2
34
39
18
218
0
0
0
92
21
121
3
11
18
398
0
O
0
329
103
38
163
2
4
12
463
0
O
0
309
189
48
38
404
7
13
19
39
502
0
O
1
SEPTEMBER
208
93
106
53
218
22
9
33
102
416
0
O
3
OCTOBER
69
83
29
22
103
3
28
16
108
366
0
0
1
NOVEMBER
109
169
J02
47
102
8
41
38 •
102
362
28
0
2
DECEMBER
102
106
108
48’
192
6
38
10
162 ‘
398
6
0
1
2060
1047
1201
464
1975
87
313
311
549
4173
34
0
9
H’Worm
Amoeba
6
32
104
9
39
182
106
38
182
Pneumonia
Anaemia
Trauma
98
40
92
94
104
49
MARCH
116
123
APRIL
182
MAY
106
JUNE
106
JULY
368
AUGUST
Coryza
Bronch
itis
JANUARY
292
39
FEBRUARY
39
) •
Ascariasis
Vit def,
0 - NO CASES
- = NOT RECORDED
••••••
• •••
• ••••
•••••••••••••
TABLE 11
—r
~I
Cholera
T.B.
Leprosy
Allergy
Furun
culosis
Scabies
ConJune
tivitis
G-Enteritis
Diarrhoea
212
Otitis
0
16
0
3
1
0
18
1
28
1
0
9
0
33
30
0
0
16
0
0
306
34
6
0
18
1
139
0
1
0
162
91
• 292
42
2
0
0
31
0
MAY
0
92
189
0
6
3
1.
16
89
JUNE
0
29
1
3
0
61
89
29
6
0
0
0
JANUARY
FEBRUARY
MARCH
APRIL
0
JULY
0
1
106
73
1
36
0
0
0
18
62
13
OCTOBER
0
2
1
0
68
162
31
NOVEMBER
0
11
0
0
106
128
39
16
0
0
*1
0
506
786
128
334
0
SEPTEMBER
DECEMBER
1
0 = NO CASES
- - NOT RECORDED
♦Same case each month
••••••••••
32
189
6
63
81
0
AUGUST
92
• ••••
• •••
208
93
198
71
193
102
396
122
2,183
839
2
■
1
30
recorded by diagnosis.
The quality of the data is suspect since the monthly
totals are absent from some of the diagnostic headings.
the diagnoses are suspect.
In addition some of
For instance it is difficulty to accept that only
1 TB patient attended the clinic'during 1975 (See Table 11).
However, 78.9%
of the attendances are recorded by diagnosis and''a summary of the commonest
diseases appears in Table 12.
The largest proportion of consultations was due. to the common cold
and influenza for which only inadequate palliative medicines were available.
No information was available to exclude the possibility that antibiotics were
prescribed in some of these cases at a high cost and little or no benefit.
For planning purposes tho information in Table 12 is useful in 2 ways.
it enables an estimate to be made of which drugs and
(a)
equipment should be available and the likely quantities required.
by identifying the commonly occurring diseases the medical
(b)
officer may be able to train other PHCU staff to diagnose and
treat them thereby releasing him for other wo’rk such as
preventive medicine.
The high numbers of cases of diarrhoea means that all staff should
be capable of coping with the rehydration of at least mildly dehydrated infants.-
They should also be able to deal with cases of minoi’ trauma.
The high incidence of bowel infestation may indicate the need for
A recent study in U.P.
routine deworming of all children.
(23) has demonstrated
the effectiveness of such a programme in improving the nutritional status of
under 5s’.
TABLE 12
DISTRIBUTION of attendances by diagnosis (MALLUR)
% ATTENDANCE
diagnosis
COLD
28.8
+ ’FLU
DIARRHOEAS
BOWEL INFESTATIONS
13.6
RESPIRATORY DISEASES (BRONCHITIS +
PNEUMONIA )
10.0
TRAUMA
9.0
SKIN DISEASE
6.6
OTITIS MEDIA
3.9
VITAMIN DEFICIENCY
2.1
TOTAL
1_
-33“
74.0
■■ > MtVi-
PROCESS
There are so many activities carried out in a PHCU under the heading
of Medical Relief that it would be unreasonable and impractical to attempt to
record them routinely.
However, in the study some measures of process were
possible from PHCU data.
For example from Table 8 the following rates were calculated for
Mallur
Total attendance for 1975
21,604
Average weekly attendance
415
Average daily attendance
75 (5| day wk.)
Average throughput per hour
14 (5.4 hour day)
These are important indicators of clinic workload.
In comparison PHCU (B) had a considerably greater total outpatient
However, because the PHCU was available
attendance in the same year of 32,096.
for 9 hours per day on 6 days each week the average attendance per day was 103
patients but the average throughput per hour was only 11 patients.
Thus,
although the annual attendance was approximately 50% greater than at Mallur
each patient had approximately 1 more minute of clinic time available to him.
However, the significance of this fact depends on what was done for each
patient during his time in the clinic.
Thus for example, there were laboratory facilities available
in Mallur but not in the other PHCU.
During the year 1975 the following examinations were carried out-.-
Stool examinations
742 of which 31 were normal
326 Hookworm positive
74 Ascarasis positive
SllAmOOba positive
Haemoglobin estimations
489 of which 157 less than 50%
205 50% to 60%
79 60% to 70%
48 more than 70%
Urine examinations
263 of which 244 were normal
11 were sugar positive
8 albumen positive
No details were available about drugs prescribed other than total annus
cost.
A comprehensive study of the appropriateness of drugs purchased and
prescribed would be too time consuming and, therefore, not feasible in this
type of evaluation.
The assessment of process is not just a matter of quantifying the
many health care activities performed.
The activities must be appropriate in
the circumstances and in an attempt to assess what can be termed the Qua1iuv
of Care 3 indicator cases were chosen for study.
-34-
Analysis of Quality of Care '
Mention has already been made of the unsatisfactory aspects of
this part of the study, in particular the very small numbers of patients
interviewed.
In addition, attempts to achieve good repeatability and
validity were inadequate, the former consisting of the maximum use of closed
ended questions with yes or no answers and the latter of checks on the answers
to questions (iii), (v) and (vi), by reference to records maintained in the PHCU.
Despite these shortcomings the findings have been included in this
report in order to demonstrate the criteria used for the assessment of ’quality*
and the fact that the necessary information was easily obtainable from patients
who had recently received medical care.
Thus this part of the development of an evaluation methodology has
been regarded as
as a
a Pilot study requiring further refinement and testing.
A summary of the data obtained from the questionnaire relating to
anaemia in pregnancy is set out below.
(i)
7 out of 10 patients had a history taken
(ii)
All the patients were examined and the doctor
used his stethoscope in all cases.
