HEALTH SYSTEMS RESEARCH
Item
- Title
- HEALTH SYSTEMS RESEARCH
- extracted text
-
I
RF_PH_22_SUDHA
Scaling Criteria
You will now use the criteria discussed in this session to select one of the problems to be studied as the first
priority.
i
I
Use the criterion scales below to help you decide which of the problems should be studied as the first priority.
Enter these ratings in the table on page 17.
Relevance:
1.
2.
3.
Not relevant
Relevant
Very relevant
r
20
i
Avoidance of Duplication:
1.
3.
Related studies available
Related studies available but major issues not covered
No related studies available
Feasibility:
1.
2.
3.
Not feasible considering available resources
Feasible considering available resources
Very feasible considering available resources
Political Acceptability:
1.
2.
Not acceptable
Acceptable
Applicability:
1.
2.
3.
No chance of findings being implemented
Some chance of findings being implemented.
Good chances of findings being implemented.
Cost-Effectiveness: (Judge whether the difference the results can make to existing programs are
substantial in view of the cost of the study)
1.
2.
3.
Cost too high in view of expected effect of study
Cost justified in view of expected effect of study
Cost low in view of expected effect of study
Timeliness:
1.
2.
3.
Not timely
Timely
Very timely
Ethical Considerations:
1.
2.
I
Not ethical
Ethical
21
1
Avm
OPERATIONS RESEARCH IN HEALTH CARE
I
¥.PBGUPTAS
1.
INTRgpyCTintll
Operations Research or OR in its abbreviated form,
is
the name given to an omnibus,
if
sprawling,
activity
initially designed to apply science in the service of war i n
entirely novel way.
OR.,
OR,.
hswever, is a different kind of
sci ent i -fie work in that it adopts the scientific method of
exhaustive
investigation
and experimentati
on,
not
experimentation
to
inventing new armaments, but to improving the utilization of
existing ones.
An early example of this deliberate turning to science
to increase the effectiveness of existing weapons occurred
during the World War II and concerned how to set the time
fuse of
a bomb to be dropped from an aircraft on to
submarine.
The coastal command of the Royal Air Force had
decided that the fuse should be set to explode at a depth of
1000 feet below the water surface,
on the plausible
expectation that the submarine would sight the approach!ng
aircraft about 2 minutes before the instant of
attack
and
thus dive that far below the water surface by the time the
encounter with the bomb would take place,
But as the
results of the offensive were disappointing, the problem was
handed over to a team of
scientists since nick named
Blackett’s Circus.
After exhaustive field observations,
Blackett’s team found that in actual combat,
the bomber had
only a small
chance of aiming the bomb right if
the
submarine dived as promptly as the theory under1yi ng the
existing practice assumed.
It,
therefore ,
followed that
when the aim could be true,
the bomb exploded too far below
the surface to affect the submarine.
On the other hand,
when the aim was necessarilx poor,
it exploded at the right
depth but at the wrong location.
It is obvious that in such
a situation,
it would be better to set the fuse to denote
almost on impact instead of 1000 ft. below water, because it
proc
i^ p,
on such occasions,
when the submarine has not
yet fully dived,
that there is any chance of seeing the
submarine and aiming true.
The result of the subsequent
change
(based on this) doubled the nt
ni mber of submarine
ki 1 Is.
Thereb are plenty of other examples of
successful
war
time use,
in operations,
of the concept of this scientific
study of |processes as they take place not in the laboratory
but in the field hence te name Operations Research.
i
$ Associate Professor (Operations Research),
National Institute of Health & Family Welfare,
New Mehrauli Road, Munirka, New Delhi-110067.
1
!
Fol lowing the end of war, the success of military teams
attracted the attention of industrial managers who also were
seeking solution to their complex executive problems.
Since
he subject has been gaining more and more importance
1950,
today it has found application
in most of
the
and
such as management,
public administration,
disciplines
heal th ? engineer!rig etc.
2.
T
Nature
research techniques
OR is many things to many people
to diagnose and correct operational problems,
pilot project
to demonstrate new approaches,
evaluation of
crongoing
programmes,
experimental
efforts
to
test
different
activities etc.etc.
By Operations Research,
however,
is
however,
meant any formalized quantitative analysis (now researchers
include all types of analysis) whose purpose is to improve
the efficiency in a given situation where
’efficiency’
is
clearly difined.
Efficiency may be interms of minimiuzation
of costs or maximization of benefits.
A distinctive feature of operational problems, however,
is that the experimenter outside the laboratory is unable to
actual1y manipulate and control his environment.
As a
result,
Operations research could also be said to be an
application
of
sci enti tic
scientific
method
to
artificial
representati ons
of
real
life
situations,
These
representations
ei ther
are
either
in the form
of
models
(mathematical or diagrammatic) to be studied analytically or
simulation
of
actual
experience
hypotheti cal
under
condi ti ons.
In either case,
the relevant
i np* t vari ables are
related to operational ac levements or outputs.
In the
heal th -Field,
the inputs may include the total
heal th
budget,
its geog; aphical and functional
breakdown,
i nf1ow
and outflow of personnel;,
. _"“ti ty and quality of training;
inflow,
storage and consumption of drugs and supplies;
distribution
of
all
types of
servi
ces
health
services
and
institutions etc.
Outputs or operational achievements may
include such things as clinic attendence,
vacci nations
performed,
number of child births assisted,
drugs and
contracepti ves distributed !>
wel 1 s cleaned and other such
act i vi ti es.
Typically,
some
of
the
inputs
are
i ncontrol1able whereas others are subject to control.
In
emergency care services,
the administrator has little
control
over patient needs but can control,
to a
1 arge
extent, the staff input to the service process.
In
the mathematical
models,
the input
- output
relati onships
are
made expli ci t and
an
appropriate
analytical technique such as differential calculus or linear
programming
is applied in order to determine the optimum
relationship in some sense.
2
x
YTovlIegj
OR vs Systems Analysis
The
dt-f-f-ei^enee-between—OR- -and-.-Sy.s±eais.^ana.L?/5.is in that
-the----- ]reeeF^--i--e---u-suial 1 y concerned wi th prob 1 ems i n
wj^i ch
the
lEtefeHf in deciding
what ought toj ■ixeeTi'dDnjfe^rn'Xin^l:
si m^Ty' How “To]”do~'i t most efficiently.
'— ---------- j--------------r1 —
. — --fn,r | - y*, T -
- _---T
ddlW
T. II . !|
- ■ n - ---
•+
-f -+ T.T
iuorfdlw
_uiv
i
I
j
wi th
Anr 7/si s and OR have to de^l
" .• deoxlqmc
^r^ctid
:idal|
problems" ot
olif ji choice or decisions but with a
--- ---------- ---------------------- £
■dTrf^erencei— Systems ]A«nalysis (research) is concerned mainly
$5rh
The
---- —4rhe--e^rategy-..-ofJ^.chiai.ce. whereas OR with______ tactics.
jtfrdbi;ein^?’^&“^ more ccrmpl etel^vrspkejci f ied in 0R< than inr Systems
TlBlUDSH
£■
ripid 'ioi
YTseiun
UDI Isd-sn-ooK
sdnsTni
3?
u___ )
gonerav----------
‘ ‘__
aspecjt
Isxosqpg model is a simplified representation
of' some
s^jE^tHeT-rjeal world,
isometimes of an object'7
, ’'7' some tithed of 'a
■e-f—-a
sitTiatTon
or a processt:iBHCltvd>J purposeful rdd-u-c-tton
so
Q£ .^forrnat^n.
pageableapd
shape and
IS
; -QdF-- ..
i f—ial n-i mnnF-f-
+- r*i
riatf-il i r*
^oflPf
rtrtl •» /—toirfT-
’
;:.p^.nU’£ A
±in
-"- ■ away
the non-essentials ~ t
thaV^Triultf^
fW(f a
must
strip
problem
to
expose
the structural relationship
among
the
important
variables,
so
that
the
consequences
o-F
of
a
particular choice may be predicted.
Ty^es of Models
i
modefs^ 'A roaql map is a fami liar example of
such modelsThe traditional*- map tries to capture the
essential features of -a road network to assist us in getting
from one place.to another.
It regards as npn-essential such
details as which route has the best restaurants/r although
t"h Jt; mi girt be given high p r i or i t y in guide for - a^ t our 1st --. >
' 3 ".—i |
■ f-? :. -n ■■■
I
Ej™.
^r®:- Particularly valuable sort of
dlagramatic
.model, especially in situations where some commodity or some
portion of the population passes at a regular rate of flow
p-from one ..condi t ion uto. another.
For example,
flow charts
^indicating the processes by which crude oil becomes gasoline
H. or some other petroleum product.
A flow ctiart may be useful
i n describing various stages of a process as shown in Fig.l
r f or optimum new borrr. care;—
I
•’
I
:
it i: osqc
[
'
“ ZZE~Z
'■^1
Conceptual models are often used to
(i i) Conceptual model:
of
make non
non quanti+iea
quantified predictions about the behaviour
individuals or situations. One of the most famous conceptual
models of
of tn®
of recent years is Garrett Hardin’s model
commons,
the common grazing ground of
medieval
English
village (Hardin, 1968).
Cattleowners holding the rights to pasture their animals on
like car owners who have the right to use a
the commons,
ignore the cost that their own use
common expressway,
The
inevitable result is overgrazing or
imposes on others,
As Hardin observed,
overcongest!on that is costly to all..
arises
in
countless
contexts that
the problem of the commons c.
or
cars.
have nothing to do with cows
such terms as feedback or
On a more complex level,
processes
are
often
used
as
conceptual
models of
contag ion
•fami 1 i ar
a
widely
understood..
A
thermostat
is
that are
example of a feed back mechanism.
The models are classified as:
Petermi.ni.sti.c , when the parameters are constant or vary
in a predictable manner or the output is proportionate to
i nput-
Stochastic,
when their magnitude depends partly on
chance or the model takes account of variability, e.g. rates
of
arrivals at clinics,
hospital
admission rates, queue
formation etc.
Alternatively the models may discribed as:
4
the parameters
Static. ' when
the
magnitude' .of
t h g?
. Static.
independent of each other, and does not vary wit'w time.
are
* Dynamic when- this is not sb;; such models take account
in which the
changing circumstances and the manner
of
affected
by
these
circumstances.
deci sions will affect or be
■Final 1 yji
according t.o purpose they serve.
the
models
are said to be;
(a)
(b)
(c)
4.
descriptive;
predictive; or
prescri pti ve
PHASES of operations
research
Churchman, et al_ (1957) as well as. Hiller and Lieberman
(1972) discuss in detail the six phases of an O.R.
problem
or project y n ame1y:
Formulating the problem
(i )
Constructing a mathematical model to represent the
(ii )
system under study
(i i i ) Deriving a solution from the model
Testing the model and the solution derived from it
(iv)
Establishing controls over the solution
(v)
Putting the solutions to workz implementation
(vi )
Here we shall primarily concentrate on phases <i> to (m)
This is in no
and discuss them briefly through an example.
way meant to minimize the importance of the other necessary
project but our emphasis is on the
phases of every O.R.
used
to
derive
an optional solution for the given
methods
problem
(i )
be
the
E
Firstly the problem must
Fqcm.y.lat.i.Qn. of
the
f r m u 1 a t. e d in &n appropriate form clearly stating
variables involved and the constraints on them-
<ii> Construct.ing. a mathematicAL model. : The model should include
three basic elements.
a)
Decision variables and parameters:
Decision variables are those unknowns that are to be
determined from the solution of the problem whereas
the
parametersf are given uncontrol 1able variables of
model«
b)
Constraints or restrictions:
Constraints
system.
account
for physical limitations
5
of
the
c)
Objective function:
•
Objective function defines'the measure of effectiveness
of the system as a -function of its decision variables.
soluti_9.nL
' ( i i i ) Der i_Y.i.na the solut.
ion:_ Generally solution to a model means
those values of the decision variables that optimize one of
performance
the • Objectives and give permissible levels of
on any other, of the objectives.
The various methods and
discussed
techniques used for obtaining solutions are
sep ar at el y.
gfiamelg 1
h ■
In a family planning programme,
suppose we want to
max i mi ze
through
a
the
number of
births prevented
The
contraceptive mix
of
IUD and oral
contraceptives.
objective is to find the ideal contraceptive mix in order to
max i mize. the births prevented,
The information available
regarding contraceptives,
at-cpti.
manpower
and
effectiveness,
financial resources is given as under:
(i )
Four woman years of IUD protection or 2 woman years of
oral (contraception are required to prevent one birth.
(ii )
ami 1 y
The -f
family
planning clinic has only 4 workers
each
aivai.lafole for 2000 hours per year and they irequi re
three hours of activity for each oral contraceptor and
two hours for each IUD user.
(iii) The programme has a total budget of Rs. 9,00,000.
Each
woman yeaur of IUD protection costs Rs.
Rs«
135/— whereas
each woman year of oral protection costs Rs- 720/-
Decision variables are X
X
and ¥ (say) where
number cf oral contraceptors
¥ = number of IUD users
Parameters are (i)
(i ) the activity (in hours) required for each
oral
contraceptor/IUD user and (ii) the cost of each woman
year of oral protect!on/IUD protect! on.
Constraints are
3X + 2 Y < 8000
720X + 135 Y < 9,00,000
Objective function i s
2
0-5X + 0.25Y
Tht? problem is to determine X and V subject to
constraints such that Z is maximum.
6
the
above
5-.
t^IbapOLOGY* AND AREAS-OF APPLICATION
OR does not however,
necessarily require the use of
special techniquesdoes not exclude other management and
planning procedures.
With this quantification,
OR can be
said to be the use of the scientific approach in the study of
□perations.
The special’techniques it employs include:
i)
ii )
iii)
i v)
v)
Linear Programming and other optimization procedures
Applications of queuing theory
NetworIk Anal, ysis
Computer Simulation Procedures
Inventory Control and forecasting
It also employs statistical analysis,
projection and
ex t r ap ol at i on methods (regression analysis),
Budgetory and
al1ocating
methods
(Cost
effectiveness
analysis
in
parti cular)
workstudi eSy Decision theory and game theory-
Here we shall discuss the most commonly used techniques
such as Regression analysis.
Linear Programming,
Queuing
theory,
cost-benefit and cost-effectiveness analysis and
inventory control
through their applications in widely
different problem areas such
as ideal
mix
of
health
programmes,
hospital
size and location,
waiting time
problems in
outpatient departments and for hospital beds,
manpower planning,
drug inventory levels etc.
Some of the
other techniques such as F’ERT/CPM.
si mulation,
simulation,
replacement
etc. are also described briefly.
The techniques as well as
some of the applications illustrated in this secti
section
on have
been taken from Grundy et al. (1973), Reinke et al (1972) and
Bailey (1964).
However it may be pointed out that the
description of both,
problems and techniques given here is
highly simplified and artificial-.
5- 1
Regression Analysis
It can be used when a suitable relationship can be
established between dependent and independent variables.
It can also be used for
forcasting,
studying the effect of
one or a group of
variables on another such as IMR,
life
expectati on
utilization of health facilities etc.
V
9
“7
9
‘Example
Let us now see how regression model can be
simulate the effects of a pcii cy.
D
~a
(
H , N
9*
W
y
L
o
used
to
D
is
)
Here D denotes the conditional mortality rate
cor.di tinned on morbidity as a function of s
i.e.
Hs
Vector of
curative health
heal th measures including
doctor per capita,
hospital
beds per capita,
primary health care and related instituions etc.
M:
Index of nutritional s tat u.s
Ws
Accessibility to potable water
Ls
Literacy rate
V;
Stochastic disturbances
W and L have been specified in the above relation on the
independent
grounds
that they must probably have an
Thus,
for
example
influence upon mortality rates.
eKcmipic,
nutritional
status
individuates
not only
only affects
affects the
the
status not
susceptibility to disease but it also affects his ability to
overal 1
survive the disease once he has caught.
The
The above
mortality is AD where A denotes morbidity rates,
to
relationship could be expressed by a regression equation
the
study
the effect of changes in one or more of
independent variables.
In the above model,
the government determines H, W and
It may therefore use the model to calculate the effects
DC,. mor ality of different policies with regard to H ,
W and
Most p r ob ab 1 y,
the most important determinent of N is
income uer cap! a which is outside the direct control
of
However, in so
as the government is able to
influence the rate of economic growth,
its policies have
jncirect implications for N.
•L ,■
•— z-r'i
If
the goals of economic development include reduction
in death rates,
the model may be used at the macro level to
determine the optimum resource allocation that minimizes the
r umber of deathsThe major problem in use of regression analysis is the
: - hoi ce of
independent variables which are not correlated
£...Tong themselves.
This fact restricts the choice of
independent variables.
To overcome this problem,
methods
such as discriminant analysis are used.
8
5.2
Linear PrpgrMWsing
Linear programming (LP) is a technique for allocating
resources when the supplies of these resources are strictly
limited
It can be a valuable aid for making policy choices
that range from allocating the. budget for a sg.uI I library to
selecting the components for a gigantic hydroelectric plant.
A linear programming model of the entire Ganges -Brahmaputra
ixya.
system takes into account flood, control,
power
generation, irrigation,
i rrigati on„ navigation etc. (Rogers, 1969).
Li near programming is a technique of optimization i e u
it
is concerned with choosing the best levels for various
acti vi titles in situations where these activities compete
for scare resources,
or with choosing the minimum-cost
method
of
producing
results.
Every linear
required
pr og r amm i ng formulation requires
requires the specification of an
objective function,? . whether in
i n terms of
benefits to be
max i m i z ed or costs to be minimized and constraints.
Thi s
wi 11 be illustrated by the following simple exampless
Examole 1 iOB.ti.mal.
mix. I
A Health officer is charged with the task of minimizing
the number of days of disability among children in his
district.
There are 160,OOO children in the district.
The
State Health Department has chosen two programmes,
namely
n u t r i t i on an d i mmun i z a t i on services,,
for
f or mi
mi ni m
mii z i n g t h •?
disability days.
The Health officer
of
20
of f i cer has
h as a staff
staf f
workers to provide these services. Each of these workers are
avai1 able for 2,000 hours of work in a yearv
year«. Nutrition
Nutrition ser—
vices require say, 0.216 hours of staff time per child per
year.
Rs.. 30, 00, 000 have bee?i budgetted for the di st r i ct far
the two programmes for the next year,
year.
Jmmunizatior■ cnst
R's.22/per child per yea. and nutrition ser<-ir?s Rs- 30/-'
per child per year.
For each child coverrdc b‘■ immuni zati a;"
programme 5 days cf disability are avoided where
co/er ago
•
nut’-ition programme avoids 6 days of disability.
The problem is to find the best mix at
chi Id.--on
t □ be
covered under the two programmes so as to avoid the maximum
number of disability days.
Let
X
number" of children covered under i mmun i zat ion programme
Y
number of children covered under nutrition programme
then objective function is
Z =
5X+6Y
which gives the number of disability days avoided
and is to be maximized.
The constraints are
0.5 X + .216Y
22 x + 30 Y a
40OOO
30,00,000
9
i
the solution i.e.
values of X and Y,
For obtaining
i
method
could
be
used.
In
complex
problems
simplex
i
ari ables (to be dertermined) and
i nvolving large number of variables
use of compute s could be made
(Gil let, 1976).
contrai ts.
it
One could also find out using computers that how much
would be worth if we could relax one of the constraints.
However , one could obtain graphic solution to the model
in case of two variables as given hereunder (Fig-2).
1
Graph the first constraint as if it were an
i - e. graph 0.5 X + 0.216 Y = 40,000
equi1i ty,
Step 2
satisfy the first
Determine the set of points that
This can be done
constraint
0.5 X ■+• 0.216 Y <_ 40,000.
easily by checking to see if the origin (o,o) satisfies the
constraints.
in this case,
0-5 sc O + 0.216 x O
O < 40,000
So all
points below the line 0.5 X + 0.216 Y = 40,000
satisfy the first constraint (note that the origin satisfies
all 1ess-than-or-equal to constraints if the right side bf
the constraint is non—negative>
Step 3
Graph the second and final constraint as if it were an
Let
22 X + 30 Y < 30,00,000.
equali tyu
Step 4
Determine the set of points that satisfy the second
constaint 22 X + 30 Y <. 30,00,000.
Again the origin
satisfies the constraint; so all points below the line 22 X
30 Y ^30,00,000 must also satisfy th-' constraint.
Step §
Determine the region where X > O
10
and Y 2s 0
I
Step■&
7
Take the intersection ot the regions • determined, in Steps 2.,
4 and-5.:
Thife gives the region o4 -feasible solutions.
That is,
each -point in this region satisfies all of the *constraints and
is a .candidate for ’ provi ding the fraximum benefit.
But the
contains an infinity of feasible solutions.
The problem
is. how to reduce this nunitjer to a manageable finite number.
2.03
L
b tdixve 2
5z 135,185180.
160-.
N
k P
' O
140-
W
E
.
R
C
120--
0
N
V:
k
S
T
lOO^t
? 0.5X + 0.216Y = 40,000
80
2 2x4-3 OY = 30, 00,000
»
60,500
60
40
20
10
20
30
40
f
J..140
50
(In Thousands)
X = 55633
Y = 60500
11
1,36,363
x
Step 7
The region of
feasible solutions is shown
ih Fig.
2 by
Fig-2
shaded ines.
With a little thought one could observe that the
objective function Z = 5X +• 6Y will take on its maximum value at
one of the corners denoted by 0 9
A,
B and C.
This in fact i s
the thumb rule that optimal
solutions lies at on& of the corner
poi nc-=>.
Step §
For obtaining the set of
correspond to the corner point
equations of constraints, namely,
values of
X and Y... which
sol ve the two equali ty
G..S X -a- 0.216 ¥
40,OOO
22X + 30 ¥
= .30,00,000
This gives me
is
the values of X = 55633 and Y = 60500 which
to
the
corner
Bcorresponding
Now one can easily
verify that
■
objective
oi3
jective function Z = 5x +
4- 6y is maximum corresponding to
x-^.j.864
and y = 60500 and the maximum number of disability days
avoided
are 632320.
For any other
x
and
632320other value
value of
ot
y
corresponding to corner points A and C,. the value of Z will
be
less than this number.
E'iamB.Le. 2 (Facility size and location and . uti 1 i zat i on )
One iis often faced with the problem of deciding as to how
many
beds
s should be there in a hospital
in"
di fferent
1ocati ons.
A rule of thumb suggests that accommodation to this
variability will be satisfactory if the number of hospital
beds
provided exceeds the average.
Hence if there were an average
r 200 beds, capacity should be
demand
200 + 3
x/200
= 242 beds.
Then,
if
two hospitals are operated to serve an average demand
of 200 beds each
they should each contain 242 beds or 484 in
all .
In contrast a single unit would require only
400 + 3/400
= 460 beds.
Moreover,
say.
one-fourth of the admissions- could be
scheduled and therefore removed from the
influence of
chance
variation , total capacity could be reduced to
1OO +
--OO
3 t/300) = 452 beds. However, in order to match the
number of beds to their demand
so that waiting time is minimum
and also beds r~
are utilized to maximum capacity/,
one could use
queuing theory techniques discussed in subsequent section.
(
• 12 .
r
Another ‘important <aspect* of“ facility
"
planning is that
of
referral relationship between scatteredJ clinics such as PHCs and
sub centres and district hospitals,
Suppose that 5 PHCs are to
refer the cases to three such hospitals,,
the referral beds and
hospital
capacities for acceptance of referrals are as shown in
table 1.
Certain referral patterns £.-"e likely to be more costly
than ' others
t ‘
due to factors of transportaion and inconvinience.
To the extent that the cost differentials
------- i can be quantified and
are
constant,
transportation
technique (a form of
linear
programming ) can be very useful.
TABI...E
1
Data for assessing PHC — Hospital
Di stri ct/
Sub District
hospitals
relati onshi ps
PHCs
A
B
C
D
E
Hosp i tal
Capaci ty
Unit Cost - PHCs to Hospitals
X
10
20
5
D
10
90
Y
2
10
8
30
6
40
Z
1
20
7
10
4
80
30
50
40
60
30
210
PHC referrals
In the illustration summarized in table
1„
the opti mal
1,
arrangement can ibe obtained by use of transportation technique
as displayed in table 2.
label
2
Optimal solution for PHC referrals based on data of Table 1
To Hospital
From PHCs
A
B
X
Y
Z
Hospi tai
uti1ization
C
D
40
50
E
90
40
30
40
10
10
30
80
The actual
technique used is similar to one illustrated
in
example 1
but the important element is that
it provides an
economi cal
basis for allocation of resources in the
face of
vari ous limiting constraints which
in this case is the hospitai
capacity.
13
5-2-1
froblems and limitations of linear programming
(i )
For many public issues, it will be more difficult to define
ard weight the objecives than was the
.he case
case in
in examples
discussed here.
The choice of an objective
is
objective is by
by no means
a CLlt and dried affair.
It may require the
affair,
the kind of
explicit thinking that a linear programming
approach
programmi ng
entorces.
“
(ii ) The restrictions that certain variables
of them) may not be negative.
(and frequently
all
(iii)Some of the relationships may be non-linear.
Some of the
variables rmay take
‘ '
only integral values.
These situations
can be handled with
-- other
---- programming techniques available.
(iv) Constraints may be
I
such that no feasible
acceptable scores on the objective function.solution yields
In that case
one fpossibility is merely to do the
best we can
with the
onerous
set of
constraints or <alternatively we can
go
back and see if the original problem
i can be respecified.
Finally,
the linear pprogramming approach forces us to
find out and formulate what
-- we are trying to accomplish,
what ways there are to go about,
how effective these
activities are relative to one another.
. and how the choices
among them are limited?
5.3
Queuing Theory
Di stributi ng medical
care as wi dely and efficiently as
possible
requires research
i nto functi ons and designs of
hospi ta... ~ and other medi c .1
institutions so as to meet the
requi rements of both patients and staff
in an optimal way.
Similar to situati □ns in industry, sue!
1
problems
of organisati on
frequently entai1
same . kind zf waiting line‘ or queue.
And s^,
so.
the theory of
queuing process is of considerable practical
importance.
There are usually three ma in aspects of the system
which may be called
(i)
the input process
(ii) the queue
discipline and (i i i ) the service mechanism.
( i > Input proc ess11 i s simply a mathematical
description of
the
way the customers arrive,
In many cases, this can be taken to be
a Poisson distribution wi th some suitable parameter.
1Q Fqi_sson d^istri_bution, the
obabi 1 i ty that a patient
the pr
probability
will
arrive at the clinic at a given moment of time remains
constant
i.e. arrival at 9.10 A. M. is equally likely as at 8.30
The
etc"
can
&
accidents, cne
the demand
demand for
for hospital
hospital beds
can
distribution
l"^9ar'ded
being approximately o* this
kind.
The
—i kind.
r«ii:nk„;:n.""P1't'Iy d»“dibsd once th. average arrival
14
<ii>
Queue disci.gil.LneL
d isclE.l.i.ne
It refers to the way in wh i c h a wa i t i n g
line is formed,
maintained and dealt
with.
The simplest
new
arrangerent is the “first some’
,
“-first served’
rule;
arrivals join at the end of the queue and the customers at the
head of the queue are served in order of arrival.
the
It de
deals
output end
end of
iii)
Ser yi.ce m
L "2.T!. L
It
a 1 s wwith
i t !"i tthe
h e output
It is specified in sufficient detail once wa know
whole process,
The
the number of servers and the distribution of service time.
distribution of
service time is usually taken to be negative
exponential distribution.
1" E^Dnential. dL®.Lr Lfeu t Lon ? the probability that a patient
finish his
who entered the doctor’s room 20 minutes ago will
consultation during the next-minute is exactly the same as that
of a patient who has been with the doctor for only 10
1.0 minutes.
The distribution is completely determined
once the average
service rate (R) is known.
The ratio A/R should never be greater
than one otherwise the queue size will increase indefinitely;
patients will arrive faster than they can be accommodated.
In health service operations, we normally consider steady
state
system
(i.e.,
as time becomes large)
because this
si mpli f i es the calculations to a great extent.
In such a
system,
we are mainly interested in five characteristics:
(i )
the mean number of waiting individuals, (ii)
the mean number in
the system either waiting or in the process of being served,
(i i i)
the mean time a patient waits to be served after his
arri val,
(iv)the mean time a patient
spends in the clinic
i ncl tiding both waiting and service time and (v) the proportion
of
time the clinic staff is idle because of no patients waiting
to
be
served.
All
these
characteristics are completely
dependent on the mean serveice time and the ratio (A/R).
k (Single server queue) »
Const der a 24-hour functing of a non-appointment paediatrics
clinic with one doctor only,
to attend the patients.
Let us
assume that on an average, 10 minutes of service time is required
per patient,
The operating characteristics for various patients
loads are given in the table 3.
15
TABLE 3
Steady-state oEe.rati.ng characteristics of
a Baediatric clinic;!
Mean
i nterv'al
between
arri vals
(60/A)
(min.)
Mean
Service
(A/R)
time
(60/A)
(min.)
Patient oriented
characteristics
Staff oriented
characteristics
Mean time
(min)
Mean no. of
patients in
clini c
Wai t ing
Total
Wai ting
Total
Proportion of
time doctor is
idle (X).
40. 00
IO
3
13
0. 1
0. 3
75
20. 00
IO
10
20
0.5
1.0
50
13.3
IO
30
40
2.2
3.0
25
11.1
IO
90
100
8. 1
9.0
10
10.5
IO
190
200
18.0
19.0
5
10.5
IO
990
1000
98.0
99.0
1
&<amQle 2 lMu.Ltip_l.e
server
gueuesT^
, ______
How rmany inpatient
beds
should be provided in any particular speciality in a hospital
I so
that the waiting for admission
—i is minimum?
This problem is of particular j
importance when there is
acute shortage of hospital accommodation .
— .4
It is clear that if
accommodation is inadequate;,
patients will
have to wait for
admi ssi on.
If it is lavish, the demand will rarely outstrip
the
supply but overal1 cost may be unduly high.
In order to achieve
a certain balance between
these
problem can be formulated in terms two extreme situations, the
of queuing process so that
the length of waiting list can be (
-rC—■.7~ exPressed as a function of the
number of beds.
The mathematical
solution of this problem has
been presented by Bailey (1964).
Example 3
server
hospital Manpower Planning)^
How 1
‘ - - is
1 arge a <staff
required to give adequate
.
J service from a
hospital messeriger unit?
16
The •functions of hospital messenger unit are assumed
patients., specimens and reports, and
to be transoortation of
mi seel1aneous objects in response to requests from any section
of
the hospital,
hospital.
Whenever a call is received, the dispatcher
sends a messenger.
messenger,
if
one is available, to provide service,
The service facility is shown in Fig.3. (Queuing system: One
1 ine, multiple server in parallel).
Figure 3
1
i
I
I
i
i
i
i
I
I
i
i
1
i
7
'I
71
7
I
• i
i.
r
I
i
ii
i
i
L
i
i
i
I
i____
17
i
i
Messengers Ml,
M2,
M3,
etc.
may be viewed as separate
servi ce faci1ities.
and incoming calls are served by any free
messenger
If no r~
messenger is free calls must wait and may thus
accumulate,
forming a queue.
The service is first-come firstserved basis.
The problem can be solved using queuing
theory .
For details one may refer to Gupta et al (1971).
5.4
Inventory Control and Forecasting
An inventory can be defined as a stock of goods which
is held for the purpose of future production or sales.
Inventory Control is the process of deciding what and how
much of
vari ous items are to be kept in stock.
It also
determi nes the time and quantity of various items to be
procured.
The Ibasic objective of inventory control is to
reduce the i nvestment
-------- in inventories and ensuring that
production process does not suffer at the same time.
The
objectives of inventory control are generallys
(i )
(ii )
to reduce financial investment in inventories,
to facilitate product i onoperati ons & services i.e by
keep i ng balanced inventories of materials and spare
parts of
components so that all
required
i terns of
stores will be available for providinq service at the
right time?!!1
(iii ) to avoid losses from inventory obsolescence, and
(i v)
to improve customer/patient servi ces,
To attain these objectives
(a)
(b)
(c)
(d)
5.^. 1
the inventory must:
determine items to br stocked,
aet-ermine when and how much to replenish,
keep suitable records, and
weed out obsolete items
gconomic order Quantity
lsUEEli.es inventories).
In order to decide how much of a particular drug or
an
item of store to order at a t i me,
one must recognize two
conflicting factors;
one
is the ordering cost associated
wi th the paper work and transportation
of material and the
other
is storage cost and blocking of working capitalL
that
might be employed profitably elsewhere when large quantities
are ordered inorder to reduce the ordering cost.
The
cost.
variation of the two costs according to ordered quantity
is
ahown in Fig.
4.
The luf
vt? aepiccmg
curve
depicting their
their sum
sum is al so
shown
in figure 4,
which reveal that there is an
i deal
compromise of the two conflicting factors at the order
quantity level designated as GU.
18
Figure 4
M
Boo
z—? Annual c.f> RRy in g.
Costs
2.5’0
Zoo| 50
\o>o
50
Annual. a&DePjNO)
costs
■
cl 1 o & AO
30 4-0
-4—
-7P
^*0
OrelcY' QuOLnli tvj (<6^
Economic order quantity: Supplies inventories
In
practice
Gt need not be determined
graphically.
Mathematical 1y ( by the use of calculus) !• Q* is given by
=
J 2SU/I
Where Q$= miminum cost order quantity;
S = fixed cost of supplying an or er;
U ~ annual usage of the item in question (in units);
Lin its); and
I — Cost of carrying one unit of the item in inventory
for one year.
In general ?
the higher the ordering cost,
the hi gher
the order quantity
and the higher the carrying cost ,
the
lower the order Quantity.
Suppose that two similar drugs ,
A and
are each
B,
priced at Rs.1000/- per box of 1000 tablets, wi th 1000 boxes
of A beinq
being used per year compared toi 250 boxes of
B.
Experience has shown that placement of an order costs Rs.
160/- and the iannual carrying cost after delivery i s 207.
of the purchase price.
Hence
2S/I
2(160)
/ 200
1-6 and Qt
For _thi s,,
the minimum cost order
_/fl.6 X 1000) or 40. For B, the Qb
US/1 =
quantity of A (Qa ) is
is
(T;6 X 250) or 20-
19
•tr
1
1-6 U
Given the minimum cost order quant i ties',
an order for
40 boxes of A should be placed on 25 occasions during
duri ng the
year and for 20 boxes of B on 12.5 occasions during
duri ng the
year.
If
on 25 occasions during the year,
an order is
placed for 40 boxes of Aj,
stocks wi.ll vary between 0 and 40
boxes. averaging. 20.
The ordering p*us carrying costs for A
□e
23 (160) + 20 (200)
7000
Corresponding costs for B will be (12.5)(160)
per year.
5(200)= 3000.
Since A and B are similar drugs,
standardization on a single product C.
may
we
Where C is a replacement -For both A and B,
□c = <yaT6 x (1000+250)) =
/(T.6 x
1250)
considec
Thus
.72000
44. 7
Annual
cost of ordering and storing this single item woul d
amount to 28(160) + 22.3 (200) = Rs. 8940 compared to the
total of Rs.10,000 calculated for A and B above.
5.4.. 2
Techni.gues of inventorY control
It is well
wel 1
known that every organisation consumes
several items of stores.
As al 1 the items are not of equal
importance, alhigh degree of control on each item is neither
practical nor worthwhile.
Therefore, it becomes necessary to
--- ---- classify items in groups depending upon their
importance.
thei r
Such
i classification
is known as the
the principle
of
principle
selective control as applied to inventories,
There are -Five
techniques of selective control:
a)
anatysi s.t
classes say A,
value i.e.
It separates inventory items into three
B and C in descending order of
usage
ot
A
i terns small
i n number but
in
account
f or major
portion of total consumption i n terms of value
need higher degree of control.
B =
items of medium importance lying between A and C
C
i terns which are large in number but account •for
only a very small portion of the total consumption
in terms of value — donot call1 -For strict control .
20
b)
This analysis is done to consider the
VED Analysis:
vitality of
an item and its effect on production and
foi
used
special 1y
is
other
servicesIt
st
ores
of
items
classification of drug -and other
denoting their essentiality of stocking.
stands for vital items without which production or
service would come to a halt.
E
D =
c)
is for essential items without which temporary
services
production or dislocation o-f
1osses of
oc c ur s denotes desirable items — all other items whi ch
are necessary but donot cause any immediate 1 oss
in production or service.
S.DE alia 1 .ysis.f Thi s analysis is
availability position of an item,
based
upon
the
i terns
S
specially imparted
refers to scarce items ,
and those which are in short supply.
D =
which are available in
refers to difficult items,
indi genecus market but cannot be procured easily.
For example, items which have to come from far off
are
cities
or for which reliable suppliers
difficult to find.
E =
refers to items which are easily avai 1 abl e;
(most 1 y 1 c»cal items) .
d)
is
(per
un i t)
cost per i tem
The
analxsisi
HML
are
i terns
considered
•f or this analysi s and all
(H)
Medium Cost (M) and Low
? ,5
classi f i ed a'- High Cost mi
r■■ —, j— 4..
L ?
f or
items.
This type of analysis is useful
1evel,
keeping control over consumption at department
physi cal
frequency
o-f
the
and
for
deciding
veri fi cati on.
e)
fsn
of
H e r e t h e q u a n t i t y and rates
a_n§.-LY.s-Ls.:_
consumption are analysed to classify the items as
(N)
s1ow moving (S) and Non-moving
Fast moving
(F) ,
iterns.
21
5■-3Fgrecasti
Forecasting is nothing but projecting the likely demand
future.
This is not only important but also necessary
4-' QFproper planning.
For example,’
planning for facilities
to be provided in hospitals,
formulation of
population
stabi1i zati on policy
(based on population projections),
projecti ng the demand for various stores in a hospital
during the next year,
demand for hospital beds in each
speci ali ty etc.
There are various methods for forecasting
depending upon specific situations,
The most commonly used
method of forecasting is using the straight line i.e.
by
fitting the straight line Y a+bX using the time series data.
in
5.5
Cost Benefi_t and Cost-Effectiveness Analysis
The technique
recnnique of cost
cost-benefit
by relating
—benefit analysis,
benefits of
helps the deci si on
a programme to its costs,
maker in deciding which of the alternative programmes should
be
g i ven
priority and if a particular programme
is
undertaken, to what extent the programme will be benef i ci al„
Cost-benefit
analysis has been used
in
two
senses,
Narrowly,
it
is a method of aggregating all costs and all
benefits associated with a given project,
programme.
or
decision in monetary terms, converting them to present value
and combining them in a single index,
such as the present
value of net benefits.
In the broader sense,
cost-benefit
analysis is an activity which investigates the cost and
benefits that are associated with a project,
programme or
deci sion.
It may also deal with the distribution of costs
and benef=ts and the sensitivity of results to di fferent
contingencies.
In
some
situations,,
the purpose of
government
expend!ture is specific ai.g we
wei1 x u. idarstood „
yet benefits
and costs are hard to r
compare directly because some or all
the benefits which accrue due to certain action
cannot be
expressed in monetary terms.
In such cases often Cost
efieciveness analysis is used in which alternate ve systems
are investigated to determines
i)
which of them is least costly when the alternatives are
equally effective in achieving the objective 5* and
ii )
which
of the alternatives is most effective
in
achieving the given objective when they are equal 1y
costly..