(Hi)
(iv)
(v)
(vi)
(vii)
(viii)
in all cases a blood .sample was taken for Hb estimation
The doctor informed the patient of the diagnosis in
only 1 case out of 10 (2 could net remember)
9 out of 10 knew correctly the drug regimen prescribed
8 out of 10 received advice about looking after themselves
4 being given dietary advice only, and 4 advice about
diet and taking rest
7 out of 10 reported definite improvement after treatment
no
with 1 reporting a little.improvement and 2 reporting
improvement at all.
The average time spent waiting to see the doctor was
25.5 mins;. (Range 20 to 35 mins.)
A further assessment of. process was carried out by studying the
Efficiency of the PHCU staff in performing their duties but as this involved
all 7 Health Functions the results and discussion are presented in a separate
section, p. 52.
Similarly Resources + Cost are presented separately
(see p.49&51).
OUTCOME
Since no follow up information about the patients was
available no assessment of Outcome was possible at this stage.
The Quality of
Care study assessed outcome of care as perceived by each patient but this cannot
be considered as a valid measure of Outcome if one adheres to the definition of
-35"
Outcome as an objective measure of the change in health status following treatment.
In order to conform to the constraints necessary for an evaluation of this kind
measurement of Outcome would have to take the form of ad hoc prospective studies
of selected groups of patients.
(•
Maternal & Child Health
The services provided under this category were as follows:~
Antenatal examinations
Assistance at deliveries by trained midwives
Postnatal visits
Health care of children aged under 5 years
Antenatal examinations
DEMAND
If it is accepted that all pregnant women should be
routinely examined at regular intervals then it is possible to estimate the
potential demand for antenatal examinations if the birth rate and the population
size are known.
For example, the Crude Birth Rate for the Mallur population
was calculated to be 31.6 births per 1000 population (see section on Family
Planning,p
42 )
Thus with a total population of 4124 one would expect
This information can
approximately 125 births each year in all 5 villages,
be used in the planning of the numbers of staff and piICU time required to
carry out the antenatal examinations.
However, actual demand measured in terms of the numbers of women
who attend for examination may be considerably less than potential demand.
Table 13 shows that only 36% of the 94 women whose births were recorded at Mallur
in 1975 ever attended for antenatal examinations.
In addition 70.6% of these 34 did not make their first visit until
after 28 weeks gestation.
Table
13
NUMBER OF PATIENTS
GESTATION PERIOD AT
FIRST VISIT_____
less than 12 wks
o
12
28 wks
10
28
32 wks
6
32
36 wks
12
36
40 wks
6
34
TOTAL
-DR
One reason for low demand may be that the local people do not
appreciate any need for antenatal care (see Table 8) and so a Health Education
programme may be required in- an attempt to alter ,such attitudes.
Another reason may be a failure to identify pregnant women in the
population so that they can be encouraged, to attend for examination.
The information in Table 13 indicates that this was a likely factor.
If this problem is to be overcome then it is vital that the field
in Mallur.
workers, expecially the females such as the Auxiliary Nurse Midwives (ANM),
should make enquiries about possible pregnancies as they visit each household.
Poor attendance at antenatal clinics by known pregnant women may
indicate that regular attendance is not possible due to other committments e.g.
in the home.
Consideration should then 'be given to planning domiciliary
examinations with visits to the clinic only for special reasons e.g. taking
blood samples for Hb tests and blood group determination.
PROCESS
The amount of information v/hich can reasonably be recorded
routinely under this heading will depend to a large extent on the demand
and the number of available staff.
In each of the PHCUs in the study
the quality of available data was poor.
In an attempt to improve this the Mallur PHCU had recently introduced
a record card for each woman which contained details of all antenatal examinations
as well as information about her subsequent delivery and post natal progress.
Future evaluation of process would then be possible by scrutiny of these cards
for details of examinations performed, results of tests e.g. Hb estimations,
and the outcome of any clinical intervention.
In circumstances where records could not feasibly be maintained
evaluation would have to rely on special studies such as that described under
Quality of Care.
OUTCOME
In the study lack of information and poor maintenance of records
made it impossible to make a valid measurement of Outcome.
For example,
although 34 women registered for antenatal care the standard of recording was
such that the numbers of subsequent attendances of each patient was unknown,
Had this information been available then assuming that regular attendance for
antenatal examinations is likely to be beneficial both to the mother and the
foetus, an appropriate measure of Outcome would have been the proportion of
women attending regularly for antenatal examination from an early stage of
pregnancy e.g. 16 weeks.
Assistance at deliveries
DEMAND
Once again potential demand can be calculated from population size
and knowledge of the local birth rate.
-37-
?rhe most accurate estimate can bo
made if the number of fertile couples is known i.e. whore the wives are in the
15 to 44 years age groups. and where the number of births is recorded routinely
and accurate.
Table 14 shows the numbers of deliveries recorded in each PHCU.
information helps in the planning of midwife requirements.
This
However, much more
accurate planning can be made if, as recommended in the previous section, the
field staff make early indentifications of pregnant women.
Planning can then
utilise knowledge of the estimated delivery date for each woman,
Furthermore,
information from antenatal examinations will identify those women likely to
need a hospital delivery e.g. abnormal foetal presentation, cephalo-pelvic
disproportion.
Table 14
Recordod Deliveries (1975)
J
B
M
JANUARY
5
18
50
FEBRUARY
9
13-
51
MARCH
5
6 (2)
38
APRIL
2
.18 (2)
33
MAY
8
15 (1)
42
JUNE
10
17 (4)
46
JULY
6
28 (6)
57
AUGUST
6
28 (1)
60
SEPTEMBER
14
41 (1)
60
OCTOBER
13
27
54
NOVEMBER
12
35
59
DECEMBER
4
24
56
TOTAL
94
270 (17)
606
( ) = stillbirths which were
recorded at PHCU (B) only
Total population M =4,124
B =11,875
J
PROCESS
15,402
Since the activities involved in a routine domiciliary delivery
by trained personnel tend to have limited scope for variation it was considered
that a suitable measurement of process would be the numbers of deliveries was
not available from PHCU records and so an estimate was obtained from the
Questionnaire Survey.(see Table 15).
-38-
Table
PERSONS CONDUCTING IAST DELIVERY
N -
N
N ~ 100
100
DR -
24
TRAINED = 51/--- ANM \
HOSPITAL
26
DR TRAINED
100
DR " 8
1
TRAINED :: 9
45 —--- ANN - 44
ANM = 1
1
UNTRAINED = 48 i.e.
relative
UNKNOWN
J
B
M
88
UNTRAINED “ 90---- rel
V* husband - 1
•no help •: 1
UNTRAINED - 55-- rel ~ 53
no help
2
1
Dai r. 1
ANM = Auxiliary Nurse. Midwife
DR = Doctor
OUTCOME
rel
relative
The ideal outcome of a delivery is a live, healthy baby and-
a healthy mother.
As a more practical objective for PHCUs in a developing
country this can be expressed as the prevention of stillbirths and maternal
deaths.