Occassional1y
an extension of
cost-effectiveness
analysis is useful for investigating budgetary allocation
for a public purpose,
In that case,one attempts to measure
benefits and costs in different units, detailing the maximum
benefits
that
can be achieved for each
amount
of
expenditure,
leaving the final choice of amount to higher
level decision makers.
2
/
Problems in analysing Inguts and gutguts
5.5. 1
The concept of input (or cost) and output (or benefit)
provides
an extremely
framework for organ!si ng
extremely useful
pertinent facts
f acts a^d relationships in dealing with policy
prchiefs hut there are various . problems associated wi th
inputs and outputs or both...
i)
What tangible costs are associated with actions
Costss
regarding personnel,
material,
capital
expenditure?
What other kinds of costs are involved -■ spillover
of
costs,
goodwill cost,
community disruption,
cost
suffer!ngs?
In what time period these costs fall? Who
will pay them?
the
However,
given the importance of such questions;,
For
i nformati on.
problem is from where to get such
in
example, if a highway project cuts a neighbourhood
What
half,
di eruption?
what is the social cost of
alternati ve type of compensation - and in what amounts
- would leave people no worse off?
Many costs, including many of those which are most
i mportant9 are not computable in financial terms or are
not even quantifiable.
But they are real
costs and
may be susceptible to logical analysis.
ii)
The sec and prob1em i s speci f ying the relat ionship
between
inputs of
resources and outputs.
These
relationships are production functions which def i ne
alternative courses of action available.
ill) The Lhird problem is defining outputs™
i v)
What outputs should be produced?
Benefi t—cost analysis is a framework for keeping
our
thinking straight in evaluating projects,
a framework
that
demands explicit attention to determining the impacts of
a
proposal
and assigning values to these impacts.
It is no
more than a tool which provides information
helpful
in
deci si ons.
taking decisions.
Cost-effectiveness analysis i s 1 argely
used for evaluation of alternatives such as:
Clinic vs Camp approach
Sales
bonuses vs mobile c ommun i c at i on
team
f or
increasing
contraceptive sales in c ommun i t y based
di stri buti on prograrnme.
Alternative strategies for delivery of health services.
23
5.6
Beg lac emen t.
So far,
we have been dealing with individuals who fall
sick,
treated
as outpatient or are hospitali zed and
discharged
in due course or time,
The cr worse o-f
this
iis
s one in which pieces of equipment operate
situatio
satisfactori1y for a time and then fail and must be replaced
or repaired. The pr iblems of replacement and maintenance are
of considerable interes’.
hospitals havt not only
sophisticated equipments but also imported ones.
Some of
them may be vital and if they are down even for
a short
duration,
it can cause a lot of misery to the patients,
In
such cases it is important to determine what number of
spares of vital equipment/imported equipment will be needed
(depending upon its failure rate) over a certain period of
time,
what type of maintenance and replacement policies
should be followed,
etc.
It is obvious that failure of a
component will depend upon its life.
In some situations,
failure of a particular item may lead to failure of
the
machine whereas in some cases it may lead to inefficient
operation of the whole system.
Replacement policies could
also be considered for major components or assemblies which
need not be replaced by a new one on a failure but could be
repaired again and again.
Similar situation may be faced with vehicles and
i n the hospital whether in a district or a big
ambulances in
city.
The
replacement policies generally
constdered
(depending upon particular situations) ’vary considerably but
some of the simple policies discussed byj Cox
(1962),
are
listed as under:
5.7
a)
Service replacements only (replacement on -failure o-f
component policy)B
b)
Planned
replacements at fixed intervals of
t i me
irrespective of the age of the component then in use
and service replacement on failure.
c)
Make
service replacement
■failure
on failure
or
planned
replacement as soon as the age of the component reaches
certain age,
whichever
(age
is earlier
(aqe
replacement
policy).
d)
Replacements are made as in policy(b) except that if a
failure occurs during
an
iinterval
nterval
around
planned
replacement point,
the <system is left idle till
the
planned replacement point is
reached
— ------ 1 (policy i nvolving
idle time).
a
Sequencing and Coordination 1PERI/CPM).
Sequencing is nothing but selection of an appropriate
order
in which waiting customers are to
be served.
But
projects or jobs consisting of tasks require that they must
24
be performed
in a specified sequence.
These problems
involve determining how much effort should be put in to the
performance of each task and when to schedule
it so as
PERT
performance.
to optimise measures of overall project
CPM
and
Evaluati on
and Revi ew Technique)
(Programme
SUCh
of
Path Method)
are used for solution
(Criti cal
Sequentng problems can be represented by a matrix
problems.
as shorn helofegs
Figure 5
Customers
or Patients
(or jobs)
- -Serv ice—fsc il’irt ies----S1
S2
S3
Ready to
start
time
S-h
J1
J2
J
n
Number of facilities in the
sequence through which
the job. must pass
Process
time
25
Time
due
The <sequentng
--problems can arise even
if only one service
facility is involved,
For ex amp1e, if each of the Jobs has
a due date and an ~-associated cost of delay,
minimization of
the total
cost of delay
/ may not be easy to obtai n.
Such
problems
are quite
common,
f or examp
1
e,
ex
amp
1
n
Limb
er
of
emergencies waitinq for a doctor, i
number of calls waiting i n
the
tho
limit."for a oess
miger etc..
5.
1
FSI > cm
r'tfiT anc CPM both are network
techniques for jplanning
and scheduling work.
They i nvolve three main stages:
i)
Breaking down the
into a set of individual
and arranging them project
into a logical.sequence/network. Jobs
ii)
Esti mating the duration r“ ‘
and r esour ce r equi r emen t s of
each Job" deducing a schedul
control the completion of the e and finding which Jobs
project, and
i i ). Re-al 1 ocating money or other
resources to improve the
schedule
PE&T was1 developed by U.S.
Navy in 1958 for F_1
Pal aris Missil e
Project whereas CPM was :
independently
developed i n 1957 by
Waiker ot
of the
the
Integrated
-d Engineering Control
group of
American Chemical Fi rm,
<
and
kelley of Remington Corporati on
in search for r
new ways of
“5 scheduling new projects.
Both
rJiese techniques define
--- » and coordinate• various activities
of
a Pr°jetL:nd —accomplish
These techniques have been discussed the objective on time.
in detail by Sri nath
(1?73), and wiest at al (1974).
••J n
7-2
fi^WR^ri son of PERT itCd CPM
PERT was d eve 1 op e d f o r
and has been used mostly i n
research and deveI op men t
types
of
programmes which are
relatively new and not much
i
nformat!
on
is available.
leads
to
This
uncercainty
in
cal cll at i ng
timings
for
vari o«s
x vt.i es-,..
On the other hand, CPM is applied i
to repetitive type of
projects
where activities are standardized
and
properties known.
thei r
ihey utilize more or less a
t ec h n o 1 og y,
standard
Changes occur ma inly in size,
shapes and
amangements rather -Ban in design f-concepts.
allow (•-It
does not
uncartainties
-f '
. .
■ — in time
estimates
-J and uses (only one time
c?st i mate
(deterministic).
Moreover3
times are
related to costs,
costs.
According to CPMj, in CPM,
most Jobs can be
reduced in duration if extra
resources are assigned to them.
The cost of
getting
job done may increase j. but i f
other advantages outwei gh'
this
added cost, the Job should
be expedited or crashed.
i
On
the
other
hand j, if there is no
reason to shorten
a particular Job - if it has
a generous
26
f
amount of
si ack
then the job should be done at its normal
pace with a lesser assignment of resources.
Only the
cri ti cal
Jobs
need to be expedited.
CFM attempts to solve
problems such as
which Job
is to be expedited and by how
much.
5.7.3
Advantages of network techniques
11 f C'F' c: es a thorough pre-planning of the task.
b)
It increases coordination.
c)
It identifies trouble spots !•
points responsibility.
d)
It
focusses
management7 s
attention
onto
those
activities iwhich are,
or
are likely to be,
i
in
diff i culties 9
rather than on to activities which are
fdLiier
progressing smoothly and hence need no attention.
e)
It indicates optimum start and finish times for
activity in an operation.
f)
It enables the plan to be revised in the best
suit changed circumstances-
g>
It suggests where alternative methods should be sought.
h)
CPM enables the manager to reasonably estimate total
project costs for various completion dates,
These
vari ous trade off possibilities.
possibi1ities,
along with other
decision criteria,
enable him to select an optimum or
near-optimum schedule.
often in advance, and pin
way
each
to
Now,
wi th the extension of network analysis into surgery
e.g.
its
use
in open heart
surgery
and
i ’kidney
transplantation,
it may well be described at times,
as
a
matter of life and oeath.
5.8
Simulation
Simulation is a method of
solving decision making
problems by designing,
constructing and
manipulating a
model
of the real system because real systems are normally
quite complex and too intricate to solve directly using
analytical methods.
In such a case, we can try to construct
a laboratory model of the system. A simulation thus attempts
to reproduce a system in what is the equivalent of
a
laboratory setting and computer provides the speed for
generation
of
large number of
samples
and
multiple
comparisions that may be involved which otherwise are rather
impossible for a human being.
27
Simulation may thus,, be defined as a quantitative
technique that uses a computerised symbolic model in order
to represent actual decision making under uncertainty for
determining alternative courses of action based upon
facts
and assumptions.
Cnie of the applications in policy analysis
will
Le in evaluating the alternative health care delivery
systems, programme interventions at various budget levels so
as
to
obtain
complete
information
regarding
the
effectiveness of various policy alternatives given the
budget constraints and the programmes to be implemented
under these alternative health care delivery systems.
This
will involve generation of a disease profile, calculation of
effect of various progrmme interventions under alternative
health care delivery system,
population coverage at certin
budget level i-e.
effect of budgetary constraints,
ranking
of
alternatives
according to
health
indicator
and
sensitivity
analysis by varying the effectiveness
of
interventions,
utilization of health services by community
and changes in the resource availability etc.
Such
applications have been discussed in detail by Grorse et al.
(1979).
Similarly,
it can be used for solving queuing
problems where specific methods available are not ,applicable
because
of violation of
basic assumptions for
their
applicabi1ity by the system.
An application of
simulation
method
to appointment system in a hospital
has been
discussed in detail by Bailey (1964).
Consider out-patient
sessions conducted by a consultant.
Here, large queues may
build up in short run even though the total
amount of
consultant’s time is normally more than sufficient to deal
with the demands because a great many patients arrive at the
beginning of the session.
The unscheduled arrivals can be
restricted by introducing appointment system and making the
inter-arrival
time equal to average consultation time i.e.
if average consultation time is 5 minutes, give appointments
at an interval of 5 minutes.
Even then,
the problem is to
determine how many patients ought to be present when the
consultant starts the work i.e.
the initial
batch of
patients so that the waiting time of the patients is minimim
as wel1 as the idle period for the consultant.
This problem
coul d be solved by applying computer simulation to generate
1 arge
number
of
sample observations
using
suitable
distribution based on average service time,
the sample
observations representing consultation times of
successive
patients
and then
investigating different
appointment
systems to these sajimpies
REFERENCES
1.
Bailey, N.T.J
N.T.J.
,(1964):
(1964):
Elements pt Stochastic Processes with
AeBllSatipns to Natural Sciences., John Wi ley & Co. New York..
Churchman C.UJest, Acko-f-F Russell L. , and Arno+f E.Leonlardj,
(1957)s Introduction to operations Research3Wi1ey,New York.
28
3.
Cox.D.R. (19&2) : Renewal, theory, Matheun and Con LtdB ■t Lon don«
4.
Feldstein, M. Setal(1973): Resource
F
Allocation Model -For Public
Health Planning: A Case Study ot F.B.Control.WHO Bui 1etin, Geneva.
it;
(1976)
E.
(1976):.:
Introduction to fiflerations
Research^ A rcomputgr oriented Algorithamic ABBCoach
___ i j, TataHcGraN Hill Publishing Co.. Hew Delhi,,.
Gupta,
Iswar,
Zoreda,
Juan and Kramer Nathan
Hospital Manpower Planning by use o f Q u e u i n g T h e o r y
Services Research &x Ng,_ i_._
7.
Grosse,
R.N.
devries J.L., Tilden, R.L., Dievler, B.S. and
Day,
S.R.
(1979): Health Development Model : AfiBlicatign to
Rural JjSSax
Lhivarsity of Michigan, Ann Arbor..
8.
Grundy,
F. and Reinke, W.A.(1973): Health Practice Research
and ±O!2mall.zed Maggaerial Methods, Public Health Paper No.
51,
WiKJs Gewevau.
9.
Hardin,
Garrett (1968): The tragedy of the Commons, Science
162 3059 „ 1243-40.
10.
Hiller,
Frederick,
S,
Li eberman,
Gerald
J.(1972):
Introductten. to QB.erati.ons Research
n? Holden
Day Inc ?
San
Franci sco.
11.
Reinke,
William
A,
Kat h1een N(1972):
Health Planning
Qualitative aspects and Quantitative Technigues,
The John
Hopkins Uni versi ty,
School
of
Hygiene and Public Health ?
Department of International Heal th , Baltimore, Maryland.
12.
Rogers Peter (1969)i A game Theory Approach to the
problems
of
International
Ri ver Basins,
Water Resources Research
O BX.O.
13.
14.
<
«
(1971):
Health
/W-6O..
Srinath,
L.S.(1973)
and
PERI
Be.B.l.icatj^qns,
At -t i 1 i ated East West
Delhi.
Stokey Edith and Zeckhouser,
Analysis^ SMLW™ IMortan torapany.
Press
Ecincwies and
Pvt.
Ltd. Mew
~ -----E for Policy
15.
Wiest,
J-D.
and levy F.K.
(1974):
PERT/CPH, Fremtice Hall of India™ Pvt™ B management Guide to
Ltd. Stew Delhi...
16.
Mehta A-K. ji Thaiagrajan TR. , Jai swals,
N.K. - A collection
of some Operational
Research Problems fr om Wor1d War-11
Defence R
p
D organisation, IMinistry of Defence,
, Govt. of
India, New Delhi.
29
PROBLEM SELECTION AND
FORMULATION OF HYPOTHESIS
Dr. (Mrs.)s.Bhatnagar
It is important to identify criteria for selection of
\*
health related problems to be given priority in research.
Source of the important consideration to choose a
problem for research are given belowiIt
Urgency
Is it a national problem ?
How severe is the problem ?
2.
Relevance
Who is affected ?
What is the magni.tude
3.
Timeliness
Time relevance
4.
Applicability
Will the results find immediate use ?
5.
Practicability
Administrative constraints and resource
availability
6.
Political
Acceptability
A national solution may be accepted
7.
Cost-'f s. ect .l ’•
effectiveness
Will the results help in reducing the
costs.
8.
External i n f 1 .u <. r ~<"
influences
Central Govt.
State Govt.
International Agencies
9.
Individual's Interest
10.
Ethical Issues1. socio-cultural sensitivity
2. Informed consent
3. Research may delay the treatment
-2-
MAJOR DETERril^AHlo
WILL ALWAYS BE THE FOLLOWING
problem must focus at a purpose and policy issue. It
should be a factual problem/ should recognise contemporary
social importance,
should examine extensive Vs Intensive
coverage and feasibility of collection of Data.
F QRMULATION OF HYPOTHESIS
HYPOTHESIS
1 •
A plausible Guess as to some aspects of relationships
existing between two or more variables.
Hypothesis is generally a positive statement of the
basic questions a study is investigating.
Dosajh/ Kataria and Ramaiah
2.
Hypothesis has been
..fined as a tentative solution posed
on a cursory observation of known andavailable data
and adopted provisionally to explain certain events and
to guide m the investigations of others.
A hypothesis
is essentially tentative, likely to be modified during
the investigation.
M.H.Gopal
-3-
3.
When
A hypothesis states what we are looking for*
facts are assembled^ ordered and - seen in a relationship
they constitute a theory.
The theory is not built on
speculations but upon facts.
Various facts in a theory
may be logically, analysed and relationship-other than
thcse stated in the theory can be deduced.
The formu-
lation of deductions constitutes a Hypothesis If
verified it becomes part of the theory^
Goode & HATT
4.
NLLL
NULL HYPOTHESIS
A statement .(a plausible Gues) that some observed
result is due to chance and not due to any meaningful
relationship in the variables observed.
The null Hypothesis
is used in test of significance to permit the acceptance of
alternative hypothesis
due to chance)
(that the observed results are not
when the null hypothesis can only suitable
calculations be rejected.
For example - a difference is noted in proportion of
individuals vaccinated in two samples. it is desired to
know if this difference is a real one. indicating difference
in the populations sampled or merely the sampling fluctua
tions •
-4-
ROLE AND FUNCTIONS OF HYPOTHESIS
Hypothesis is to indicate the direction of the
Investigation and‘to suggest what facts are to be collected.
The hypothesis provides entry point to the enquiry
without which the Investigator may collect non-essential
and may be useless data and may overlook really significant
and useful one.
O'
TYPES OF HYPOTHESIS
Low order of abstraction
CRUDE
Indicates only the type of data to be
collected.
(Descriptive methods of Research)
REFINED Hypothesis tests logically derived relationships and
examines multiple variables.
Is Hypothesis always necessary ?
Not in exploratory and in fact finding researches
©
Methods for formulation of Hypothesis
1.
Method of Difference
..r
-5-
2.
Meunoa of Agreement
3.
Metnod of Concomitant Variation
4.
Method of Analogy-
Problems in formulation of Hypothesis
1.
Absence of clear Theoretical frame work
2.
Lack of ability to utilise the theoretical
frame-work.
3.
Lack of acquaintance with available research
techniques.
*
LECTURE OUTLINE
Management of Research Project
By
Dr.R.S.Gupta*
I.
Work Planning & Scheduling:
After the research design its objectives and strategies
have
been
detail.
finalised3
The
v/e begin to plan the project work in
first step is lo specify what major activities
will be carried out:
It may include:
The tasks to be performed.
When the tasks will be performed.
Who will perform the tasks and the time each from will
spend on them.
Thus,
a
work
summarizes
in
a
research
dear
fashions
various
components
project and how taey fit to-gnther.
duration
the
plan is a schedule3 chart or graph that
of
of a
It also tells
each activity and ;ho will be responsible
for its conduction.
Create
Activity
schedule
This
is- established through a 3 step
procedure.
1)
Construct
Preliminary Activity network: The first step is to
determine the sequence in which the activities
l
should be carried
out. This is done by asking 3 questions.
a)
What
other
activities
activity can be undertaken.
must
be
completed before this
b)What
other
activities
can not begin, -.until this "activity
is completed,
c) What activities can be performed c_t the- same . time as this
activity.
. This
the
establishes.
i'ntei’
relationship
in time between
-
activities.
2.
Estimate
Resource
Requirements
for
a c c omp1i s hmc nt
of
activities: (the major limitations will be on money and manpower)
3.
Schedule Project Activities; First estimate how much time is
needed to complete each activity. In this task 3 elements must be
considered
activity
together
(a)
the
inherent time requirement of each
(b) its relationships in time with other activities and
(c)the limits on resource availabily by type and period.
The
activities/jobs
in Fig.l.
shown
are represented in the form of Bars as
The length cf bar* indicates the time duration
taken for completion of that job.
j ob s
can
be
taken
up
Generally
.in any project same
concurrently and some before or after a
particular job.
To
conclude,
those who are to carry out or manage a health
systems
research activities■should
e,xamine
and-
discuss
the
prepare a detailed workplan.
same with other concerned colleagues.
before initiating the actual work.
r-
2
GANTT CHART
The first step would be to write down beside each listed activity
the activity' numbers of all activities that must be accomplished
before it can begin. Then draw a Gantt Chart.
This
Gantt
Chart
can
also
be
for
usc-d
monitoring and
progress of project.
TITLE OF PROJECT
Acti- Activity
vity
Responsible
No o
1990
i Jan. ; Feb. . Marj Apr.! May
a
~~
2.1
Prepare proposal
and submit to
..-Govt*. _ _____ __
2.2
Discussion with
Govt.
P.I.
2.3
Receive reply
from Goyt.
P.I.
2.4
Research Staff
P.I.
2.5
Purchase equipment P.I./R.O
2.6
Conduct training
to staff
R.O.
2.7
Collect data
from field
Field Worker
(F.W.)
2.8
Analyse data
F.N./R.O.
2.9
Prepare draft
report
R.O.
Principal
investigator
“
-1.- -A I
I
-t.-
SI
-t
4 __
■L
I
'
.-1--
...
.. . . o . _ -.L _ . ...i
2.10 Workshop on
findings
P.I./R.O.
2.11 Revise Draft
R.O.
2.12 Submission of
report
P.I.
I
2.13
3
II.
Preparation of budget:
The
following things should be kept in mind while preparing
the budget:
a.
Duration of the pr;-jcct (huw many months)
b.
Manpower
for
secretarial
(Normally
the
help
the
project
-■ their
salaries
research
and
and
allowances
research staff which is appointed for a
project 3 fixed salary is
c.
both
paid-and not the sca’le) .
Expenditure on data collection - TA/DA of the staff and
POL expenditure.
d.
Computer
time
(cost)
and
purchase
of any equipment
(cost).
e.
Stationary
requirement
including the printing charges
of the final report.
f.
Overhead
costs
(depending
upon
the
policy
of
the
funds sometimes already available within
the
organisation).
III. Source of funding:
1.
The Agency asking you.to ourry cut research
2.
Research
organisation.
3.
Funding agencies within the country:
Concerned Ministry/Department of the.Government
.JCMR
ICSSR
UGC
Planning Commission etc.
4
4.
In t e rn a t i o n a1 Agencies:
WHO
UNICEF
ILO
UNFPA
World Bank
Asian Development Bank etc.
5*
B-ilateral Agencies;
USAID
DANIDA
SIDA
ODA, etc.
IV.
j?-,JIicies and Procedures
application and approval.
Every organisation has its own policies and procedures,
Mo s t organisations have
their own proformae in which such
requests
for
are accepted. They take their own time and may ask
further
detailssif
lie
oS.ai’y.
Usually
they take the
opinion of experts on such proposals.
For International/Bilateral AGencies, such requests
are
normally routed through
respective Ministries/Departments of
the Government.
V.
4
Administration and Monitoring:
Composition of Steering Committee for the Project
Constituting the study team
Assigning
responsibilities
different
components of research3 depending upon their
speciality.
5
to
different
members for
Appointment of staff-both research and secretarial
Accommodation for staff, etc.
VI.
Monitoring
to
be
done
by
the Steering Committee and/or
ProJect Director:
Stages of the project
Person
Time frame
responsible
1.Review of literature
2.Preparation of questionnaire/
toolss etc.
3.Pre-test of questionnaire
and training of research
staff
4.Finalising the questionnaire
and its printing/stenciling
5.Data collection in the field
6.Data analysis
7.Draft report
8.Final report,print .ng and
submission
6
Guidelines to Investigators
All Investigators proposing reserches which may
involvehuman experimentation in any form are required
ubmit the details of the research proposals to the
ics Committee for approval. Tte format of the
research proposal should contain the following :
k
Title of the Study
Name^Designation & Qualifications
Of the Investigator(s).
Rationale of the S tudy.
I
Reieyant background information from
the literature including data ort. animal
experimentation with detailed preclinical.
studies!UbaCUtS 3nd chronic toxicological’
Precise protocol of research
proposed..
The possible hazards to the human
subjects involved.
All proposals needing the clearance of the Ethics
Committee should be submitted to the Member-Secretary of
the Ethics Committee after it has been
approved by the
Research Committee of the Institute.
The Investigator will be given a chance to
c anfy his/her position before the Ethics Committee,
r needed* T^e decision of the
Ethics Committee
of the
will however be final.
f
I
CONSENT FORM FOR CLINICAT.
TRIALS
Name of the Project;
PART I - GENERAL
I__
living at
s/d/w of
hereby give my consent freely to
trial. I have been explained the participate in the above clinical
nature and ]purpose
--- -of the clinical
trial, and its potential
risks/discomforts/benefit (
— - expected, i arn
free to withdraw from
this trial as when I feel
so inclined.
PART II
SPECIFIC
(Nature and scope of the trial
and the role o? the
Signature/Thumb Impression
of Volunteer
Date
Certified that the above
consent form hasbeen signed in my
presence. The potential■ risks and benefits
so far as they are
known to us, have been ,
explained to the above named
volunteer,
The Volunteer is free at his/her
own risks to withdraw from the
clinical trial as and when he/she
feels so inclined. He/she has
been informed of the risks, if any/ involved in
such
discontinuation<
Signature of the Clinical
Investigator
Date
. JI
RESUME OF TALK BY DR>3.DjKAPOQR
Meaning, of Experiment
(1) Experimentation is an activity,• and (2) it is the kind
inquiry” •
Experimental enqviry has historical! been distinguished
from common-sense enquiry in two different w-?ys: one on the
bais of (1) what problems are investigated/ that is, subJect
matter; and (2) how problems are investigated, that is, method
Though control is not synonymous with manipulation, some
scientists consider it useful to make a distinction between the
general class of controlled inquiries and the special class of
inquiries in which control is obtained by manipulation. The
general class they call "research”; the special class they call
"experimentation". This practice has had the unfortunate conse
quence of giving non-manipulative inquiry a lower status than
manipulative inquiry. It would be preferable to call all contro
lled inquiry "experimentation". Then the emphasis would not be
on manipulation but on control, where it belongs. But, in order
to avoid doing too much violence to scientific usage, we shall
employ the term "research" to designate the general class of
controlled inquiries. Our concern, henceforth, will be with
methodologically designed research.
Research Design: Meaningz Purpose and Principles
Meaning:
Research design is the plan, structure, and strategy of
investigation conceived so as ■ j obtain answers to research
questions and to control variance. The Plan is the overall
scheme or programme of the research. It includes an outline
of what the investigator will do from writing the hypotheses
and their operational implications to the final analysis of
data. The structure of the research is more specific.. It is
the outline^ the scheme, the paradigm of the operation of the
variables. When we draw diagrams that outline the variables
and their relation and juxtaposition, we build structural schemes
for accomplishing operational research purposes. Strategy, as
used here, is also more specific than plan. It includes the
methods to be used to gather and analyse the data. In other
words, strategy implies how the research objectives will be
reached and how the problems encountered in the research will
be tackled.
*
2
Purposes
Research Design has two basic purposes* (1) to provide
S. to research questionsr and (2) to controT variance
Naturally, research design (RD) does not do~these things; "
only the investigators (I) does.. Design helps the I obtain
answers to the questions of research and helps him to control
the experimental, extraneous, and error variances of the
particuiar research problems, under study. Since all research
activity can be said to have the purpose of providing answers
to research questions, it is possible to omit this purpose
rrom the discussion and to say that research design has one
gr_and prupose; to control variance. Such a delimination of
the purpose of design is dangerous, however. Without strong
stress on the research questions and on the use of design to
help provide answers to these questions, the study of design
exercisfnerate int°
interesting/ but sterile, technical
Experimental Designs
of.exPetimental designs applied to the study
of social factors
---- is
—5 to measure all effect of one social factor
at a time; all other factors in the: situation (unknown as vzell
as known) being held constant, i.e.,, controlled. Difficulties
arise at once because it is seldom ppossible to measure one
factor alone. Usually the best that we can do is to try to
measure a group of factors which operate as social treatment
or as a social programme - as, for instance, in an effort
to isolate and measure the effect of improved housing upon
a group of families formerly slum residents. a second
difficuity arises because there are unknown factors at work,
that is, unmeasured factors which are not controlled.
3 sCheck sheet for use in comparing the
efficiency of alternative experimental
designs.
Check sheet for design
Final reaction
Source of Invalidity
Comme nt" 1
Status
; Controlled
+
» History
Not * Not
sure*: Contro- <
? | lied
1
Maturation
*
co
•H
W
(D
:
Testing
*
Instrumentation ;
t
T
I
■P
O
a
i s
-p
Regression
(D
>
-P
rU
O
P
(D
-P
I H
•j
Selection
f Mortality
■H
Interaction
between
sei and mat.
i
I
? Reactivity
Testing inter*
Selection inter.
c c
o o
•H
'H W -P
£ O
' £ o ra
; p -H *H
' Reactivity
t
Multiple treat
'■ (D -P r-4
, -p o nJ
• X -H P
. M p (D
-P C
I
W O
(D tP
P
»
I
|
Other
J Other
Other
*
Depends partly upon presently unknown characteristics
or tlie variables and their relationships.
-
WHO assisted Course on HEALJi-i SYSTEMS RE SEARCH
HAiWOUT
by
Dr. S.D. Kapoor
Session c^X-
AC.T1OH RESEARCH APPROACH CO HSR
Obj ective:
To come to an understanding how Action Research(AR)
may perform any of number of functions/ consciously
or unconsciously:
1.
a.
To provide ^objective” daca in order to justify
a policy decision made or a course of action
already taken.
b.
To furnish information that can be used in
administration as a basis for future action.
c>
To justify a delaying action or taking of no
act ion.
d.
To sensitize action workers to the dynamics of
change and the implications of different means
of achieving ends.
If research and action meet/
it is only because action
programmes are ways of solving social problems and are
therefore characterised by structural and behavioural
patterns evidenced in any form of social activity.
2.
Behaviour al/So ci al Sciences research and HSR will alv/ays
have some relevance for action/ because the subject of our
research is human behaviour and action programmes repnesen
one form of human behaviour.
3.
The Action Research is both^ijan ap pro act to problem solvingj
L.e.fa model or paradigm;
anqiga problem-solving process^a
series of activities and events.
It is a data based/ problems solving model which involves;
a•
data oohlection.
b.
feed- back of the data collected/ and
c.
action planning based on the data
2.
ACTION R£o£A^Ci-i AS A PROCESS
4.
This is the process of systematically collectingresearch
data fibout an ongoing system relevant to seme objectives/
goal/ or need of that system;
feeding these data baJck
into that system; taking actions by altering selected
variables within the system based both on the data and
on hypotheses; and evaluating the results of action by
collecting-more data.
The key aspects of the A.R.model a.; ex
a.
Diagnosi s
b.
Oat a gathering
c.
Feed-back to the client group.
d.
bata discussion and work by the client group
e.
Action planning/
f.
ACTION"
and
The sequences tends to be cyclical/ with the focus on new
and advanced problems as a client group learns to work more
effectively together.
5.
Action research is a process in two different ways:-
a.
It is a sequence of events and activities within each
iteration (data collection. feed-back and working the
data, and
b.
saking action based on the data)
It is a cycle of intorsions of these activities semetimes
treating, the same problems through several cyc^s and
sometimes moving to different problems in each cycle.
aspects point^ip the\enqoing nature
of action research.
Both
3.
AdJlwi\
6.
/-i J Ad
.
Ro Adri
This is the application of the scientific method o
fact-findings and experimentation to practical
problems
recTuirin;; action solutions and evolving trie collaboration
and cooperation of scientists, practitioners and lay-men.
The desired results of action approach are solutions to
imme^cl^te problems and a contribution to scientific knowledge
and theory.
The action research model is a normative model
or a model for
case o
for 1 naming/
Its main features ares- In the
; 1 ann ed ch .an. • e.
intelligent human action there should be an objactive/
be it ever so fuzzy or distorted.
And in advance of human
action there should be 31anning.
although knowledge of paths
to the objectives is always inadequate.
be taken as a steo ar a time.
do some fact- f~:
Action itself should
and after/each step it is well to
dings.
The fac--findings may disclose whether the objective is
reali sti c.
whether it is nearer or more distant than before.
whether it needs alternation.
situation can be as massed.
Through fact-findings.
and this information, together
with information about the objective.
the second step.
the or-sent
can be used irjplanning
Movement toward on objective consists of a
series of such cycles of plannin-—^acting—>fact-finding—>planninc
7.
There are 4 types of actimi research;-
a.
Diagnostic action research:- probable diagnosis—
recommendations for remedial treatment to the clients.
.4 .
be
Parti cipant acti-n research: -
fhe action t iking people ire involved in the entire
f rora tn e beg i nn i ng .
r e s e a r ci“i an < i action researon
Jri i s
keeps the recommended actions feasible and sorkaola.
c.
Smpirical action research:-
.■^tor keeps systematic,
extensive records of •..s.?_t he
did ?nd what effects it had.
Sxoerimental action researcn:It is a controlled research on the relevant effectiveness
of various cechnicues•
8.
of action and research.
fne re is s orn e i n t e rd ep end en c e
the instrumentality not connected
ction is carried out
with a research institution.
Research functions are
relegated to an independents but sympathetic research
body •
Yet res ^arch is being conducted on action is
guided by res ;arch findings.
9.
fhere is always a mutual
artnership of research and
action in a common endeavour.
in-action,
In an action-research-
the ac^ion-research-leader-decision-m^er oi
person involved in the project plays the role of an
experiment-list; he isolated the variables beforehand.
He is in possession of these variables and the mechanics
of field and office operation are prepared before
plunging into action.
com ? 1 e t a d tn e
once the operational design is
act ion-resear ch-leader-decision-maker feeds!
che var i a :1 as discriminate!'/ into structure.
He records
what haopens under hi oily specified conditions.
vary the conditions if he want* to.
the situation
He may
He is in command of
and so he must be knowledgeable bouh in
r23earch acthodslry
action programme.
in substantive concent of the
5.
10 .
The criterion of reievinc? for
action or for policy-
making is of cricicil imnortence in ?icti on re s o ?.rch.
the focus shoul
wr. i ch c an
be on the ".aanicul •ble" variables
actively be taken into account when formulating
policy or development strategy.
11.
For a more insightful approach to the task of defining
problems areas for evaluation research the following
questions may be very useful.
All action programmed are
designed to oring about change of one^r another; these
are the dimensions for which data are usually sought.
1.
What is it that has changed?
2.
How much h\s it changed (extent)?
3.
How quickly has it changed (rate)?
4.
What were the conditions before and after the change?
5.
.-.hat occurred during the transition?
6.
What were the stimuli that induced the change?
7.
Through what mechanisms aid change occur?
8.
What brought stabilization at a particular point in
9.
change?
Can directionality be Observed in the change?
These questions lead one to identify specific problem
areas which can be the focus of evaluation:
(a)
Impact of the programme on intended recipients (indivi-
du al s, groups,
communties.
and institutions)
as well as
on the change agent themselves.
(b)
Extent of the programme1 s effect on the intended
recipi ?nts*
for example.
by the programme and
the number of persons influenced
the degree to which they were
influenced.
(c)
Time involved in cringing about the impact of effects.
(d)
before-after measures of the variables supposed to
undergo change.
If chan e in health practices is sought^
indicesof tnese practices ar j taken oefore and afteA
6.
the programme.
Other things being equal, the difference
between the two measures is attributed to the impact of the
programme.
e.
Identification of the events and processes that took
place between the before and after-measures.
Because of
the difficulty of keeping track of what is happening.
an intermediate measure is sometimes taken.
f.
Analysis of the subject-matter or substantive content of
the programme (different components and how they relate
to each other and to the objectives of the programme) •
If the aim is to increase rate of utilization of immuni
sation, the nature of the subject matter in the programme
of the change to be introduced must be such that they
will contribute toward increase in utilizing the
immunization services.
g*
Organizational, structural, and operational aspects of
the programme (how the programme was carried out and
administered, by whom, and with tthat techniques) .
h.
Factors in the change introducted, in the change-agent,
and in the entity or unity being changed which led to
the acceptance and incorporation of the innovation into
the life of the unit or entity being changed.
Direction of the changes observed relative to the
objectives of the programme.
The direct, indirect, anti-
cipated and unanticipated consequences of the programme
are observed and inter-preted in the light of programme
objectives and their corresponding implications.
9-
7 .
12.
dhen and how to use action rose arc. . in lie altli Organization
DevelODment(liOJ) •
)
fhe Health Organization Development;-;
is Xijically -in
action research orograrnfne designed to improve ure
functions of a health organization or system.
Effective
imp rovem en t p rog ramra e almost always require a data base/
i.e. they rely on systematically obtained empirical facts
for planning.
action.
takin. action and evaluating action.
Action research supplied an aporoach and a process for
generating and utilising information about the systems
itself that will provide a base for the action programme.
The collaborative enquiry features of action research
suggests to practitioners and lay-men alike
for jointly determining control needs;
and hypotheses/
and (3)
(2)
he( 1) desirabi 1 ity
critical proolems
actions.
The potential experimental natur:
■ f actions inherent in
action research provides a different set for managers as they
try to solve problems. ice. viewing problems in cause—effect
terms and viewing solutions to problems as only one action
hypothesis from a nange of several.
The systematic collection
of data about variables related to the organisation' s culture—
wh i ch m 2 an 1 ay-m en are only now going to view as important
determinants of performance----- and testing the effects of
managerial actions on these variable after new/ tools for
understanding organization dynamics.
.All these features fit
with a programme to improve the organization.
The nature of organizational development an j action rosearc
are very similar.
They are both variants of appli jd behaviour
sciences; they are ooch action orient ec ; the j. .are both d a t a b as ed;
they both call for close collabor i-ion between insiders and
outsiders and they are both problem solving social inventions.
Prepared by Dr. S.D, Kapoor
Before we discuss, methods of HSR, it is
necessary to define a number of terms such as
"Scientific Tool", "Scientific Techniques",
"Scientific Method" and "Methodology" as defined
by Ackoff:
1.
By a scientific tool we mean a physical
or conceptual instrument that is used in
scientific inquiry”. Mathematical symbols
and tables of random numbers are -therefore
tools, according to this definition;
2*
"By sci entific techni qve we refer to a
way of accomplishing a scientific objective^
ii
a scientific course of action.
Techniques,
therefore, are ways of using scientific
tools”e Stratified random sampling is
thus a t 'chnique; it employs the random
number of tools;
©
Prepared by
3<
Dr. S.D. Kapocfl?
"By a scientific method we refer to the way
techniques are selected in science; that is,
to the evaluation of alternative courses of
scientific action. Thus, whereas the techniques
used by a s ientist are results of his decisions,
the way these definitions are made is the result
of his decision rules# Methods are rules of
choice; techniques are the choices themselves”.
Thus the choices between simple random sampling,
stratefication, and other sampling designs is a
matter of scientific method. The rules for making
the choice must be such that the result is based
on fact and is reproducible, and not the biased
product of the technique selected.
4.
The study of scienti fic methods is frequently
referred to as methodology. The objective of
methodology is the improvement of the procedures
and criteria employed in the conduct of scientific
research.
'2
Prepared by Dr*. 3
Kapoor
RESEARCH and the HSR
Research to improve primary health care must be
useful*
The usefulness of research i s determined by the
importance and relevance of the questions asked*
Asking the right questions is the starting-point
of all good research^ as it is of any productive
effort to improve health and the quality of life.
Health Systems Research can be distinguished from
basic research mainly by its direct-focus on solving
practical and relevant problems^
It (HSR) will soon be discredited if it merely
results in repetitive or irrelevant studies and
evaluations that simply endorse the status quo
or accumulate data that are of no avail in produ
cing a tangible improvement in health.
In view of the severe limitations in research
capacity and resources, specially in developing
countries, studies should be focused sharply on
hiqh-priority everyday problems from field experi
ence. Bor many priority problems scientific know
ledge is available, and the obvious need is for
research on the ways in which to apply that know
ledge vdthin local cultural and resource constraints.