Therefore, in order to calculate reliable rates it is important
that all live and stillbirths and maternal deaths should be routinely recorded.
in the study only PHCU (B) recorded stillbirths.
The measure of Outcome which was used in the study was the proportion
of deliveries which were attended
by trained staff since this was assumed to
be beneficial to both mother and baby.
Table 15 shows that 51% of deliveries at Mallur were attended by
trained staff.
When tested statistically (by calculating the difference
between Proportion there was no significant difference between Mallur and
PHCU (B) but the difference between Mallur and PHCU (J) was highly significant
(p < 0.001).
To establish the reason for this would require deeper
investigation.
Table 16 below shows the percentage of deliveries attended by
qualified persons in the state (Karnataka) as a whole and in 2 neighbouring states.
Table 13
STATE
DELIVERIES ATTENDED
BY QUALIFIED PERSON (%)
KARNATAKA
15.83
KERALA
25.72
TAMIL NADU
21.92
J
-39-
TABLE 17
UNDER FIVES EXAMINATIONS 1975 (MALLUft)
MALLUR
MUTHER
KACHAHALLI
ALLUNAGANAHALLI
BHATRENHALLI
TOTAL
JANUARY
156
0
42
0
0
198
FEBRUARY
110
0
32
0
0
142
MARCH
108
0
0
0
0
108
SEPTEMBER
342
0
0
O
0
342
OCTOBER
398
0
0
0
0
398
NOVEMBER
406
0
0
0
0
406
DECEMBER
470
0
49
•21
41
581
TOTAL
1990
0
123
21
41
2175
APRIL
MAY
JUNE
JULY
AUGUST
NOTE: From April to September no ANM was available
Total No. of under 5s - 682
• •••
(From October 1975 census)
••••
>••••
bicix.i.uA
y
showed a significant difference.
vavn
vne rvsuiis
(p‘ < 0.001)
I
Postnatal visits
The purpose of the visits in each of the PHCUs was to
check on the progress of mother and baby in the first 2 weeks after birth,
The
quality of available data was very poor so that no assessment could be performed
and no results tabulated.
Health Care of under fives
The programme consists of 2 main parts i.e.:
(a)
Immunisation
(b)
Developmental screening with an emphasis on
detection of nutritional problems.
demand
This is arguably an inappropriate term since the services
are instigated by PHCU personnel who make domiciliary visits throughout
each village on a rota basis.
The term is retained for the sake of uniformity.
The potential demand for services is composed of the entire under
5’s population within the catchment area.
It can be measured by a census oi'
estimated from a sample survey as described at the beginning of this Chapter.
I
It is important that this initial estimate should be updated by the addition
of all new births.
This information can be used together with details of the geographical
distribution of the population to calculate the number of field workers required
to make regular domiciliary visits.
Where the number of staff is restricted
through lack of resources the information can be used to devise the most
rationa1 programme of visits which is possible with the few staff available.
PROCESS
In the study this was measured in teijms of the number of examinations
and the number of vaccinations performed during 1 year.
Table 17 shows the number of examinations performed in Mallur during
1975.
Apart from the absence of the field worker (ANN) for 5 months the
bias towards the headquarters village is striking,
During the 7 months when
examinations were performed the 379 under 5s in Mallur (from Table 6) were
each examined approximately every 5 weeks on average. This compares with an
examination approximately every 4 months on average for the remaining villages
except Muthur where no examinations were carried out.
The most likely reason for the above discrepancy is a poorly
organised visiting programme but poor record keeping raay have been another
possible factor.
Table 18 shows the numbers of vaccina.tions done in Mallur in 1975.
-40-
J-M
VACCINATIONS DONE JN 1975
Information gathered from Road to Health Charts
1°
SMALLPOX
MALLUR
MUT1IER
KACHAHALLI
40
0
0
0
0
40
0
0
0
0
0
O
40
0
0
0
0
40
148
0
3
1
4
156
129
0
8
1
0
138
30
0
0
1
0
30
307
0
11
2
4
324
232
0
30
1
3
266
169
0
22
1
0
192
1
O
0
0
0
1
402
0
52
2
3
469
Revacc.
Total .
DPT
1
0
2°
3°
- Total
-I0
'polio
2°
3°
Total
BHATRENIIALLI ALLUNGANAHALLI
TOTAL
The same problem of bias towards Mallur is highlighted.
OUTCOME
On the assumption that vaccination is of automatic benefit the
measure of Outcome employed in the study was the porportion of under 5s who
were protected by immunisation.
The information was obtained from the Questionnaire Survey and is
presented in Table 19.
TABLE 19
%• VACCINATION COVERAGE'XUNDER 5s)
M
N =
B
132
N
J
122
N = 132
No vaccination
3.8%
4.9%
6.1%
Smallpox only
29.0%
73.7%
90,0%
Smallpox + DPT
67.4%
1.2%
3.8%
BCG
0%
BCG+ Sma11pox
19.9%
Sma 1 Ipox-rBCG ’’-DPT
0.4%
-41-
BCG
0%
I
The results were tested statistically.
The proportions of under
5s who had received- smallpox vaccinations showed no significant differences
between the 3 PHCUs.
The higher DPT coverage at Mallur was significantly
different from the coverage at the other PHCUs.
(p < 0.001).
The fact that none of the children in the Mallur sample had
received BCG vaccination was a serious omission in view of India’s known high
incidence of TB infection.
Since one could not assume that clinical examination was automatically
beneficial per se it was not possible to use the proportion of under 5s regularly
examined as a valid measure of Outcome.
Family Planning
in developing countries where rapidly increasing populations are
consuming limited resources there is an obvious need for an effective Family
Planning programme,
Since the production of children is such a fundamental
part of human behaviour any attempt to bring about a change must be directed
on a personal level at individual couples.
in order to obtain the goodwill and cp-operation of the population a
Family Planning programme, planned and executed at a local level e.g. by PHCU
staff, should include an assessment of the attitudes and beliefs of the local
In the study the Questionnaire
people concerning the production of children,
Survey included a number of questions which attempted to obtain such information
ardnone was available at any■ of the PHCUs (see Section or Need,p- 29 )•
This
information would be useful in planning-a motivation campaign e.g. by identifying
those coupl.es sympathetic to Family Planning who could be further encouraged to
adopt a contraceptive method. •
DEMAND^
as measured by the numbers of people coming forward for contraceptive
supplies, sterilisation or even advice, is useful for the planning of resource
requirements i.e. staff and equipment, as described in the section on Medical
Relief, p 31.
Table 20 shows the demand for Tubectomy during 1975 at the Mallur PHCU.
No information was available about demand for contraceptive advice or supplies,
Since a Family Planning programme is heavily dependent on promotion
personal level by PHCU staff, the planning of a home visiting programme will
at a
be aided by knowledge of the potential demand for services.
This can be defined
crudely in terms of the number of women in the fertile age group together with
their respective spouses. An estimation of these numbers can be obtained from
knowledge of the numbers of the catchment population.