(3
Prepared by Dr. StD, Kapoor
APPROPRIATENESS OF RESEARCH METHODS FOR H.S.R.
Health Systems Research is concerned with the strengthening of health care.
It begins with real field problems and uses of
variety of research disciplines to apply practical
scientific knowledge for improving health care and health
status.
It remains in the realm of research because it uses
methods that yield :
(1) Bep ro du ci bi 1 i ty (i.e. solutions applied to similar
problems and conditions elsewhere would
produce similar results) .
(2)
Abilify to identify the key elements of a problem
(i.e. the characteristics that make research
approaches from one setting applicable in
another setting) 9
(3) An understanding of Cr nation (i.e., relationships
between actions and impact can be organised
in concepts and theories that may be usefully
applied to the prevention or modification of
health problems generally) .
The fundamental principles is thab the q uestions
studied should be relevant to the everyday concerns of
policy-makers, administrators, er anyone responsible for
'
-
care, Appropriate research methods that fit those
problems and situations should then be applied.
-."■ii tinr,j
Prepared by Dr,
Kapoor
RANG3 OF i-^TODS AVAILABLE FOR HdALTII SYSTEMS RESEARCH
APPROACHES
TO
HSR
1. Simplified Systems Analysis (°f heal ill care situations
and policies)
2. Operational Research (for incremental improvement of
management) o
3. Field Projects (to test alternative interventions and
management procedures)
4> Field Trials
5» Natural Experiments and Comparative Analyses
6. Field Experiments (for testing health packages)
The first three can best be done vzithin the health system;
the rest are more likely to need special studies by an
academic institution/research institution.
Prepared by iJr,
H. S. R.
HSR is the systematic study of means by which
biomedical and other relevant knowledge is
brought to bear on the health of individuals and
communities under a given set of conditions.
H. B. R.
Health Behaviour Research is a part of HSR in its
scope and purpose^
HBR is concerned with finding out what people
know, beli eve/
think and feel about health,
and
how such cognitive and effective bases are related
to what they cb •
HEALTH
BEHAVIOUR
BIO-P SYCHO-SOCI AL
CONCERNED
HUMAN
WITH
BEHAVIOUR
ECONOMICr
RESEARCH IS BASED UPON
MODEL
THE
OF
ORIGINS
IT
IS
AND CAUSES
OF
HEALTH;
IN RELATIONSHIP
CULTURAL AND BEHAVIOURAL
TO
SOCIAL,
CHANGES
AFFECTING HEALTH.
%
6-
oor -
7"A<> Cgvx CA^U" /jL
EtrcJUJAu
I
I
© SnTe comtawti
o O N DMy Ce>NI>l
I
" oC
I
I
®
I
i
I
BACk____ T
I
I
I
I
✓
I
I
\
• FLokJ
\
\
\
X
Z
z
I
I
9 Mrs
/
I
I
I
X
/
I
I
©
iNTfcepg^'
I
——,—
\pMC.e$SS$ ) [pur A>r
i h purs
I
, ,_SL_
Sub SySTCNS \
IfiTCWcm/
CMAN6L CMMCTM-'i
-/$7/e5 i
**
c^cept o
4
In this diagram, the basic system is depicted as
consisting of :
(1) Inputs (persons needing health care)
(2) Processes (Health care delivered by Services)
(3) Outputs (treated patients) ; and
(4) Feedback (effects of processes on future inputs)
Operating within a number of constraints,
such as
buildings, equipment, geography and population distri
bution©
HEALTH
SYSTEM
FRAMETORK
Components of the total health system as well as the
interaction of other sectors with the health system.
. IMPROVE!
Quality of life
t
IMPROVED WEALTH STATUS’
Bokouin,' ^tx/L
CO
C3)
INTB<A4TIOH
WITH OTHAA
3 ter#as
>
>
H
G
T
H
$
E
R
v
i
c
E
CS)s_
^AWAdte/HeNT
The importance of the environment and social and
economic development in
relationship to health
promoting behaviour.
>MMtMtvr
in vikohmEnt X^ >•
. *
■
I
■
9
LwiAMl&aJkr*'
9
t
^^^LylK^r<S
/\
»IbvctUi^^^A-tt
oiaaUs
fro^^rtt^ \
fUlAVfoJ^ (BcUamI
***>. MZfk
xX^ WJSiis^
A MoPC L
FP &
| NDMDUAL
(. JwtrCMCt.
Social
c £xivtnn «L«,\s«X )
tAakyJiC^^
Seci«JL
\/aAvu< §✓
| Sy^tosA.
W4fc«r
Fv«4«lviy
e>eH*wHX
Rcva^cjU^ £------------- -—l£x|5<*U**ey^
s
Wtf
CO 16 fa kxAei***
CX^4AA4**«A4«
*4
ff*^rfL<U
/<» oni ic><so
O/~ ST^tTlJi
Well
A
b&f-INEHS
ILL
A. N^y***L
(I
o
3
C
f
V)
X
C. MAlUsftAt/V,
®JU Mtt
_
My^ockon«l>u'AX
X
Rc,QAJt«iA«4uLf) bdwo^ f
f4*Au
tta. ua#U'«h^LuX qua4, <-C« s^XAju« 4a^c<U«^ .
f*
i
i
Seventh Five Year Plan
the
■^3.ternal_ran.d Chi^ld. Jie\alth
*
M
►
SECTION B: MATERNAL AND CHILD HEALTH
I.
INTRODUCTION
Maternal and Child Health forms the core of any sourik
health care system.
Realising the close inter-relationship
r
■
?•
■
between population central arid* improved health of mothers and
children, and
integrated approach f»r both these programmes
has been suggested in the Seventh Five Year Plan.
II.
t-
CURRENT STATUS
*
Efforts in the previous plans have brought in some
successLife expectancy of vomen has improved, althtugh the
sex rati* has remained adverse.
According to 19<31 Census,
theie is a slight improvement and the sex ratio stands at 933
as against 930 in 1971«
of by vomen increasingly 9
Health services are being availed
although it still remains less as
compared to males.
Infant mortality after remaining around 125 for. many
years has shpwn'a decline and is estimated to be 114 per 1000
live-births in 1980.
tional
However, this is still high.
The nutri
status of children has improved due.to child welfare
pr ogramm.es,
which are making steady progress. but not to the •
desired extent.
if
1.-
2-
III. POLICY THRUST AREAS IN THE SEVENTH FIVE YEAR PLAN
The major thrust of Maternal and Child Health Care,
in accordance with the National Health Policy, in the
Seventh Five Year Plan would be as under s
1 ,
Health care for mothers and children will be
strengthened
through the primary health care
approach , which includes integrated^ comprehensive
MCH care and suitable strengthening of referral
services.
2.
*
Recognising the close relationship that exist1
between high biri^rate and high infant mortality,
a-high priority should be given to MCH Programmes.
3.
Increased emphasis should be given on people's «
participation in MCH activities by supporting
voluntary organisations, NGOs, village health
committees, women’s clubs and traditional birth
attendants.
4.
A close linkage of health and health related sectors
with MCH activities will be strived at.
5.
t
Preventive , promotive and educational aspects of MCH
services will be given the highest priority.
IV.
«
BASIC APPROACH FOR MCH SERVICES
The basic approach for the implementation of the MCH
programmes would be as given below "
1,
MCH services would be provided based on the
’high risk1
approach.
-32.
Health and family planning services would be extended
ts ^all vjomen in the reproductive age group, and all'
chiidren,
3.
The existing bed situation in the districts vould be
assessed and depending upon the needj adequate beds'
would be provided for .women and children.
4.
The recommendation of the Rural Health forking Group
•s
ti
that 40 per cent of all new beds in the Seventh Five
Year Plan would be for women and children^ is endorsed
5,
Logistics and technical. consultative and and referral
support for primary health care in the form-of qualified
professional manpower (obstetric. Gynaecology,
Pediatric and trained professional workers who have rtiana-
genial skills’, should be provided at secondary level*
In order to bring more women and children within the easy
reach of MCH services, the health infrastructure
should be augmented and population coverage should be
progressively reduced below 5000 per sub-centre,
wherever possible.
7>
Efforts should be made to maximise the use of TCDS
infrastructure for the enhancement of MCH programmes.
8»
Special I^^campaigns would need to be organised t»
educate women on the advantages of prolonged., breast
feeding»
-
I
*
0
-4V.
TARGETS TO BE ACHIEVED BY THE END OF THE SEVENTH
FIVE YEAR PLAN
1 .
Infant Mortality rate
2e
Maternal Mortality rate
2-3 per 1000 live
births
3e
Perinatal mortality
40 per 1000 live
births
4,
Percentage of babies with birth
weight less than 2500 grams
18
.5..-
Mortality in- the
—pre-school
child ((1 to under 6 years
of age)
15 per lOOOchildren
in that age group
6.
percentage of pregnant mothers
receiving ante-natal
—-- care
60-75
7.
Percentage of deliveries by
trained birth attendants
80
8. .
Percentage of TT for pregnant women
100
Percentage of TT for school
100
9.
children upto 16
. > years
10.
DPT coverage for infants
11\
Polio
12.
BCG
13. ^T, (School entrants)
14.
Typhoid (school entrants)
87 per 1000 live births
85 percent
85 percent(proposed
target are
85 percent higher
than those
set in the
National
Health
Policy)
?•' ■ ‘
85 percent
85 percent
-5VII.- PR0GRA1W1E OUTLINES
1.
Medical Termination of Pregnancy and Care of ■■.■omen
1.
The health of mothers, and in particular, maternal
mortality insignificantly affected^ by induced abortions
^.performed by unqualified persons under unhygienic
conditions. The Medical Termination of Pregnancy
Act (1971) is a legislative measure for improving
maternal health. Services for Medical Termination of
Pregnancy (MTP) are to be increasingly provided. By the end
of the Seventh Plan period, MTP services should be
provided by all Primary Health Centres. In urban
areas, it should be available in all maternity homes
and centres. Central Government should provide
appropriate fin.ncial supportfor this purpose, .MTP
services should be an integral part of maternal and
child health services, and be closely linked with the
ante-natal programme. At the centre and States, one
officer should be responsible for coordinating and
implementing this programme.
Training programmes ri
should be vigorously pursued for improve^ delivery of
services. Enhanced compensation should be given to
.States where women.accept sterilisation or IUD after
MTP to enable them to meet the cost of drugs,
dressings, etc.
An intensive educational and
publicity programme, commensurate with available
facilities, is suggested for improvet utilisation.
2.
2,
Care of women, in addition to services provided
through the general health core system, will aim at
raising their health consciousness, A comprehensive,
field-based information, education and communicc.tion
programme will have to be developed. Women should be
organised around some economic activities to enable
them to actively participate in the entire process of
socio-economic development including health.
Care of ...Preg^ncmt and J^urs ing^ jj/omen
Pregnant and nursing women are a vulnerable segment of the
population.
Maternal mortality rate is estimated to be around
4-5 per 1000 live births>
Sepsis, haemorrhage, toxemia,
illegal abortion and anaemia are some of the important causes..
•
o f mate rna 1 m )r t a lity <
j
■
,
■.
Services for health care of mothers
during ante-natal, intra-nr. cal
and post-natal period will be
■.
«*»•■$.■'
-6-
strengthened.
Additional inputs will be provided for
training, increased infrastructural facilities at various
levels, adequate drugs and equipment and improved mobility.
All eligible pregnant women will be covered by prophylaxis against anaemia.
Birth rooms wherever possible will be
provided at sub-centres, and at village level, preferably
through community participation.
Services of obstetricians
and gynaecologists will be made available at all community
health centres, sub-district.and district levels.
•
Care of Newborns :
It is accepted that about half of infant deaths occur
%
during the fi^rst month of life, and in particular, durirg
the first week of life.
The main causes of death are,
directly or indirectly , low birth weight, inadequate ante
natal care of the mothers, poor care of the newborn soon
afteb; birfh (resulting in deaths due to asphyxia, pulmonary
complications, etc.) and
inadequate referral facilities from
primary to secondary and tertiary levels of care.
The
importance of preventive and promotive aspects of newborn care
was realised. and it was felt that specific funds for this
should be earmarked, particularly for IEC activities 9 and
organising rural women.
Fifty pilot projects are suggested for a comprehensive
programme for care of newborns.
At primary or level I care
for the newborn, improved facilities for conducting deliveries
will be provided
at sub-centres, primary health centres.
It
-7maternity centres, as well as for domiciliary deliveries.
High risk cases will be referred appropriately.
For
instituting neonatal care services at intermediate or leve 1
II care, all medical colleges, post-partum centres (or
Category A and B) and institutions with annual deliveries bet
ween 1500-2000, will be suitably strengthened in terms of
manpower, equipment and supplies.
Special baby^ufti^s.with a
capacity of twenty beds will be set up at the above mentioned
institutions,
Regional perinatal and neonatal centres will be
established for providing tertiary, or level
leve1
III care,
These
centres would provide the necessary linkage for medical care
and impart the required training to medical and para-medical
personnel.
There should be one such in each state.
Allocation of beds should be made available at district level
and below.
Existing beds for newborns should be fully
utilised.
Care of the Young Child :
The infant mortality rate in India is still very high and
deaths in the pre-school child is responsible for half the
total mortality. . Major causes of deaths are infections (such
as. respiratory diseases, diarrhoeal diseases and others)
9
dehydration and malnutrition.
Most of these are preventable.
Reduction in deaths due to diarrhoea, respiratory
infections and malnutrition is to be brought about by
training multipurpose workers and traditional birth attendants
-8-
in the recognition of these problems, administering primary
care, as well as in referral of selected patients.
Facilities for secondary level care have to be created.
Support facilities and supply of drugs at primary and secondary
levels of care will be augmented.
Additional pediatricians will be trained to provide the
required services for the Seventh Five Year Plan period.
Facilities for training doctor^ from primary health centres,
as well as para-medical personnel, in aspects related to
delivery of MCH services, should be strengthened.
Training
centres for trainers of multipurpose workers in delivery of
child health services will be established.
Pediatricians
would be involved in the planning and implementation of
training.
5.
Immunisation :
The high morbidity and mortality among infants and
children may be attributed to several common infectious
diseases, which can be prevented by immunisation.
Under the
expanded programme of immunisation, vaccination against seven
diseases is provided.
The objectives of the immunisation
programme during the Seventh Five Year Plan will be to reduce
the incidence of diphtheria, whooping cough, tetanus, polio,
myelitis, childhood tuberculosis, and typhoid fever by making
vaccination services available to all eligible children and
women by 1990, and to achieve sejf-sufficiency in the
production of vaccines.
-9-
Immunisation against measles is to be included.
In order to achieve the objective of the programme it
will be essential to provide additional inputs for trained
manpower,cold chain equipment, transport facilities and other
essential supplies and equipment.
Immunisation services will
be provided through fixed centres, camps, and mobile teams.
The epidemiological pattern of diseases will form the basis
for programme operations.
The ’cold chain1 is to be suitably
strengthened for vaccine storage.
should be suitably
services.
Surveillance of diseases
strengthened to document the impact of
Information dissemination and health education is
to be promoted with a view to raise the health consiousness of pe
people, as wall as to provide support and knowledge to health
workers.
6.
Health Services for School Age Children
:
Emphasis in the health care of school age children (6-14)
is to'be given for the early detection of correctable dis
abilities -which will prevent major handicaps later and
infectious diseases.
The Multiplier effect-of education to
children. and the child to child extension, are important
aspects of- comprehensive health care programmes for children
in this age group.
Health services
for children, both in
schools as well as those in the community, are to be organised
and augmented.
Paramedical workers will be trained in the
-10-
detection of visual and hearing defects. as -well as postural
defects. They will adminster primary care for common
diseases.
Children will be examined and screened at primary
school entry 9 before leaving primary school and at completion
of high school.
Their immunisation should be complete.
Screening kits and medicine kits will be made available.
School teachers will be trained in the required areas of
health care.
The health programme for school age children
will be integrally linked with general health services.
Budget provisions for this programme should be made separately.
7.
A National Institute of Maternal and Child Health should
be created to develop training programmes in practical and
applied aspects of iMCH.
VII. RESEARCH NEEDS
■
1,
Identification of the -infras-tructural requirements to
Pril?ary health care and maternal and child health
care at various levels
- ----------•- sub-centres primary health centres,
community health centres district health centres and
referral hospitals.
2.
Assessment of manpower requirements, based
I
on scientific
estimates of work-load and populationl coverage.
3.
Development of referral services 9 including facilities
required at each level of health care.
4.
Development of mechanisms for monitoring, evaluation and
feed-back.
5.
strategies/modeIs to involve all sections
o± the community in programmes for maternal and child
health care.
4
-11-
6.
Cessment of training needs - both technical and
managerial for improvement of skills and knowledge
of all categories of health workers.
7.
Studies in, and development of, tidluxiig
training technologies
for qualitative improvement of technical and managerial
skills.
8.
Development of a comprehensive information education
and communication strategy for improved MCH care.
1
As^
-> c- li”
*
SAMPLING PLAN : HOUSEHOLDS
/
RATIONALE : SAMPLE SIZE
*
The characteristics that are considered most important
in the evaluation exercise are the vital rates,
especially the :
- Birth rate
- Infant mortality rate
*
The estimate of the vital rates are required for the
Project Area ^ls a whole
*
It is also assumed that the confidence range of the
estimates is required at a level of + 10% of the
parameter value.
*
And the estimates are required to be significant at
95% level.
*
The Baseline Survey of the area done in 1993 indicate
a Crude Birth Rate of 4.14% and Infant Mortality Rate
of 11.19%
*
Therefore the sample size was based on :
- An expected lowest parameter value of 4%
- Confidence interval of 3.64% - 4.39% at 95% significance
- The estimate requirement being at the project 1-vel
*
The total sample size, thus worksout to 10,300 hous-holds
in the Project Area
*
Though it will be possible to estimate the characteristics
with 4% expected value only at the project area level,
those with higher value can be estimated ar lower
geographical area levels, namely regions and districts
.
. SAMPLE SIZ^': 10,000 HOUSEHOLDS IN THE THOJECT AREA
SAMPLING DESIGN : RATIONALE
*
The indicators are to be estimated for the rural areas
only, as in the base line study. Therefore the study will
be restricted to the rural areas only.
*
In order to make the results comparable with the base line
study, as far as possible, the same design considerations
have been considered for this evaluation exercise. Therefo
it is suggested that the villages be stratified into 2
strata:
- Headquarter villages i.e. villages where PHCs, Mini-PHCs
or subcentres were located
- Non Headquarter villages ; where such service units
are not present
The rationale for this stratification is:
- An important objective of the study is to study
utilisation of health and family welfare services
- And utilisation is expected to vary between these 2
village types
- Therefore we need to study them separately
#
To make the estimates self weighting, the total samplis proposed to be allocated to the 7 districts, (Gwalior
& Datia merged) and further to each stratum within a
district in proportion to the population
*
For ensuring variability, the villages within a district
and stratum will be further stratified by the population
of the villages.
HQ villages
: Into 2 strata
Non HQ villages: Into 3 strata
*
The household sample size proposed to be 40 per village,
total no. of villages to be covered, therefor* works out
to 250
SAMPLING MECHANISM
*
A multistage stratified sampling procedure will b*
used in the sampling method
*1
Step 1; Distribution of No. of villages to the
districts in proportion to population
Step 2:
Distribution of the no. of villages in a
district to the 2 strata )HQ village/Non
HQ village) in proportion to the population
Step 3A: All villages within a district and Stratum I
(HQ village) are arranged in descending order
of population
They are then divided into 2 groups with
equal population
From each group equal no. of villages (i.e.
half of th- no. of villages allocated to that
stratum in that district) will be chosen. The
specific villages will be selected by Probability
Proportional to size (PPS)
3B:
The Stratum II villages in a district (Non
HQ villages) will be in descending order of
population
They will be then divided into 3 groups of equal
population size
Equal no. of sillages will be chosen from
each group. Th- specific villages will be
sel-?ted_by PPS.
Step 4: Within each village 40 households will
be chosen by Systematic Random Sampling
from the list of households for the village
per 1931 census.
*
In each stage substitutes will be preseletted which will
be used if needed.
I•
V
SAMPLING DESIGN ; VILLAGE & HEALTH
CENTRE LEVELS
DESIGN : VILLAGE LEVEL
*
Village level information will be collected from
- Village Health Guides
- Trained Dais
- Village Opinion Leaders - formal & informal
- Other voluntary organisations, if any
*
The villages to be covered will be the sam- as in thHousehold survey
*
All HGs and TBs in the villages will b° cov-r-d
*
In each village coverage 4 opinion leaders J2 formal
& 2 informal) is plann-d subject to availability
*
1 Voluntary organisation, p-r village will be covered,
if available
<•
.• *
<
.
DESIGN : HEALTH CENTRE LEVEL
I
*
All th* PHs conn*ct*d with the selected HQ villages
(stratum 1) will be covered
*
Subcentres located in the selected HQ villages will
also be covered in this phase of the study
Il '
/
fix;.
Mf 0
uj5.5 W
tn
._
;K
^y
A\
<
hid Hi <5
in c\
it5
X
u.
Vi U.
5
in
0
0
PUL.
.
$
^1
4
z®
\
Zt A
/r Ztd\
/^XC?\
$ EiHiS
(a^e % )
UJ
5
2
<)
^14
V ^2,
0? J
0-
’C.
Ik
o
i
t
!s
I r-
:□
AU
Z
I
i_
Id
7*
20
r)
H
V>
\32£^
ff 2 y
\?o/
—» o
?.?§
<
?
M
,<r co
vi
r
fcM?H
r
M 15
h Z
QC
1
V)
o
%
o
xz:,
'S
u
o
<3
ro'
S Ua a
V9
£ <
£
r
UJ
m
nJ
/
\w^/
\v)/
3\
1° '
15‘j
CT
z
□a.
««
Ul
U
3U
\^
\^l
\—* I
/S£5£ ^
AI?.
/^
\o /
\<J /
u 0
X ID
in j
a -a
5^
£
s
bi
\in /
a
ixJ
H
in
U-
\<5^/
<
v>
Ul
V)
Z2
o
o
K
£ i
U E
1
c
h?
\dD3/
3
r
V
V
<r
x
X
v< 111
UJ
l
v>
tp
\|7
*
u
u
■‘-^i a
g
x
.<£
o
•z.
>
25
o aZ
u <
cn tn
Sampling
1,
What is PQPn^^iQ1?/17n 1 vesg
It is the aggregate of all events, subjects, units,ele
merits those are to fee studied for a particular problem.
; A
What is samp]ing
It is the technique applied for:
i)
ii)
3.
determining the number of ob 1 ec^elements/individuals
needed to be studied and
their- selection from the total population and any
characteristic’s value is to be estimated or studied
Need of Sampling
Most often data collection by complete enumeration covering
each individual element in population is beyond our capacity,
keeping in view the cost involved, man power required and time
needed for collection and analysis of huge data. Thus to overcome
these problems, we resort to sampling. Moreover sampling approach'
has other advantage of getting more detailed and better quality
data.
4.
What is sample
It is a segment or a part of population through which the
characteristics of the population can be studied.
5.
Qualities of good sample
i)
i
ii)
ill)
It should be large enough to draw valid conclusions in
regard to population characteristics.
It should be unbiased. Various types of biases can
invade the sampling procedure because of convinience,
cheapness, speed and personal likings and dislikings.
It should be representative of the population. As
based on sample statistical inferences are to be drawn
about population parameter.
2/-
2
iv)
It should be correctly approached to provide com
plete and correct information.
6.
When not to use sample
There are exceptional situations where vie cannot resort to
sampling to study about the population such as s
i)
ii)
iii)
7.
entire count is needed,
• population is small and manageable-/
event to be studied are rare in the population.
•I'
I
Steps in sampling
i)
Before drawing a sample from the population decide
about What is to be studied, ii) identify and define the study
population and its characteristics, iii) define sampling units
(which could be experimental subjects respondents, speci fic
define scope
groups of persons, organisational units, etc.), iv)
of study and describe the area to be studied (state, district.
institutions, etc.).
8.
Sampling Unit
It is the basic or smallest ' unit ’/which the study is / with
concerned.
9^
Sampling frame
It is the list of sampling units from which the sample is
to be selected.
It should be ensured that sampling frame is
complete, relevant, upto date and correct.
10.
Methods of sampling
There are various procedures that can be used to obtain‘
a sample.
i)
ii).
Probability sampling.
Non Probability sampling.
P rob ability Samplings It employs random procedures to en11.
sure that the sampling unit is selected on the basis of chance.
Every member of the population should have a known chance of being
included in the sample.
This is also called as random sampling.
3/-
3
Hpwever, probability sampling is no guarantee for the
representativeness of the sample.
Even random sampling some
times produces non representative sample, particularly so when
the individuals in the population vary
"
in respect of
characteristics under study and/or when the sample size is very
small.
When to use:
If the aim is to guantity variables and to
generalize 'the findings to population.
Why probability Sampling: It is the only approach that
makes possible the formulation of representative sampling
plan iie. to estimate size of the sample needed.
A.
uni-stage Sampling: In this type^of sampling the ultimate
sampling unit is selected at the first.stage of selection itself.
a)
Simple Random Sampling: It is random sampling with furthen provision that each sampling unit in the population
list (sampling frame) has an equal chance of being
selected in the sample, It is the basic probability
sampling design.
If sampling frame is not available and construction
, is costly as well as tedious for large popuiati^ simple
random sampling is not preferred but otter methods of
b).
cluster or two stage sampling preferred.
Systematic Sampling: In this procedure, the sampling units
are selected after a fixed interval which is fixed accord
ing to the predetermined sampling fraction (n/N).
The
random start is however selected randomly (less than fixed
interval).
When to use:
(!) the sampling frame not available and costly
to construct,
(ii) the elements in the population do not have specific
cyclical pattern for the characteristics of interest but are
randomly arranged,
eS(iii) there is a flow of people to be sampled.
4/-
4
Advantages:
(i) Simple to draw sample and less time consuming.
(ii) .Concurrent listing is needed to draw a sample
and on occasions listing not needed.
(iii)More precise compared to simple random sample
when study units within the same sample are
hetrogeneous.
Limitations sFor unknown population size (M)z for higher estimate
of N you are likely to draw smaller sample than
intended and for lower estimate of N you are lately
to draw larger sample than required.
c)
Stratified Sampling;
With good knowledge about the
population, the total population under study can be
divided into well defined mutually exclusive sub
groups (each subgroup called a stratum). A random
samole from each stratum is selected with uniform
or varying sampling ffaction.
This procedure helps in improving representa
tiveness of the sample and hence precision-.
To
maximise precision (i) the stratification should be
done in such a way that the individuals within a
stratum are .similar as far as possible in respect
of characteristics under study,
(ii) the sampling
’ fraction should.be in proportion to the square root
of the variation of that stratum.
(i) population is hetrogeneous in regard to
When to use:
certain characteristics; under study,
(ii)
to ensure desired representation of subgroups
comprising similar study units.
(iii) Comparisons are to be made between similar
groups of study units.
Advantages;
Any number of subgroups possible.
can draw relatively larger sample fr«m smaller
subgroups
more reliable information for same sample size
comparison between strata easy.
5/-
i
- 6
Limitations: only possible when it is known tohat proportion
of study population belongs to each subgroup/^fstum.
Using unequal sampling fraction for drawing
sample from each
corrected
while generalizing the findings.
d)
of
Cluster Sampling: It is the selection ot
groups
of sampling units together. In this technique, total
population is divided into subgroups of sampling
units and then a random sample of predetermined
number of clusters are selected and each sampling unit
in the cluster is included in the sample.
Cluster has
defined boundaries and study units are close together
within the' boundaries-.
To .maximise precision, the cluster should be
framed in such a way, that the individuals within each
cluster vary from each other as far as possible in
respect of characteristics under study. nSecondly,
cluster should be small in size, with each cluster
containing small number of study units.
More clusters of samller size in the survey area
is better for survey results and managing field work.
(i) Sampling frame does not exist or difficult
When to use:
to list all study units
(ii) sampling units to be visited are scattered and
operationally not feasible to cover at random.
Limitations:
(i) requires larger sample size than SRS
but economic to draw sample
(ii) difficult to form small and more or less evenly
sized clusters
(iii) complicates mathematics of computing.sampling
variance and thus test of significance
(iv) less efficient compared to SRS and stratified
random sampling.
6/~
6
“T’
ft
B.
'
Multi-Stage Sampling: In this type of sampling, ultimate
sampling units are. -selected by stages and the list of ulti
mate sampling units are required at the last stage only.
-^it’is a combination of cluster sampling with SRS or syste
matic random sampling dr stratified random sampling at
various sampling, •stages depending on the homogeneity nr
hetrogeneity in characteristics under study at various
stages.
When to use:
(i) for large and diverse population making
it difficult and uneconomic to use uni-stage sampling
(ii) population is divided in natural clusters.
sampling frame is required for the last stage
Advantages-:
cluster finally selected
A
logistically feasible
Allows larger number of clusters into sample with
out over burdening with too many study units.
Limitations: problem of listing all study units in selected
clusters at various, stages
— lesser probability that final sample is not represen
tative for total sample.
C.
Sampling by Probability Proportional to Size:
(PPS Sampling)
It is the selection of sampling unit, which is
determined by the size of the- population and the chance of
selecting a sampling unit is proportional to the population
size, of that unit.
When to use:
,
It is applicable for cluster- sampling-, stra
tified sampling, multistage sampling for drawing
sampling units in the situation when clusters/strata
vary greatly in size.
Advantage: It reduces sampling variance and provides
reliable estimates.
7/-
■
'■
■
/and
,
12. Non Probability sampling are partial/ohly useful for pi o
studies where the aim is to get Only a’ rough impression of how
certain variables are distributed in the populationa)
Convinience sampling:a sampling method in
which-cthe ’ sample1 ^that happens to be available at
at the time or period of the research is selected
fur convenience sake.
(i) Too much accuracy about estimate frSm
When to use:
sample is not needed
(ii) to examine specific situation
(iii) probability sampling is too expensive «r too
difficult.
Limitations: (i) Probability of selection of element from
’■ population is not known.
(ii) It tend to over select.some elements and under
select others.
(iii) No precise way to adjust over representation
or under representation of elements and distortion
b)
caused.
Quotaamp1ing:
It is a method of sampling which
ensures that all known elements in the population
occur in the sample.
It is a .typical judgement sampling in whi«h
quotas are set up according to some specific
characteristics such as age groups/ income groups#
etc.
Within the quotas selection of units or ele
ments depends on personal judgement.
When to use:
(i)-When it is felt that convinience sampling
would not provide the desired balance ofelements in
the population. It is also used in the*population
where strata cannot be identified in advance.•
<*.
8
Procedure of Sampling: Decide in. advance the quota that
are wanted from each stratum* Continue sampling
from the population until.necessary quota is obtained
in each stratum*
Limitations: (if Time consuming and requires considerable
work on the part of enumerator as most of the.units
contacted may fall in state where qAota has already
been met.
(ii) Selection is not random but utilizes knowledge
available about characteristics.
(iii) Lattitude is permitted to enumerator to reduce
cost and time
v
(iv) sampling theory cannot be applied.
(v) Precision of estimates obtained^can be Judged
only when•comparison is possible with total population
or ether similar sample<-
13.
Sample Size
The sample size to be drawn from the population ab
as tointerpret about population characteristics under
study depends/?!)1 availability of resources determining
feasible upper limit and (ii) requirement of proposed plan
of analysis to answer questions raised in the study deter—
mines desirable lower limit.
The proposed plan of analysis could be in terms of
desired level of accuracy in estimates ef population
characteristics
relative analysis of cross tabulation
variations in population characteristics and test
—
v
of significance.
In the situation when desirable sample size exceeds fea
sible sample size the researcher needs to change plan ef
analysis or reduce scope or increase resources.
9/-
9
Very often sample' size required for (i) Qualitative ,
studies is larger to reflect important population variations
but small sample ts enough to allow for intensive study,
(ii) Quantitative Descriptive studies;
For quantitative
type of studies in which estimates of population character
istics or outcomes are expected, it is essential to have
an adequate sample size^to get an accurate estimate from
sample. Sample size is always determined in terms of the
denominator used for computing indicators such as mean,
proportion or rate. For example, for estimating birth
rate or death rate or incidence rate of a disease, the
sample size is determined interms of population required
to be surveyed where ^as estimating-' infant mortality ra
,
the sample size shal1 be^determined as number of live
births to be covered in the survey.
in: estiTo provide minimum sample size for accuracy
of the characteristics
mating mean or proportion as parameter
following guestions first,
of study units answer* the
estimates of key proportions
What are the reasonable
1)
' (if not known
•• or averages to be measured in the study?
or cannot guess use proportion as 50%).
What degree of; accur.acy required in estimate or How
2)
far can allow the sample estimate to deviate from the
(For high degree of
true value in the population?
smaller the error to be tolerated the
accuracy with
longer is the sample size r>equired).
What is the variability of the characteristics under
stud*^ in the population or the incidence of occurence
of the event under study (larger the variability/
larger is the sample size required, smaller the inci
-1 I-'
3)
w
dence larger the sample size required).
10/-
10
4)
/so
How confident do you want to be/that the ^sample
estimate is as accurate-as you wish.
For estimating a proportion (p) or mean (m) , sample
size (n) is determined by the formula
n
S2 pq
n
22 s2
d2
d2
where ]□ = prior estimate of percentage
;..
• ./;
Q ±= 100 - p
g = standard normal variate corresponding to
stated confidence level
For 95% confidence level, Z = 1-96
Fpr 99% confidence level Z = 2-6
d = error to be tolerated in the estimate
s2= measure of variation in population.
In case information is collected on more than one
/are
item and- number of indicators/estimated for the population,
select vital indicators and estimate sample size needed
for each of the important item.
If the largest of the
1 n*s estimated in within the feasible limit of the sample
size then this <nl is selected.
In case there is large
variation in ’n1 for each item and outside feasible limit
then the desired standard of precision may be reduced
for
certain items which are not vital.
(iii) Quantitative; Analytical Studies:
The basic requirement for cross tabulation analysis
■ are..:i) each category of independant variable '" included
in cross, tabulation should contain at least 5^
observations
(it) expected number of cases in each $ell. of a ijcLble
should be at least -five.
11/-
11
minimum sample size is deterUnder the above conditions
mined as:
50
50
i)
n
ii)
n
0705
cases
expected
in
smallest
cateproportion of
independant
variable
(Ps)
gory of vital
5
5
C .0 5x0.2.
Psi x Ps2
product of proportioni in smallest category of
variables in cross table
to
Education attainment^
Contraceptive
No
Secondary
Primary
College 5%
practices
Schooling 40%
School 15%
School 40%
Current user 50%
Past users 20%
Never users 30%
In cross tabulation select larger sample size to
meet both
(iv)
criteria of education and contraceptive practices,
Sample Size and Sampling, fraction for stratified
Sampling
estimate total sample size needed as explained
earlier
distribute it in number of stratas equally nr
allocate in propcrtion to the size of strata
a.
or
optimum allocation to stratas where sampling
b.
fraction could be MbSh/EMbSb
estimate
' minimum or maximum
For each strata
mo^t. impo^pyop/average with which
likely value of
survey is concerned•
find samples size for each strata corresponding
-
to these values
If resources allow, for each strata choose the
larger of these sample sizes.
12/-
12
For comparisons of proportions/averages from diffe~rent stratas sample size required can be large. With
" smaller .s-ample size results will be acceptable but will
not be as reliable- as they could be using the larger
sample sizes.
.r'
Larger the number of stratas (more than 8) variance
of estimate reduces giving more’precise estimate. -Where
-stratas can not be identified in advance.
take, gne important variable under study
examine frequency;distibutior. of the variable
- determine' number of■stratas with variable values
- (if necessary examine distrl o^o^er highly
....
(v)
correlated variable)
assign the sample size to each strata based on
proportional allocation ^qual distribution.
Cluster sample size
count no. of clusters formed i.e. the no. of
clusters into which population has been divided
make preliminary estimate of likely value ®f
the most important proportion with which survey
is concerned.
look for desired cluster sample size from table
using estimate of the proportion and number »f
clusters. In this case, if population (3000) is
divided into*
i) Few large clusters..
50 (of size 60)
ii) large no.of clusters
of smaller size
30 0 (o’f size 10)
No.of clusters to
be sampled
decreases
. 15
Greater the
survey field
area
15x60= 900
No.of clusters to
be sampled
increases
21
smaller the
survey work
21x10= 210
13/-
13
(vi)
Two stage sampling
Decisions are needed for - how many clusters should
be Hampled
- how many study units should
be taken from chosen clustersi)
ii)
increase number of clusters chosen than to increase
number of study units chosen from clusters
if possible select at least three study units from
each cluster in the sample.
Steps to determine sample size
i)
ii)
Estimate the value of most important proportion of
the likely upper and lower limit
Use table to decide on total sample size 'n'
,
(study
units), if sample size thus determined is too large
for the available resources, decide about largest no.
iii)
of study units the resources allow
use table II, to decide least no. of clusters needed
based on estimated proportion and number of clusters
formed.Increase this cluster sample size by 20%
iv)
v)
Determine the no. of study units in the sample of
clusters by listing them (m)
No.of study units regd.
estimate regd. sampling fraction^ by sample(n)--------Total no.of study units
in the sqmple of
clusters (m)
empty clusters are also included.
I
INTRODUCTION TO HEALTH SYSTEMS RESEARCH
Y.P.GUPTA^
1.
System .
different disciplines/sectors are linked together
it may be called a system. A
to achieve a common objective,
it also includes
system is not merely the sum of its parts;
When
the interaction between the parts.
not merely of the boxes
Example: An organisation consists
of an organogram but also of the pattern of interIt also
relations between them and within them.
involves the clientale served and the resources
consumed
and
the services and products resulting
from the organisation’activities.
Therefores however a system may be conceived.
«
is
it
relationships
operate,
Information''flows’
how’
know
to
important
how
how
the
inter-
they are managed and
through
to
system
the
facilitate management.
2.
Health System
Health
system is the complex of inter-related elements
that contribute to health.
Health
National
System
is
made
up
certain
of
functionally inter-related components which include.
Individuals,
agents
of
families,
their
communities
as change
acting
own health care and demanding health
services.
Professor (Operations Research), National
^Associate of
Health & Family Welfare, New Delhi-11006(
Institute
-
.
.•v.
//
Health
care
delivery
centres
or
private
units
like
hospitals ^health
practitioners 9providing
health
services.
units steering and coordinating health care
Managerial
and supporting activities.
delivery
like training
of
health manpower.
Political
Institutions
formalating
health
nat ional
policy.
COMPONENTS OF HEALTH SYSTEM
/ potiTy
f Politics
i HEALTH
\ policy
»
\ n \ $T.Q A 4
/
.M
I
/COfAnONiTy
/ Ur£
\___ - ..../
I ^rPOCTU^ES
.
- ■
hnh
I I
S£^Vicl5
**.*»
MM. W
MM
c
I\
\ SUPPLY of I
I Mt ALTI
I L- k r t. &
MM N> MM
■*«
*■ MM MM n-B
X
*> • MM m MV MW
—■■ «■
MW MM ««• MV MM OTM MW «MB MM MM MM M>
W*« MM M4« «M» WW «M
.MT MM MM<
Components are not only inter-related but also interacting.
'
2
3.