-42-
MALLUR FAMILY-PMNN1 NG 1975
TUBECTOMY ONLY i.e. Reference to PHC Sidlaghatta
ALUNAGANAHALLI biiatbenahalli
MALLUR
MUTHUR
kachaiialli
i JANUARY
1
0
0
0
0
1 FEBRUARY
3
O
0
0
0
( MARCH
1
0•
0
0
0
1 APRIL
’ MAY
2
0
0
0
0
2
0
0
0
0
I JUNE
2
0
0
0
0
I JULY
2
0
0
0
0
AUGUST
3
0
0
0
0
■ SEPTEMBER
5
0
0
0
0
OCTOBER
4
0
0
O
0
NOVEMBER
3
0
1
1
1
DECEMBER
4
0
0
0
0
TOTAL
32
0
1
1
1
For example, from Table 6 it was known that there were 988 males
and 932 females in the age group from 15 to 50 years in the Mallur population.
In the
The numbers of this group living in each village were also known,
other PHCUs estimation of the numbers could only be made from knowledge of
the total population and the proportions of males and females in each age group
Thus in PHCU (B)
calculated from the results of the Questionnaire Survey.
there were an estimated 2371 males and 2460 females in the 15 to 50 year age
group and in PIICU (J) 3314 males and 3573 females.
It was not possible to
estimate the numbers in each village.
of those
These highly theoretical estimates can be refined by taking account
factors likely to influence acceptance of contraceptive methods by the
population.
Local attitudes and beliefs have been discussed in the section
on Need.
Another important factor is the local Infant Mortality Rate since,
if the rate is high, people may wish to produce more children in order that some
will survive to adulthood.
This is related to another relevant factor i.e.
the number of live children currently within a family.
damaging for a
It may be potentially
Family Planning programme if parents with relatively few children
are approached by PHCU
staff and encouraged to adopt a contraception method
be aroused. For this
or undergo sterilisation since unnecessary resentment may
be advisable to set an arbitrary limit, of the number of live
reason it may
should be made by PHCU staf
children in a family, below which no routine approach
-4'?-
TABLE 21
Family Planning Practice-
TOTAL PRACTISING
N = 100
N = 100
100
N
J
B
M
20
i.e. 18 Tubectomy
TOTAL PRACTISING - 21
i. e.
7 Vasectomy
TOTAL PRACTISING = 13
11 Tubectomy
11 Tubectomy
2 Pill
2 I.U.D.
DON’T KNOW ~ 6
1 Vasectomy
i .e.
1 Rhythm
DON’T KNOW = 1
1 Condom
Age of all mothers
X = 27.0 years, S.D.jf 6.1
X - 28.7 years, SD I 6.5
X
27.7 years, SD X 5.9
Age of mothers practising F.P.
X
25.4 years, SD .+ 5.8
N = 20
X - 25.1 years, SD +5.9
X x 31.6 years, SD + 6.1
N -21.1 years
N
- 13
• Number of live children (all mothers
X
3.1, SD + 1.8
X
3.2, SD + 1.7
X = 3.0, SD + 1.8
Number of live children (acceptors
X = 3.0, SD + 1.7
X
3.1, SD + 1.9
X = 2.9, SD + 1.6
From Table 21 ail arbitrary limit of 2 live children would appear
to be appropriate since there vzere couples in each catchment area who wore
using contraceptive methods and yet had only 2 children.
PROCESS
The only information available in the study was that presented in
Table 20 i.e. the monthly referrals of tubectomy patients.
For planning purposes
it would have been useful to have details of contraceptives issued as well as
procedures carried out, such as vasectomies and tubectomies, since they have
implications for the purchase of supplies and deployment of staff.
The
Assessment of Efficiency study produced some data about staff deployment and
is discussed on p, 52.
OUTCOME
Given the many variables involved in carrying out a Family Planning
programme such as traditional attitudes, personal beliefs and efforts by
PHCU staff, a single measure of Outcome is extremely difficult to define.
This is especially true where information is either unavoidable or unreliable,
However, as a crude estimate of Outcome the study used the percentage
of couples practising family planning,
This was calculated from the results
of the Questionnaire Survey (See Table 21).
Statistical test showed that the
Outcome at’ Mallur was not significantly different from that at either PHCU (B)
or PHCU (J).
Another useful indicator of Outcome is the birth rate,
From the data
at Mallur it was possible to calculate the Crude Birth Rate since the catchment
population was known and from Table 14 it was known that there wore 94 recorded
births in 1975.
An important point was illustrated by the fact that when calculated
using the total population of -4124,
Crude birth rate = 22-. 8 per 1000 pop
n
However since it was noted that the village of Muthur was excluded from
the services then the more accurate figure is
Crude birth rate = 31.6 per 1000 pop
n
When compared statistically with the rural rates in Table 22 below
there are no significant differences except with the All India Rural Birth Rate
which is significantly higher (p< 0.01)
-44-
1 d U J *>
CRUDE BIRTH RATE - 1971
Rural
Urban
Birth Rate
Birth Rate
Kerala
31.3
29.6
Tamil Nadu
32.7
27.8
Karnataka
34.6
25.3
All India
38.9
30.1
State
Communicable Disease Control
This Function -refers to mass programmes aimed at the population
at a whole.
in the study area these programmes such as TB and smallpox
In
eradication were carried out by teams from the State Government--aacl^this lay
» I • l« >
outside the- limits of the study.
"^N
4GALO.U-3
BAXCA'
j
Environmental Sanitation
‘
\
Sanitary measures are an importan': tool in th6^control
th& C£®rtrol of the
*.•.
Az
communicable diseases which are water-borne and food-borne
may be the dramatic diarrhoeal diseases such as cholera and shigellosis
or the more insidious diseases such as typhoid, amoebiasis and infectious
hepatitis.
Unfortunately, in developing countries, with their high rates of
mortality and morbidity, the public demand is for curative rather than
An added problem in promoting sanitation is the likely
preventive measures.
clash with traditional methods of water and food storage, cooking techniques, waste
and sewage disposal.
However, examination of local morbidity rates is likely to demonstrate
the need for improved sanitation.
Thus, for example, Table 12 shows that in
Mallur diarrhoea and bowel infestations were responsible for 13.6% of all clinic
consultations in 1975, the second highest after cold and ’flu.
For reasons explained above the component of Demand is inapplicable
x
and any Sanitation programme must be actively promoted by PHCU staff.
In
the Mallur PHCU this programme consisted of routine chlorination of drinking
wells and the construction of sanitary latrines for each household.
No informatior
was available concerning sanitary measures promoted at the other PHCUs.
-45“
The total number of drinking wells is shown in Table
23% below.
Table 23
No. of drinking.wells
Village
Mallur
8
Muthur
3
Kachahalli
5
Bhatrenhalli
3
Allunaganahalli
2
TOTAL
21
L
The aim was to chlorinate each well once every 2 months.
The actual chlorination programme as carried out in 1975 is shown in
Table 24 below.
Table 24
E,7.SANITATION MALLUR 1975
NoWells Chlorinated 1975
MULLUR
MUTHUR
KACHAHALLI
allunganahalli
BHATRENAIIALLI
TOTAL
ANUARY
3
0
1
0
0
4
-ebruary
3
0
1
0
0
4
XRCH
3
0
1
0
0
4
PRIL
3
0
1
0
0
4
<X*Y
3
0
1
0
0
4
„UNE
3
0
1
0
0
’ULY
3
0
0
0
0
4
2
UGUST
0
0
0
0
0
0
SEPTEMBER
8
0
0
0
0
0
'CTOBER
8
0
0
0
0
0
NOVEMBER
0
0
0
0
0
0
OEGEMBER’
8
0
0
0
0
8
“4-3-
It is
clear that Mallur village once again received the bulk of the
services as has already been demonstrated.with the MCH and Family
Planning Services.
of wells chlorinated per
was measured in terms of percentage
Thus in Mallur village the percentage, coverage was an average
outcome
month.
The other
of 93.75%.
villages had no coverage except Kachahalli which
had 40% coverage for the first 6 months only.
Sanitary Latrines
(ii)
the number of latrines
No information was available on
However, information on the percentage of
constructed during 1975.
gathered from the Questionnaire
households having sanitary latrines was
Survey (See Table 25
below).
Table 25
HOUSEHOLDS HAVING A SANITARY LATRINE
J
B
M
100
N - 100
N
N - 100
11% have sanitary
latrines
1% households have sanitary
latrines
7% households have
sanitary latrines
(2% net known)
Mallur and PHCU (J)
When tested statistically the difference between
the difference between Mallur and PHCU (B) was
was not significant whilst
An adequate explanation of these results would not
significant (P
0.001).
probable important factor is that the
be possible without deeper study but a
majority of sanitary latrines in Mallur and PHCU (J) were in the main villages
which had good access to the modern city of Bangalore whilst access to all the
villages in the PHCU (B) catchment area was very poor.
Liaison
Liaison between PHCUs or with other
health organisations is unlikely
isolated PHCUs there may be no
to be a very important component and in some
However, where liaison exists an assessment should be made
liaison at all.
administration and future planning of health care delivery
of its implications for
had 3 important links with other health
The Mallur Health Co-op
institutions.
(a)
These were;-
S-tJohn’s Medical College .^ang^alore,
the Co-operative Committee and
Members of the college staff sat on
-47-
/
/
contributed expertise to the management and planning of health care delivery,
in addition newly graduated interns spent 6 week attachment’s at the PHCU and
contributed to the health care delivery.
Knowledge of the precise dates of
attachments were useful in planning special projects such as immunisation programme
Table 6 shows the results of a census conducted
or data collection programmes.
The medical officer was also able to plan occasional short
by the interns.
periods of leave whilst ensuring some continuity of care during his absence.
(b)
St. Martha ’ s Hospital, Bangalore.
Patients from member villages were afforded direct referral to the
hospital for more specialist treatment, where necessary.
In addition specialists
from the hospital attended periodic treatment ’camps’ at the PHCU and the
medical officer was able to draw up lists of suitable cases for treatment.
(c)
Primary Health Centre, Sidlaghatta
This government health centre was the referral unit for tubectomy
in the locality so that good liaison was necessary for the planning of referrals
from the Mallur PIICU.
Table 20 shows that 35 women were referred for tubectomy
during 1975.
The centre also stored certain special vaccines which the PHCU was
able to utilise in special circumstances.
Jn 1975 Mallur received 500 doses of
cholera vaccine for use in a suspected outbreak of disease.
Administration
In many, if not most, PHCUs administration forms part of the duties
of those staff members who are providing health care directly to the public.
Thus evaluation of all the preceding components will automatically include, to
a large extent, an evaluation of the administration of the PIICU in terms of the
quantity and quality of data routinely collected as well as decisions made on
the basis of this data.
In view of the above a separate, detailed evaluation of Administration
may be inappropriate.
The decision on whether or not to carry out a detailed
study will be influenced by the proportion of total cost and staff effort
expended on administration.
Table 27 shows that in Mallur in 1975 Administration accounted
for 6.1% of total expenditure.
This was considerably more than expenditure on
Environmental Sanitation and only marginally less than on Maternal and Child
Health at 6.9%.
’-48“
Table 30 shows that Administration accounted for approximately
21% of total staff effort, second only to Medical Relief.
This was some 7 times
greater than staff effort at PHCU (B) and some 3 tiroes higher than at PKCU (J).
On the basis of the above evidence further investigation of
Administration, would be warranted in order to discover the reason for such high
cost and effort.
An important factor was that non members of the co-operative who attendet
for care were charged for drugs and so in addition to such records as patient
registers, drug registers etc. a comprehensive record of accounts was maintained.
A full time clerk was employed to maintain the records.
RESOURCES
(i)
Manpower
In most, if not all, PHCUs in developing countries
limited resources will result in limited members of staff with varying levels
of training and capability.
Knowledge of the number of available staff and
their capabilities can be used to plan the best possible work programme,
This
is particularly important for field workers who have to carry out regular house
to house visits in the villages.
In order to plan effectively it is necessary to have good information
about the catchment area and its population.
Knowledge of village locations and
travelling distances from headquarters are necessary for calculating travelling
Details of transport facilities, if any, are also important.
times.
Information
about the numbers of houses in each village and the composition of each household
i.e.sex and age of inhabitants, is required so that a rough estimate can be
made of the time required to carry out visits in individual villages.
Table 26 shows the manpower resources in each of the 3 PHCUs in the
study.
Table 26
MALLUR
1 Medical Officer
BASETTIHALLI
1 Medical Officer
JANGAMKOTE
1 Medical Officer
1 Auxiliary Nurse Midwife 3 Auxiliary Nurse Midwives 3 Auxiliary Nurse Midwives
1 Compounder
1 Compounder
1 Compounder
1 Senior Health Inspector
1 Senior Health Inspector
1 Junior Health Inspector
2 Basic Health Workers
3 Helpers
3 Helpers
1 Clerk
1 Ayah
-49-
In comparison with Mallur the other PHCUs each had twice as many
staff members but they also had larger catchment populations which were
widely dispersed in small, often remote villages.
(see maps in Appendix).
A greater proportion of their workers was involved in field duties.
For
example, they each had 3 ANMS compared with one in Mallur.
No information was available to judge the success of the domiciliary
visiting programme or whether, as in Mallur (See Tables 17 and 18) the services
were concentrated in the main villages to the virtual exclusion of those in
peripheral positions.
(ii)
Equipment, and drugs
Little information was available at any
of the PHCUs so that no proper evaluation was possible.
A list of the equipment used by the Mallur PHCU is included in the
Appendix.