Health Systems Research(HSR)
Background
w
Health Systems Research is concerned with strengthening
of
Health Care and focuses on the entire health system
or
a
part
of
s;y stem
is
optimally.
it.
Its object is to ensure that the
planned
organised and that
and’
programmes are carried out efficiently and effectively.
&
It
is
also
interest
scientific
deals
which
a
-as
understood
much
area
of
Research,
Services
Health
than
larger
with the health services component of the
broader health system.
•a
on policy. society,
Systems
Research
focuses
Administration,
process
of production. supply. demand
Health
and use of services..
&
The
scientific
and
view
interest
extensive when compared
Research
is
Health
Research
which
are
in Health Systems
to the
areas of
oriented to the Etiology,
diagnosis and treatment- of diseases.
precisely, it begins with real field.problems and
uses
a
variety
of research disciplines to apply practical
scientific- knowledge
improving health care and health
for
status. It uses methods that yield:
1)
reproducibility
solutions
(i. e .
applied
to " similar
problems and conditions elsewhere would produce similar
results);
3
ii)
ability
the
to identify the key elements of a problem
(i.e. the characteristics that make research approaches
from one setting applicable in another setting).9 and
ill) -an
understanding
of
causation
(i.e.
relationships
between actions and impact can be organised in concepts
and
theories
prevention
that
or
may
be
usefully
modification
of
applied
health
to the
problems
generally).
In
contrast
limited
to
Health
systems
indigenous
a
to Health Services Research that has been
narrow
concern
research
with government programmes 9
comprises
the
private
sector.
health practioners. the efforts of the community
itself. and inter-sectoral lunfluences on health. It focuses
particularly
on
integration
family
primary.
health
care,
including
the
of treatment and control of disease. nutrition.
planning.
environmental
sanitation and problems of
referral.
Definiation
Health
systematic
systems
research can be broadly defined as the
Investigation
and
evaluation
of
a
country 1s
health system (or a part of it) in terms of both their inter
relationship
with
measures
feasibility,
as
all
health
need.
related
factors
coverage,
and
such
effectiveness,
utilisation.
costs and efficiency. It is multi-disciplinary
in
nature,
and ideally should result in the improvement of
the
decision making process and the optimization of the use
of resources.
4
It
borrows tools and techniques from other
disciplines
and try to ppovicje
a11<rna t i ve solutions t : the problems
related to health systems as well as
methods of determining
the best/most feasible s.-lutioru
4.
Epidemiologic^; Resadrch
HSR
is not simply a branch of epidemiology
although it
often uses epi demi o1og1ca1
methods. The ep1demiologic a1
research concerns with the
study of the distribution and
determinants
population
health
of
and
health
the
problems
related
application
while
HSR
status
of
and
events
in
this study to central
is concerned with the? study of
problems in Health Cares
5.
He:? 1th Practice Research
Health
practice
organisational
research
problems
related
is
concerned
with
the
with
the
planning.
management,
Although
with
logistics and delivery of health case services.
health practice research overlaps to
some extent
epidemiology
and
often
methods, epidemiology as such
6.
makes use of epidemiological
is outside its scope.
Systems Research
This term is used with three similar meanings*
1)
The examination of various elements of a system
view to ascertaining whether the
problem
with a
proposed solution to a
will fit into the system and in turn effect
a in
overall improvement in the system.
5
1.1)
The
analysis
of
precisely
what
can
be
best
an
order to determine
accomplished.
Is
refers
to
systematically
effectiveness
Strategies
in
is required of the system and how this
ill) Systems analysis
purpose
activity
any formal analysis whose
suggest
a
examining
the
and
and
to
risks
of
course
of
action
objectives5
by
costs
alternative policies or
designing additional ones if those are
found wanting.
It is an approach to or way of looking at
problems
complex
choice under uncertainty; it is not yet a
of
method.
From
much
the
above discussion. it can be seen that HSR is
broad
based in content as well as scope in comparison
to other research methods used in health field.
7.
£ontent Areas of Health Systems Research
The
services
main emphasis in HSR should be to improve existing
incrementally .
The
practical problems car. h
1)
Definition
of
and
causes
their
types
•- ■'■e -.i i
priority
to
of situations in which
‘ ’d ar:- given Jv re under:
pr :?b lems-identifying problems
which
limited resources should
focus to do the best;
11)
Analysis
of why current health care components are not
meeting priority health needs;
ili)
Reallocation of roles and adapting technology;
iv)
Development
of
ne w
training
categories of personnel;
v)
Improving management
6
approaches
for
all
Areas for Health Systems Research
The
WHO
study
subjects 5
group (1983) has listed following nine
which it considered most important for HSRf
1.
relationship betw?3n he-iltn systems and society as
The
a whole.
2.
Assessment of Health needs
-f the population
3.
Study
and
of
the
production
distribution of health
resources.
4.
Study of the organisational structure of health
systems
5.
Study of the delivery of health services.
6.
Study of Health Systems management
7.
Analysis of the economic support of health
systems
8.
Determination of the results of health
9.
programmes
Community participation
Under
each
area,
specific
problems/issues could be
identified given hereunder:
1.
2.
Health Systems and Society
a.
Inter
sectoral coordination involving all
sectors
(agriculture, education
housing, water supply and
sanitation,
women and cchild development,
welfare,
etc.) which influence humana health and health system.
b.
Contribution of other social sectors to health
resource
production.
Assessment of health needs of the population
a
Rates of morbidity, uloabllj.r.y
disability and mortality in various
subdivisions of the population classified by geographic
area.
b.
Utilization of health services by different sections of
the population.
7
'...... -•}
i
C o
Unmet health needs of the population
d.
Per capita government and private(personal) expenditure
on health.
This will also requirej implementation of a uniform
health information system to provide relevant information
and conduction of sample household surveys from time to
time.
3<
Production and distribution of health resources
a.
Health manpower
Number and geographic distribution of medical and para
medical staff by type
Manpower planning with respect to health needs
Training of various health personnel and per unit cost
of training.
b.
Health care delivery Institutions
Total number and distribution of hospitals and. hospital
beds
Other health service units
c.
Equipment and supplies
Use of appropriate
health care delivery
technology
at
various levels of
Inventory
of all highly technical equipment with
locations
for
sharing of such resources ; by• the
neighbouring institutions
Logistics of supply of drugs^contraceptives and other
materials from centre/state HQ to the periphery to cut
the delay.
d.
4.
Coverage and gaps in the dissemination of knowledge of
medicine and public health
Health System Organisation
a.
Comparison of organisational patterns among different
states of the country
8
5.
b.
Analysis
of various health programmes/schemes
pattern of funding between centre and states.
c.
What falls in the perview of/supported, by NGOs?
and
Delivery of health services
a.
Norms and standards
Population
which
can
be
peripheral health units/staff
effectively covered
by
Standardization of equipments, supplies and drugs at
least upto block level health units
b.
Quality of health
health care)
c.
IEC for promoting healthful ’''Life styles”.
do
Time and motion studies for assessing the functioning
and workload of various.categories of health personnel
with respect to health needs of the population
cart (specially Maternal and child
Referral system
f.
6.
Analysis in terms of performance, efficiency,costs and
utilization of various departments/services provided by
hospitals and health centres.
Health Systems Management
a.
Decentralized (grass root) planning for more realistic
representat on of local health ieeds.
b.
Uniform health Information system and its periodic
review
by
both policy makers and health system
researchers to examine the relevance and adequacy of
records and reporting system initially developed in
relation to changing needs with time.
c.
Horizontal integration of various health programmes and
their actual implementation in integrated manner
d.
Management training needs of different health personnel
e.
System of monitoring and supervision
f.
Leadership
9
7.
Financing of health services
State and non-state sector
Health expenditure as percentage of GNP
Analysis of the percentage distribution
among different types of healch services
of expenditure
Health care expenditure an investment or consumption
Cost per unit of health services
C ost-effectiveness analysis of various health programmes to
determine the most cost effective strategy for providing
services
Cost-benefit analysis for optimum allocation of resources
within health sector (between different components of health
system/programmes )
8.
Community participation
The type and nature of community participation
Determination of effects of community participation on the
operation of health systems ;
Improving the prospects of effective community participation
9.
Evaluation
of
different programme
Efficiency
and
effectiveness
desired
ach. jvement
of
intervent ions/coin, onent s
in
obj ectives
Evaluation of the programme impact in terms of health status
as well as perception and opinion of the beneficiaries.
8.
Basic Guidelines for Health Systems Research
1.
Research
should focus squarely on priority problems in
health.
It
broader
context
should also consider health issues in the
each
country’s
placed
on
of
overall
socio
economic development.
2.
Emphasis
short-term
should
be
research
comparatively simple.
designs which are likely to yield
10
w
practical results when needed for decision making.
3.
In
order
should
by
to
be
promote
cost
° f f c t i v ene s s,
the focus
• n low cost studies which can be undertaken
and
managers
s arv1c e
personnel5 but supported by
experienced researchers.
4.
All
parties
staff3
concerned
community
and
designing
(health
members)
authorities s
should’
be
implementing the study
health
involved
in
in the analysis
of the reesults and the formulation of recommendations.
5.
Results
should be presented in formats most useful for
administrators 5
and
decision-makers
the
community.
The research report should include:
A clear presentation findings
Honest
discussion
of practical or methodological
problems that could have effected the findings;
Alternative
from
courses
the isuits
of
action that could follow
nd the advantages and drawbacks
of each
6.
The
effectiveness of the research undertaken should be
judged not by the number of papers published out by its
ability
to
influence
policy3
improve
services
ultimately lead to improved health in the community.
11
and
PTOPOSA™E DEVEL0PMENT °F A HEALTH SYSTEMS RESEARCH
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
Important elements
of each step
statement of the
problem
What information is
already available?
literature review
r
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
{
literature and
other available
information
- research
questions or
*- hypotheses
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
— data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
[
How will the
results be used?
Who will do what,
and when?
I
plan for
utilization and
dissemination of
results
I
What resources do
we need to carry
out the study, and
what resources do
we have?
L- pilot study
or pre-test
- manpower
timetable
work plan
L
4
- ethical
considerations
resources required
and budget
- administration,
monitoring and
L evaluation
— material support,
and equipment
money
How will we present
our proposals to
authorities or
potential funding
sources?
summary of proposal
4
I
INTRODUCTION TO HEALTH SYSTEMS, RESEARCH
Y.P.GUPTA*
1.
System .
When
to
different disciplines/sectors are linked together
A
achieve a common objective, it may be called a system.
system
is not merely the sum of its parts; it also includes
the interaction between the parts.
not merely of the boxes
Example: An organisation consists
of an organogram but also of the pattern of interIt also
relations between them and within them.,
involves the clientale served and the resources
consumed
and
the services and products resulting
from the organisation’activities.
Therefores however a system may be conceived.
is
it
relationships -operate.
inform.- tibn' ' flows-
how
know
to
important
how
the
how
inter-
they are managed and
through
system
the
to
facilitate management-
2.
Health System
Health
system is the complex of inter-related elements
that contribute to health.
Health
National
System
made
is
up
certain
of
functionally inter-related components which include:
Individuals,
agents
of
families
their
communities
as change
acting
own health care and demanding health
services.
National
^Associate
Institute
- ■?: . -
> V-
j
•*
Health
care
delivery
centres
or.
private
units
like
hospitals ^health
practitioners ,providing
health
services.
units steelin'- and coordinating health care
Managerial
delivery
and supporting activities.,
like training
of
health manpower.
formulating
Institutions
Political
health
national
policy.
COMPONENTS OF HEALTH SYSTEM I
/ Politics
i HEM-TH
>
policy
\
vxj\9r£AA
» P,.
I
'
/
/Co^noN ivy
/ MrE
,
-
■
/ rNMVirups
•»*. *-
«BM> «* «w
*«■ *«• *• Mi
sv ppi-y or
rtt/UT'V
SEA Vice 5
ac*
mm
«■»
>r I < taiZ vl
E5
\
I
'
|
/
cm m. • «a«i an m *m mm —■ ms
• w»» wm «w m> m» mo mm mb mm'mm mb mm
mm
mm mm *■»
•• «■» ■»■
Components are not only inter-related but also•interacting.
'■
2
bi
3.
Health Systems Research(HSR)
Background
Health Systems Research is concerned with strengthening
of
Health Care and focuses on the entire health system
or
a
part
of
system
is
op t ima1ly • planned
it,
Its object is to ensure that the
organised and that
and
programmes are carried out efficiently and effectively.
%
It
is
also
interest
scientific
deals
which
a
as
understood
much
of
Research,
Services
Health
than
area
larger
with the health services component of the
broader health system.
-x-
on policy, society,
Systems
Research
focuses
Administration,
precess
of production. supply. demand
Health
and use of services.
&
The
scientific
and
view
interest
extensive when compared
Research
is
Health
Research
which
are
in Health Systems
to the
areas of
oriented to the Etiology,
diagnosis and treatment of diseases.
,e ' precisely, it begins with real field, problems and
uses
a
variety
of research disciplines to-apply practical
scientific- knowledge
improving health care and health
for
status. It uses methods that yield:
i)
reproducibility
solutions
(i.e.
applied
to
similar
problems and conditions elsewhere would produce similar
results);
3
11’} • the
ability
to identify the key elements of a problem
(i.e. the characteristics that make research approaches
from one setting applicable in another setting); and
ill) -an
understanding
of
causation
(i.e.
relationships
between actions and impact can be organised in concepts
and
theories
prevention
that
or
may
be
usefully
modification
of
applied
health
to the
problems
generally).
In
contrast
limited
to
Health
systems
indigenous
a
to Health Services Research that has been
narrow
concern
research
with government programmes
comprises
the
private
sector5
health practioners5 the efforts of the community
itselfs and inter-sectoral iunfluences on health. It focuses
particularly
on
integration
family
primary.
health
care 5
including
the
of treatment and control of diseases nutrition3
planning^
environmental
sanitation and problems of
referral.
Definiation
Health
systematic
systems
research can be broadly defined as the
investigation
and
of
evaluation
a
country’s
health system (or a part of it) in terms of both their inter
relationship
with
measures
feasibility,
as
all
health
need,
related
factors
coverage 3
and
such
effectiveness 9
utilisations
costs and efficiency. It is multi-disciplinary
in
nature,
and ideally should result in the improvement of
the
decision making process and the optimization of the use
of resources.
4
It
bo.rrovp tools and techniques from other disciplines
and try tc provide
alternative solutions t : the problems
related to health systems 's well as
methods of determining
the best/most feasible scluti ...n-.
4.
.Epidemiologic;-!^ Research
HSR
is not simply i branch of epidemiology although
it
often uses epidemiological methods.
The epidemiological
research concerns with the
study of the distribution and
determinants
population
health
of
and
health
the
problems
related
application
while
HSR
status
of
and
events
in
this study to central
Is concerned with the study of
problems in Health Care,
5.
Health Practice Research
Health
practice
organisational
research
problems
is
related
concerned
with
the
with
the
planning.
management,
Although
with
logistics and delivery of health case services.
health practice research overlaps to
some extent
epidemiology
and
often
makes use of epidemiological
methods 3 epidemiology as such is outside its
9
scope.
6.
Systems Research
This term is used with three similar
i)
meanings:
The examination of various elements of
a system with a
view to ascertaining whether the
proposed solution to a
problem will fit into the system and in turn effect
an
overall improvement in the system.
5
il)
The
analysis
of
precisely
what
can
be
best
an
order to determine
accomplished.
is
to
e f f ec11veness
to
refers
systematically
strategies
in
is required of the system and how this
ill) Systems analysis
purpose
activity
suggest
.a
examining
the
and
and
any formal analysis whose
risks
of
course
of
action
objectives.
by
cu st s
alternative policies or
designing additional ones if those are
found wanting.
It is an approach to or way of looking at
problems
complex
choice under uncertainty I it is not yet a
of
method.
From
much
the
broad
above discussion. it can be seen that HSR is
based in content as well as scope in comparison
to other research methods used in healuh field.
7.
.Content Areas of Health Systems Research
The
services
main emphasis in HSR should be to improve existing
incrementally .
practical problems car b
i)
Definition
of
and
causes
their
The
types
I
J
priority
to
of situations in which
’d ar-:- given ho re under:
pr?blems-’identifying problems
’which
limited resources should
focus to do the bests,
ii)
Analysis
of why current health care components are not
meeting priority health needs»
ill)
Reallocation of roles and adapting technology;
iv)
Development
of
ne w
training
categories of personnel^
v)
Improving management
6
approaches
for
all
$
for Health Systems Research
The
WHO
study
subjects s
group ( 1.983) has listed following nine
which it consider^d most Important for HSRj
1.
relationship betw rsr. health systems and society as
The
a whole.
2.
Assessment of Health needs
’f the population
3.
Study
and
of
the
production
distribution of health
resources.
4.
Study of the organisational structure of health
systems
5.
Study of the delivery of health services.
6.
Study of Health Systems management
7.
Analysis of the economic support of health
systems
8.
Determination of the results of health programmes
9.
Community participation
Under
each
area,
specific
problems/issues could be
identified given hereunder:
1.
2.
Health Systems and Society
a.
Inter
sectoral coordination Involving
sectors
(agriculture,5» education, housing, water all
supply and
sanitation
women and child development, we1fare5
etc.) which influence human health and health system.
b.
Contribution of other social sectors to health
resource
production.
Assessment of health needs of the population
a.
Rates of morbidity, disability and mortality in various
subdivisions of the population classified by
geographic
area.
b.
Utilization of health services by different
sections of
the population.
7
c
Unmet health needs of the population
d.
Per capita government and private(personal) expenditure
on health.
1
This will also require? implementation of a uniform
health information system to provide relevant information
and conduction of sample household surveys from time to
t ime.
3.
Production and distribution of health resources
a.
Health manpower
Number and geographic distribution of medical and para
medical staff by type
•
Manpower planning with respect to health needs
Training of various health personnel and per unit cost
of training.
b.
Health care delivery Institutions
Total number and distribution of hospitals and hospital
beds
Other health service units
c.
Equipment and supplies
Use of appropriate
health care delivery
technology
at
various levels of
Inventory
of all highly technical e qu ipment with
locations
for
sharing of such resources; by- the
neighbouring institutions
Logistics of supply of drugs^contraceptives and other
materials from centre/state HQ to the periphery to cut
the delay.
d.
li.
Coverage and gaps in the dissemination of knowledge of
medicine and public health
Health System Organisation
a.
Comparison of organisational patterns among different
states of the country
8
5.
b.
Analysis
of various health programmes/schemes
pattern of funding between centre and states.
c.
What falls in the perview of/supported by NGOs?
and
Delivery of health services
a.
Norms and standards
Population
which
can
be
peripheral health units/staff
effectively covered
by
Standardization of equipments, supplies and drugs at
least upto block level health units
6.
b.
Quality of health
health care)
c.
IEC for promoting healthful hLife styles’*.
do
Time and motion studies fci* assessing the functioning
and workload of various.categories of health personnel
with respect to health needs of the population
eo
Referral system
f.
Analysis in terms of performance, efficiency,costs and
utilization of various departments/services provided by
hospitals and health centres.
care (specially Maternal and child
Health Systems Management
a.
Decentralized (grass root) planning for more realistic
representat on of local health ieeds.
b.
Uniform health information system and its periodic
review
by
both policy makers and health system
researchers to examine the relevance and adequacy of
records and reporting system initially developed in
relation to changing needs with time.
c.
Horizontal integration of various health programmes and
their actual implementation in integi’ated manner
d.
Management training needs of different health personnel
e.
System of monitoring and supervision
f.
Leadership
9
7.
Financing of health services
State and non-state sector
Health expenditure as percentage of GNP
Analysis of the percentage distribution
among different types of healch services
of expenditure
Health care expenditure an investment or consumption
Cost per unit of health services
Cost-effectiveness analysis of various health programmes to
determine the most cost effective strategy for providing
services
Cost-benefit analysis for optimum allocation, of resources
within health sector (between different components of health
system/programmes)
8.