The major items were as follows
1 Motorcycle (for house visits by the Medical Officer)
1 Refrigerator (for storage of vaccines and heat labile drugs)
1 Microscope
1 UNICEF Midwifery Kit.
In view of the services offered by the PHCU these items were
appropriate.
The-only useful available information concerning drugs was the total
expenditure in the year 1975.
This is discussed in the following section on COST.
Information about types and quantities of drugs used was not readily
available and it was not possible to assess the planning of drug purchases.
A rough estimate of drug requirements could be made from knowledge
of attendances by diagnosis for Medical Relief services.
(see Tables 10 Sc 11).
This information would indicate the types of drugs required as well as giving
a rough estimate of quantities.
An investigation of drug prescribing and dispensing should also be
considered as a check on whether appropriate drugs were given in the proper
quantities.
For example, if a medical officer gave antibiotics routinely for
colds and ’flu a great waste of resources would ensue with little or no healtli
benefit.
If, as in Mallur, cold and ’flu ’formed the largest diagnostic group
(see Table 12) then the waste could be very serious.
-50-
COST
The resources expended in the delivery of health care, i.e.
staff effort, use of equipment and drugs., all cost money.
In developing
countrios, where the amount of money available for health care is severely
limited it is therefore important for an overall examination to be made of
total expenditure on health care delivery.
The costing method used in this study enabled a break down of
total expenditure to be made by Function.
By showing the order of magnitude
of services provided by Function the results acted as a rough measure of
performance by the system.
Table 27 presents an analysis of the total expenditure in the
Mallur Health Co-operative in 1975.
The original method called for the inclusion of the calculated
annual capital depreciation of 1%.
However, the PHCU building was rented from
a private landlord and the original capital expenditure was not known.
In
order to overcome this problem the capital expenditure was estimated arbitrarily
at Rs 20,000, and thus the capital depreciation was Rs 200.
The table shows that by far the largest proportion of expenditure
(91.8%) was on clinic based services with the remaining 8.2% on field services.
The most costly item was the drugs bill which accounted for 64% of total
expenditure and 98.4% of the total drugs bill was spent on Medical Relief.
However, Table 12 has shown that some 30% of O.P. attendances are
for the common cold and influenza whilst diarrhoea, bowel infestations accounted
Thus almost 50% of attendances were either due to ailments
for a further 16%.
where only mild symptomatic relief could be offered or where the illness itself
was preventable.
This evidence would warrant a special study into the
appropriateness of prescribing and dispensing.
Administrative costs (6.1% of total expenditure) followed closely
after MCH services as the third most expensive Function.
In terms of money it
cost Rs 3395 which was almost as much as the Rs 4000 annual drug allowance at
PHCUs (B) and (J).
The per capita expenditure on health care for the population of the
Co-operative member villages was Rs 13.50.
Table 28 shows that this was some
4 to 5 times greater than expenditure in the state as a whole as well as two
neighbouring states.
-51-
Table 2-7
Annua1 Clinic and Field Expenditures by Functions (Mallur 19751
FIELD
CLINIC
MR
MCH
FP
CDC
Env.
San
Admin
MR
MCH
FP
CDC
Env.
San .
Admin
112
13
0
0
0
75 •
0
129
0
0
1
0
Capital Costs
180
21
O
0
0
120
0
673
0
0
7
0
Maintenance Costs
34,815
0
0
0
0
0
160
25
0
0
550
0
Drugs
12,780
2231
0
0
20
3228
0
3069
0
0
0
0
Salaries
47,887
265
0
0
20
3423
160
3896
' 0
0
558
0
Total Expenditure
92.8
0.5
0
0
0.04
6.6
3.5
84.4
0
0
12.1
0
0.3
6.9
0
0
1.0
0
percentage
Distribution of
Clinic Expenditure
percentage
Distribution of
Field Expenditure
percentage
•Distribution of
clinic 8c Field
Expenditure
85.2
0.5
0
0
0.04
6.1
Total Expenditure - Rs.55,659
Catchment population - 4,124
• •••
• •••
• •••
• •••
Table 28
HEALTH EXPENDITURE PER CAPITA__
"
IN~3 S, INDIAN STATES
PER CAPITA
(RUPEES) .’
expenditure
STATE•
2.58
KARNATAKA
3.70
KERALA
3.03
TAMIL NADU
Further discussion on COST follows latei’ in this Chapter.
in the absence of suitable measures of Outcome an intuitive
reaction to the high cost of health care demonstrated above is that it is
probably not justified.
assessment of efficiency
Although this forms part of the evaluation of Process it has
been presented af the end because it gives an overall picture of staff
performance in. the delivery of health care, .It provides a more detailed
analysis of performance by individual staff members than was possible in the
analysis of Cost.
The results are presented in Table 29. The 10 days during which
this part of the study was carried out did not coincide with any unusual
occurrence such as an epidemic and thus the results can be regarded as
representative of staff effort.
Examination of the table shows that the efforts of the medical
limited fields of activity
officer and ANM in Mallut were restricted to very
compared with their counterpartsin PHCUs (B) and (J).
For example the
ANM at Mallur only participated in the MCH services while her counterpart in
PHCU (B) was involved in MCH services, Family Planning, Communicable Disease
Control and Environmental Sanitation.
The ANM at PHCU (J) was involved in all
of these with the exception of Communicable Disease Control.
The medical officer at Mallur devoted 99.9% of his time to activities
within the clinic building.
87.3% of his time was spent in Medical Relief
activities and 7.1% on administration. Previous evidence (see Table 12)
has shown that some 30% of attendances for Medical Relief were diagnosed as
common cold or influenza.
The next largest group (16%) was composed of
-52-
Table 29
ASSESSMENT OF EFFICIENCY: STAFF EFFORT (% TIME) BY FUNCTION
)
MED OFFICER
COMPOUNDER
MCH
MR
STAFF CATEGORY
CDC
FP
ENV. SAN
ADMIN
PERSONAL
TRAVEL
OTHER
M
87.3
0
0
0
0.1
7.1
4.4
0
1.1
B
60.5
16.5
1.7
0
0
10.3
2.6
1.2
7.2
J
46.5
0
2.7
0
0
8.1
30.6
12.3
O
M
65.5
0
0
0
0
10.8
3.6
0
20.1
J
34.6
0.5
0
0
0
11.9
33.3
19.2
0.5
M
0
74.3
0
0
0
0
O
25.7
0
B
0
42.3
23.7
2.0
2.1
0
3.9
25.7
O
J
O
26.9
19.0
0
19.7
10.0
14.6
18.3
O
M
31.32
14.32
0
0
0.04
21.172
3.38
13.46
16.32
B
17.86
16.98
16.52
9.60
12.62
3.05
3.90
17.34
2.13
J
19.93
6.60
11.85
0
11.25
7.57
23.42
19.11
0.25
B
AMM
TOTAL EFFORT
Average total hours
worked per day
M
B
J
• ••••
• •••
• ••••
CLINIC
FIELD
5.4
9.0
8.1
5.1
7.2
8.4
I
preventable disease, i.e. diarrhoea, bowel infestations and vitamin
deficiency.
This would tend to suggest that'both as a doctor and as manager
of the local health services more effort should be devoted to preventive
medicine in the community and less to Medical Relief Services.
On examining Total Staff Effort Table 29 highlights the high
proportion of effort in Mallur expended on Administration (21.2%) compared
with PHCU (B) (3.05%)and PHCU (J) (7.6%).
This confirms the findings of
the analysis of Cost and provides further evidence of the need for a study
aimed at finding ways of reducing Administrative Cost and effort at Mallur.
Although not directly related to Mallur it is worth noting the
high proportion of Personal time (23.4%) available to the staff at PHCU (J).
Personal time refers to time available to staff members when not engaged in
activities related to health care delivery,
The probable reason was low
demand for services by the local population.
This is supported by data in
Table 9 which shows that PHCU (J) had the lowest O.P. attendances for Medical
Relief despite having the largest catchment population.
Resources required to carq out evaluation
The importance has been stressed of keeping co a minimum the
In order
resources used to carry out an evaluation of health care delivery.
to Obtain the information, part of which is presented and discussed in this
chapter, the following resources of time and manpower were required:~
(a)
Time
(i)
The Questionnaire Survey required 10 man days
for each PHCU.
(ii)
The Quality of Care study required 2 man days
for completion of each of the 3 questionnaires.
(iii)
The Assessment of Efficiency study required 10 man
days for each PHCU.
(iv)
Examination of clinic records required 4 man days for
each PHCU.
(b)
Manpower
The study required 2 interviewers to carry out the
Questionnaire Survey and Quality of Care Study.
Three
Observers were employed to carry out the Assessment of
Efficiency study and 1 person carried out the examination
of all PHCU records.
-53- ,
RECOMMENDATIONS.
An important role of a 'first time’ evaluation is to
identify defects in the previous system of routine data collection.
Correction of these defects will enable any subsequent evaluation to be
more effective.
Although the choice of data to be collected should lie
with the local PHCU staff, as a result of this study the following are
recommended:-
Population data
Information about the catchment population is extremely
important for the purposes of evaluation, management and planning.
is recommendedf therefore that any PHCU about to introduce routine
It
evaluation should carry out a comprehensive population survey, perhaps
as part of an initial evaluation.
Although it is likely to be time
consuming the survey should include details of the geographical location
of all villages, the number of households in each village and the names,
ages and sex of all inhabitants.
Thereafter the following are recommended for routine
recording:(i)
All births and stillbirths including parents' names
and address.
(ii)
All deaths including name and address of deceased and
suspected cause of death.
(iii)
Medical Relief
All outpatient attendances should be
recorded by. age, sex and diagnosis.
If possible,
the patient's home village should be recorded as an
aid to the initiation of control measures in the event
of an epidemic.
(iv)
Matema1 and Child Health
(a)
All pregnant women should be registered by name
and address.
It is recommended that each
pregnancy should additionally be registered on an
individual Antenatal card which should contain
name, address, age and estimated delivery date.
Details of all subsequent examinations, tests and
treatment should be recorded on these cards.
(b)
All deliveries should be recorded as in (i) and
any maternal deaths.
(c)
All children under 5 years old should be
registered by name, address and date of birth.
This can be updated from (b) above.
It is
recommended that each child should have an individua.
record card containing details of immunisation,
routine examinations and any medical treatment.
(V)
Family Planning
All consultations about family planning
should be registered by name and address for follow-up
purposes.
Details of contraceptive methods or
sterilisation procedures should be included.
(Vi)
Environmental Sanitation
In sanitation programmes which
include chlorination of drinking wells, a register
should be maintained stating the location of each well
and the dates of chlorination.
However, this is only
one example since sanitation programmes will vary in
content according to local conditions and priorities.
-SS-
CONCLUSIONS
The results and discussion have demonstrated that the proposed
methods of evaluation, despite necessary resource constraints, are capable of
generating information which can be used for the improvement of management and
planning of health care delivery.
However, the merits of the Quality of Care study remain unproven
and further testing of this particular method is required to properly
validate the questionnaire and test repeatability.
Interpretation of the results of an evaluation must be made with
caution because of the necessary reliance on data collected by PHCU staff
which may contain errors and omissions..
In an attempt to improve the quality of routinely collected data
recommendations have been made concerning categories of information
which should be recorded.
However, it must be stressed that the final decision on the .
nature of data to be collected should lie with those responsible for the
day to day management of a PHCU, i.e. , the staff and in particular,
the medical officer.
The role of an evaluation system is to provide
information which may simplify decision making (27) and by implication
must be relevant to local circumstances.- The danger of making a
standard list of data recommended for routine collection is that
evaluation may degenerate into a system of processing inf-rmation vzhether
or not it is relevant to local needs.
This is a rapid means of
condemning evaluation to a state of uselessness.
For these reasons this study has proceeded on the
assumption that the process of evaluation is best carried out on a routine
basis by the local PHCU staff, This does not mean that the introduction
of ’self evaluation’ is likely to be a simple process,
On the contrary,
there is the major problem of lack of local expertise in carrying out an
evaluation study as well as probably the most difficult problem of lacx
of understanding of the basic concept of evaluation.
In order to overcome these problems, the next step should be
the development of evaluation systems in selected PHCUs under the
guidance of experienced evaluators who can both instruct the local stafx
and monitor the progress of systems so that specific organisational
problems can be identified.
When such problems and their implications
have been carefully examined, then a strategy can‘be drawn up for a
widespread introduction of routine evaluation as an integral part of
primary health care delivery systems.
-56-
REFERENCES
1. ABEL-SMITH. B.(1976)
Value for Money in I Tea1th Services; a
comparative study. Heinemann, London.
2.
ACHESON, H.W.K. '(1975)
Medical Audit and General Practice.
Lancet, March 1, pp 511-513
3-
ALDERSON, M.R. (1974)
Evaluation of Health Information Systems
Br. Med. Bull., Vol.30, No.3
« 4-
ALDERSON, M.R. (1973)
Objectives and Concept of Health infomatj
Systems
Proceedings of WHO Conference on Health
Information Systems, Copenhagen, 18“22 Jun
WHO Geneva.
5.
ALEXANDER, C.A. et al
Cost Accounting of Health Centre Expenditu
Ind.J. Med.Res., vol GO, No.12.'
6.
ASHFORD, J.R. (1975)
How Can quantitative Methods help the
Health Services Mana ger?
in McLachlan, G. (Ed)
Measuring for Management. Quantitative
Methods in Health Services Management.
Nuffield Provincial Hospitals Trust.
7.
ASHFORD, J.R. and RILEY, V.C.
1975.
An Approach to Monitoring the Quality of
Health Care.
In McLachlan, G. (Ed.)
Measuring for Management. Quantitative
Methods in Health Services Managfmenr ._
Nuffield Provincial Hospitals Trust.
8.
BABSON, J.H. (1973)
Disea so Cost ing
9.
BABSON, J.H. (1971)
Hospital Costing in Great Britain
The Hospital,
4, pp 106-111.
10.
RARKER, K. (1970) et al
(Eds).
Health in the Developing World
Cornell Univ.Press, Ithaca, New York.
11.
BIOLOGICAL SCIENCES
COMMUNICATION PROJECT (1973)
Delivery of Health Care in less developed
countries* with exphasis on integration
of family planning with mother and child
health.
George Washington Univ.
Washington D.C.
12.
BLUM, H.L. (1974)
Planning for Health. Development and
Application of Social Change Theory.
Human Sciences Press. New York.
13. BRADSHAW, J.
(1972)
Manchester Univ. Press.
A Taxonomy of Social Need, In Problems
and Progress in Medical Care, No. 7
Nuffield Provincial Hospitals Trust.
14. BRYANT, J. (1975)
Health and the Developing World
Cornell Univ. Press.
15. BURDETTE, J.A. et al (1974)
Primary Care Evaluation. The AAFP
Collaborative Study.
JAMA, vol. 230, No.12.
16. COCHRANE, A.L. (1972)
Effectiveness and Effici ency Nuffield
Provincial Hospitals Trust, London.
UNC
17.
CVJETANOVIC, B and GRAB, B.
(1976)
Uecermination of the Cost Benefit
Balance Point of Sanitation Programmes.
Bulletin of W.H.O., Vol. 54, Part 2,
pp 207-215.
18
DJUKANOVIC v and mACH, E.P.
Alternative approaches to meeting basic
heal th n eed/i
developing cqurrtjJjys
joint UNICEF/WHO Study. WHO, Geneva.
(Eds) (1975)
19.
DONABEDIAN, A (1969)
Evaluating the Quality of Health
Care in Programme Evaluation in the
Health Field, SCHULBERG, H.C. et al
(Eds.) Behavioural Publications,
New York
20. FELDSTEIN, M.S., PIOT, M.A.
and SUNDARESAN, T.K. (1973)
Resource Allocation Model for Public
Health Planning. A Case Study of
Tuberculosis Control Supplement to
Bulletin of W.H.O. Vol.48. WHO. Geneva.
21. FORD FOUNDATION (1976)
The development of health services
in Bangladesh Ford Foundation, New York
22. GRIFFITHS, A.
(1976)
Health Planning, Management and Training
Needs in Underdeveloped Countries
Unpublished.
23. GUPTA, -M.C. et al (1977)
Effeet of Periodic Dcworming on
Nutritional Status of Ascaris-infested
Preschool Children Receiving Supplementary
The Lancet, July 16, pp. 108-110.
24. HARRISON, P. (1977)
Basic He a 1th Delivery in the Third World
New Scientist, Feb.17. pp. 41-43
25.
IDRISS, A.A. et al (1976)
Sudan; National health programme and
primary health care, 1977/78 - 83/84.
Bull. World Health Organ, vol.53, part 4,
pp 461-471.
26.
L4ST, J.M. (1965)
Evaluation of Medical Care.
Med. J. of Australia, November, pp 781-785
27.
LITSIOS, S. (1976)
Developing a Cost and Outcome Evaluation
System.
Int.J. of Health Services, Vol 6, No.2
28. LOGAN, R.F.L. et al (1972)
Dynamics of Medical Care.
Memoir No. 14. LSHTM.
29.
RIASON, A.M.S. et al (1973)
Disease Costing jn Hospitals.
A review of completed work. Unpublished.
30.
MORLEY, D. (1974)
National Nutritional Planning
BMJ, October 12, pp 85-88.
31.
MORLEY,D (1975)
paediatric Priorities in the Developing
World.
Butterworths. London.
32. McDOWELL, I. -and MARTINI ,
C.J.M.(L976)
Prob1ems and New Directions in the
Evaluation of Primary Care
'Int. j. Of Epidem., Vol.5, No.3 pp.247-25’
33.
M01AC’!LAN, G, (Ed. )(1970
A Question of QuaJity. Roads to Quality
As surai ice in Mo< lie al Care.
Oxford Univ. Press for Nuffield
provincial Hospitals Trust.
34.
NEWELL, K.W. (Ed) (1975)
Health by the People WHO, Geneva.
35./
PARKER, R.L. et al (1972)
Relating Health Services to Community
Health Needs.
Ind. J. of Med. Res., Vol.60, No.12.
36.
PIACHAUD, D & WEDDELL, J.M.
(1972)
Cost of Treating Varicose Veins
Lancet 2, pp 1191-1193, Dec.2.
. 37.
PIACHAUD, D & WEDDELL, J.M.
(1972)
The Economics of Treating Varicose
Veins. int.J. Of Epidem., .1,3,
pp 287-294.
38.
POLE, J.D. (1973)
39.
POPOV, G.A. (1971)
The Use of Outcome Measures in Health
Service Planning.
Int. J. of Epidem., Vol. 2, No.l
Principles of health planning in the
U, S. S .R . Pub,.Health Paper No. 43.
WHO, Geneva.
40. RAMALINGASWAMI, P and
RAMALINGASWAMI, V. (1973)
In Health Service Prospects; an
International Survey.
Ed. by DOUGLAS-WILSON, I. and McLACHLAN,G.
The Lancet and Nuffield Provincial
Hospitals Trust.
41. RAO, P.S.S. et al (1972)
Methods of Evaluating Health Centres.
Brit; J. Prev.Soc.Med., No. 26, pp 46~52.
42. REINKE, W.A. et al (1974)
Functional Analysis of Health Needs and
Services.
Johns. Hopkins Univ., Baltimore.
Unpublished.
REINKE, W.A. (1972)(Ed.)
tative aspects
quantitatiye techniques. Johns Hopkins
Univ., Baltimore.
43.
*
r
44. ROEMER, MX (1972)
Evaluation of Community Health Centres
Public Health Paper, No. 48.
WHO, Geneva.
45< ROEMER, M.I. (1976)
Rural Health Caro.
St.louis.
46.
SANAZARO, P.J. (1974)
47.
SANJIVI, K.S. (1971)
48.
SCOTTISH HOME AND HEALTH
DEPARTMENT AND WHO JOINT TEAM
(1974)
49.
SHAPIRO, S. (1967)
The C.V. Mosby Company
Medical Audits
BMJ, February 16, pp 271.-274.
Planning India's Health Orient Longman,
Bombay.
The Child Health Services.
Planning Approach.
WHO, Geneva.
A Systenatic
End Result Measurement of Quality of
Medical Care.
Millbank Memorial Fund Quarterly,
Vol. XLV, No.2.
Position: 1206 (4 views)