Community participation
The type and nature of community participation
Determination of effects of community participation on the
operation of health systems
Improving the prospects of effective community participation
9.
Evaluation
of
different programme
Efficiency
and
effectiveness
desired
ach. jvement
of
interventions/coin. onents
in
objectives
Evaluation of the programme impact in terms of health status
as well as perception and opinion of the beneficiaries.
8.
Basic Guidelines for Hea 1th Systems Research
1.
Research
should focus squarely on priority problems in
health.
It
broader
context
should also consider health issues in the
each
country’s
placed
on
of
overall
socio
economic development.
2
Emphasis
short-term
should
be
research
comparatively simple?
designs which are likely to yield
10
w
»
I
practical results when needed for decision making.
3.
In
order
should
by
to
be
co-st
orbmote
effectiveness ,
the focus
■ n low cost studies which can be undertaken
and
managers
fJUl-VlCt:
persenna1, but supported by
experienced researchers.
4.
All
parties
concerned
(health
authorities,
health
41
staff.
community
designing
and
members)
should'
be
involved
in
implementing the study. in the analysis
of the reesults and the formulation of recommendations.
5.
Results
should be presented in formats most useful for
administrators 9
decision-makers
and
the
community.
The research report should include:
A clear presentation findings
Honest
discussion
of practical or methodological
problems that could have effected the findings;
Alternative
from
courses
the icsults
of
action that could fellow
nd the advantages and drawbacks
of each
6.
The
effectiveness of the research undertaken should be
Judged not by the number of papers published out by its
ability
to
influence
policy.
improve
services
ultimately lead to improved health in the community.
11
and
pJoposa™E development of a health systems research
Questions you
must ask
What is the problem
or what is to be
studied?
What information is
already available?
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
Steps you will
take
Important elements
of each step
statement of the
problem
literature review
formulation of
objectives
1
literature and
other available
” ’■
information
- research
questions or
*- hypotheses
r- variables
research
methodology
___ - type of study
- data collection
techniques
- sampling
- plan for data
collection
t
How will the
results be used?
Who will do what,
and when?
4
What resources do
we need to carry
out the study, and
what resources do
we have?
plan for
utilization and
dissemination of
results
- plan for data
analysis and
interpretation
- ethical
considerations
L pilot study
or pre-test
- manpower
timetable
work plan
resources required
and budget
- administration,
monitoring and
L evaluation
— material support,
and equipment
*- money
How will we present
our proposals to
authorities or
potential funding
sources?
summary of proposal
4
I
G3 HO
CH
CO
ro
o
co
co
o
U1
■b.
GO
ro
SL. NO.
>
— O
*< o
CD CD
co co
r*
<
O
O
CO
"O
8£
ZT
CD
<
CD
CD
Z3
>
m
zi
ZT
>
O
Z5
2
AGE
2.
s
CD
Q.
w
V)
f
O Q
agZT O’
O
C
CZ>
CD
ZT
O
CQ
CD
CD
S’’
O
ZJ
2
TJ
O
CO
H
O
0)
ZE
m
u
c:
m
o
o
z
o
2
SEX
£o
c
3
CO
O
O
O
m
OT
CD
Cl
CX
Z
*<
ex
>
“H
RELATIONSHIP TO
HEAD OF FAMILY
o
i
co
H
>
CD
H
C
CO
g
ao
MARRIED/
UNMARRIED
t
EDUCATION
5
Z3
ZJ
£D
c
Q
3
£0
H
w
O-
OCCUPATION
INCOME
(ANNUAL)
REMARKS
d
ZT
W’
3
JD
ZJ
Q.
£0
o
cr
□:
o
c:
w
m
□:
o
o
c
co
5O
CD
/
1
4
2.
MATERNAL & EARLY CHILDHOOD HISTORY - DEMOGRAPHIC FACTORS
2.1
Fertility Status
1989-90
2.1.1
Number of pregnant women
2.1.2
Number of women delivered
3.
FAMILY PLANNING/WELFARE PRACTICE
3.1
Does married women or her husband use
any family planning method?
3.2
If yes, specify the method
3.3
If not, why?
4.
DIETARY HABITS
Period
1990-91
Yes/No
4.1 Is the family
Vegetarian/Non-vegetarian
4.2 What is family's staple food?
Rice/Ragi/Wheat/Other
4.3 Does the family consume pulses?
Yes/No
4.4 Does the family consume Milk/Milk Products?
Yes/No
4.5 Does the family consume green Vegetables?
Yes/No
4.6 Does the family provide special food during pregnancy?
Yes/No
4.7 If yes, specify the items
4.8 When does the mother start breast feeding the baby?
a. Soon after birth
b. After 48 hours
c. Later than 48 hours
4.9 When does the mother start feeding the baby solids?
a. After 3 Months
b. After 6 Months
c. Later than 6 Months
4.10 Specify the solid items fed to the baby
Current
•>
5.
1989-90
1990-91
Current
1989-90
1990-91
Current
MORTALITY STATUS
5.1 Number of babies born dead
5.2 Number of babies died within 1 week of birth
5.3 Number of babies died from 1 week to 4 weeks
5.4 Number of babies died from 1 month to 12 months
5.5 Number of pregnant women died
a. During pregnancy
b. During delivery
c. Within 6 weeks after delivery
5.6 Number of Children (0-3 Years) died during
6.
CAUSES OF DEATH (Children 0-3)
6.1 Due to ARI
6.2 Due to diarrhoea
6.3 Infection/TB/Whooping Cough/Brancho Pneumonia
6.4 Due to accident/injury
6.5 Prematurity
6.6 Malnutrition/Low birth weight
7. CAUSES OF DEATH (Pregnant Women)
7.1 Anaemia
7.2 Toxaemia
:
Pre-eclampsia
Eclampsia
7.3 Haemorrhage
7.4 Obstructed/Prolonged labour
7.5 Infection
7.6 Others
Enumerator’s Name
5
SCHEDULE - 'C (WOMEN)
Married Women Upto 45 Year age
1.
GENERAL INFORMATION
Identification No.
of Household
1.1 Name of the village
1.2 Informant/s name
1.3 Age
11.4 Married/Unmarried
M/UM
1.5 If Married husband's name
1.6 Educational level
Primary / Secondary/Higher/College
1.7 Occupation
Housewife / Employed
1
1.8 No. of children
2
3
4
5
Sex
Age
Q
1.9 Is the marriage consanguinal?
Yes/No
1.10 If Yes, type of marriage
UN/MUS/PUS
6 1.11 Any family history of disabilities
1.12. Any history of
2.
Visual/Hearing/MR/Physical Disability
Epilepsy / Behavioural Problems
MENSTRUAL HISTORY
2.1 Age at menarche
2.2 Age at marriage
c-
2.3 LMP
3. HISTORY OF EARLIER PREGNANCIES
3.1 Age at the first pregnancy
3.2 No. of previous pregnancies
n 3.3 Spacing between subsequent pregnancies
3.4 Antenatal supervision by ANM
(Last Pregnancy)
0 3.5 Number of visits by ANM (Last Pregnancy)
Yes / No
172/3/4/5
3.6 Have any of the following administered?
(Last Pregnancy)
Test
Supplementation of
Blood/Urine
TT / Iron & Folic Acid
/
3.7 Any Medication taken during pregnancy
Yes / No
3.8 If yes, specify
3.9 Any habits
Smoking/Tobacco chewing/Drinking
3.10 Whether suffered from, during pregnancy
Diabetis/Convulsions/Anaemia/Hypertension/Taxemia/
Oedema
3.11 If yes any referral services rendered at
Sub-centre/PHC/Hospital/Private Clinic/Native Medicine
3.12 Any significant occurance during previous
pregnancy (Accident / Injury)
Yes/No
3.13 If yes, specify
HISTORY OF LAST DELIVERY
4.
(b) Before term
4.1 Delivery status
(a) at term
(c) More than 40 Weeks
4.2 Place of delivery
Home/Sub-Centre/PHC/Hospital
4.3 Conducted by
TBA-Trained/Untrained/Relative/ANM/Doctor
4.4 Type of delivery
Normal/lnduced/Forceps/CS
Hours
4.5 Duration of labour
4.6 Any complication during delivery
Yes/No
4.7 If yes, specify details
5.
POST NATAL PERIOD
5.1 Any visit made by ANM/LHV during
first ten days
Yes/No
5.2 Any subsequent visits made by ANM/LHV
during SIX weeks
Yes/No
6.
HISTORY OF PREGNANCIES
6.1 Have any of the following occured
6.1.1 Abortion
6.1.2 Premature Death
6.1.3 Still Birth
6.1.4 Neonatal Death
9
Previous
Current
SCHEDULE -’D' (CHILD)
0-3 Years completed as on 1st September 1991
Identification No. of Household
J
1.
GENERAL INFORMATION
1.1
Name of the child
1.2
Sex
1.3
Date of birth & age
Male/Female
d.4
Weight at birth
Normal/Below Normal
1.5
Place of delivery
Hospital/Home/Others
1.6
Name of the hospital / others
1.7
Delivered by whom?
2.
CONDITION IMMEDIATELY AFTER BIRTH
2.1
Did the baby cry well /cry weak/not cry
2.2
Did the baby breathe well/breathe with difficulty / not breathe
2.3
Was the baby blue in colour
2.4
Was there any evidence of malformation?
2.5
What methods were used to
resuscitate the baby?
2.6
Did the baby have convulsions
Home/ANM/TBA/Relative
Yes/No
2.6.1 If yes, when?
3.
DEVELOPMENTAL HISTORY
Normal
Delayed
3.1 Reaction to sound
3.2 Head control
J
3.3 Recognition of mother
3.4 Turning
3.5 Crawling
3.6 Sitting
3.7 Walking with aid
3.8 Walking without aid
3.9 Teething
3.10 Talking
3.11 Bladder control
3.12 Awareness of surrounding
■t
4.
UIP + Status
4.1 BCG
Yes/No
4.2 Polio Vaccine
1 st/2nd/3rd
4.3 DPT
1st/2nd/3rd
4.4 Measles
Yes/No
4.5 Booster doses (DPT)
Yes/No
Given age
Has there been any incidence of diarrhoea
during past six months?
Yes/No
If yes, has he/she been given oral rehydration
during past six months?
Yes/No
4.8
Did he/she suffer from rapid breathing?
Yes/No
4.9
If yes, was he/she treated?
Yes/No
4.10
Were the children given
a. Vitamin 'A' supplementation
b. Iron & Folic Acid
Yes/No
Yes/No
4.6
4.7
5.
NUTRITIONAL STATUS
6.
EARLY CHILDHOOD HISTORY (0-3 Years)
f
Good/Average/Poor
6.1 Any abnormality/Disability
Visual/Hearing/MR/Physical Disability
6.2 Any assessment done by any Agency/lnstitution
Yes/No
6.3 If yes, give details of the Agency/lnstitution
6.4 Any external support received
Yes/No
L
6.5 If yes, specify nature of support received
*
>i
i
CHAPTER 3
DESIGN OF A SAMPLE SURVEY TO
MEASURE CHILDHOOD MORTALITY
F
The two ways we control chance error in a sample survey are by choosing the
sample to interview in an unbiased way and by assuring that the number of
respondents is large enough for our estimates to be as precise as we would
I
I
like them to be.
i
p
i
Some variant of probability sampling must be used to select the sample
respondents. This means that in theory all individuals in the target population
(in surveys of childhood mortality, usually all women, or all ever-married
women, aged 15-49) have a known chance of being interviewed and that the
selection of one individual is independent of the selection of another individual.
Other types of samples chosen on less objective grounds, such as the judgement
of the interviewer, field supervisor or project director, or samples drawn from
• particular groups only, such as mothers coming to clinics, are not appropnate
for determining child mortality for a whole community or population. Samples
such as these can result in biased estimates, that is estimates of the true value
which have been shifted in one direction or another from the true value because
the sample is different in some systematic way from the whole population of
interest. It is not possible to measure the amount of error in the estimates
resulting from this kind of sampling bias, so estimates from such samples cannot
be assumed to represent the true values in the population.
In order to choose any form of probability sample, a sampling frame is
needed. This is a list of all the units (usually districts, households or families) in
the population which will form the basis of sample selection and which are each
given a unique number. This can be all individuals in the target populauon, if
the study is concerned with a small area where this information is available.; It
is more likely to be a list of houses, households, towns or villages from which a
sample will be selected. The amount of time and effort available for compiling
this list will determine the size of the units to be used as the Pi imarv
I
44
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
As we shall see, these will usually be districts or villages, with households
selected at a second stage of sampling. The point to remember is that all such
units in the area of interest must be included in the sampling frame for the first
stage of the sample. If national estimates of mortality are required, then the
frame must be national, including for instance, a list of all towns, villages and
cities or districts of cities in the country. If some areas which are difficult to
survey are left out of the sampling frame, then estimates should be reported as
i
excluding those areas.
-
Simple random sampling is the simplest form of probability sampling. The
units are chosen singly by taking random numbers generated by a calculator or
from a table and interviewing the person or household unit with that number on
the list. Systematic sampling, where every nth (ie. 10th or 50th) person on the
list is chosen, is another method of choosing a random sample, as long as you
are sure the list you are using is not ordered in any periodic way (for instance,
male, female, male, female ...) which might bias your sample.
&
K.
a
■
.*
Cluster Sampling
Where lists of individuals or households are not available for use as a sampling
frame for a simple random sample, even for sub-areas, or when the cost in both
time and money of using them to identify individuals in the sample would be
prohibitive, a procedure for choosing respondents from groupings or clusters
can be used. We will use the term "cluster" in its standard sampling sense to
mean a grouping within the population, such as a village or district, from which
a sub-sample may be selected, and not in its EPI usage as that sub-sample itself.
The term "village" may be interpreted as an urban block or enumeration
district or whatever grouping is appropriate.
Cluster sampling reduces costs and simplifies field work in surveys, but violates
the assumption made in random sampling that each respondent is chosen
independently of the others. Each dwelling is still chosen at random within the
primary sampling unit, which may be some geographic or political subdivision.
■■■■
45
Since these areal divisions will usually vary considerably in size, these
differences have to be controlled by choosing the first units with probability
proportional to size from the frame using random numbers. This method
ensures that, if equal numbers of households are selected from each cluster,
each household will have an equal chance of selection. The sample is then said
to be self-weighting. The first dwelling unit is usually identified at random
and further units are identified according to a pre-arranged protocol. In areas
where separate dwellings are difficult to identify such as shanty areas or
multi-household dwellings, it is best to define a geographic area in the protocol
and interview all households or individuals falling within the defined area. For
instance, if the house chosen is a multiple dwelling, all households in that
dwelling should be visited.
The use of this sampling strategy reduces costs of listing for sampling frames,
reduces travel and time costs, and makes easier the field supervision of
interviewers and survey administration. This is important, as the quality of the
data collected is likely to be much better if field workers are more closely
supervised.
Stratified Sampling
*
I
This method may be useful if the mortality of particular sub-groups is of
interest. In order to sample large enough numbers in the sub-group to keep
sampling error small and yet still allow for a separate analysis of these groups,
you may need to set up a separate initial sampling frame in each of the
sub-groups of the population under study. Then, you can choose an independent
cluster sample as above from each sub-group. These sub-groups are known as
strata. The size of the sample in each sub-group (stratum) can be increased
independently so that enough cases are available for analysis.
For example, a particular region of the country or urban areas may be of
particular interest, but using a straight forward cluster sampling procedure
would not yield a sufficiently large sample of these individuals for separate
I
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
46
analysis In this case the minority group should form a separate stratum from
.which a larger sample may be taken. When the estimates for the whole
population are made the results from each stratum will need to be weighted
results. It is clear from experience in surveys around the world that th
women who are harder to find are different from other women. Thu:
we would miss recording their child mortality experience, which is als
likely to be different from the women who are found. They may be mo;
educated women, who work at office jobs during the day or They may t
poorer women, who must leave their family in search of work far awa;
or different in some other way from the women who remain at home.
I
accordingly.
Stratifying your sample should only be undertaken when the variable you are
stratifying is easy to observe and work with. For instance, urban/rural
residence or region of residence are usually more easily identified than some
•
measure of social status such as occupation or income.
Modifying EPI Surveys for Purposes of Mortality Estimation
Tro/T... I.
-
r
Experience shows that proxy reports (that is, from someone other than tl
eligible woman) are not as good as reports from the woman herself
Proxy reports should not be allowed until at least one further attempt h<
been made to find the woman, and then only from the woman s mothe
sister (or possibly husband, if marriages are fairly stable) also residing .
the household.
The cross-sectional survey design used by the Expanded Programme o
Immunization of WHO to estimate vaccination coverage is an example of a
survey design which might be modified for the purpose of estimating childhood
mortality. Free-standing surveys of child mortality may also be conducted
using a similar cluster sample design. This is a practical solution for most
surveys conducted in countries where taking a simple random sample of
individuals across the country would be too expensive and difficult to carry out.
If the cluster sampling design is to be used for mortality estimation, the sample
size and possibly the number of clusters in the sample will need to be
modified, as will some of the field procedures. The selection of households is
47
I.
Objectives of the Survey Define Sampling Units
I
!
particularly important:
The ideal sampling unit for childhood mortality surveys in countries with lit!
or no extra-marital fertility would be ever-married women aged between
and 49. The aim is to measure the child mortality experience of these womer
However, we define our basic sampling units (bsu) as the household ai
interview all eligible women in the household. One reason for this is that mt
EPI surveys have multiple aims, and household members other than mam
women may need to be interviewed. It is also more likely that a sampli
frame, or list from which to draw the sample, already exists for househol
rather than for individuals.
The survey cannot be confined only to mothers with living children, nor
to mothers with living children of certain ages, as this would bias the
results for estimation of overall mortality levels. Mothers with young
children alive will have lost fewer of their previous children than average.
•
The households chosen randomly to visit should be re-visited if all the
eligible women usually residing there are not at home. If the household is
simply skipped and another household is substituted, this could also bias
Zj
The quantities we wish to measure are usually expressed in terms
proportions:
Number of children living
Number of children ever bom to women in the sample (7 age groups)
Number of preceding-born children dead
Number of total preceding-bom children
48
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
The relative size of villages is more important than their absolute size, so that
even an out-of-date census will be useful as long as some allowance is made for
variations in population growth by village since the census was taken. If the
original sampling frame is out of date, it is best to enumerate a fqw clusters
completely, counting all persons in each household, to get some idea of how the
population and household composition may have changed. This information can
be used to adjust the original sampling frame used to draw the PPS sample.
We want to estimate these proportions which form the basis for the mortality
estifnates within certain limits of precision which we choose in advance. For
example, we might require the sample estimate of the proportion of
previous-born children dead for mothers with a birth in the last two years to
fall within 1 percentage point of the true proportion.
In a survey in which the household was the bsu, both the numerator and the
denominator of such ratios would be unknown quantities until the survey had
been carried out. Both would be different if a different sample of households
had been selected with different numbers of women in each household. This
variability in the denominator would diminish the precision that we could assign
to our estimate of the proportion. We shall simplify matters by neglecting the
variability in the denominator, as is done implicitly in many surveys.
I
Sampling with probability proportional to size (PPS) is then carried out.
This is done by creating a cumulative list of village populations and selecting a
systematic sample using a village chosen at random to start. For example,
suppose you need to take a sample of three villages (clusters) from the list of ten
villages shown in Table 3.
Selecting the Sample
Selection of the sample may be done in several stages: for example, a country
may be split into regions, a number of districts chosen from each region, a few
villages from each district and a number of households from each village. The
basic principles for deciding sample size and structure, and the methods for
estimating rates and their standard, errors, are all the same. They will be
demonstrated first for the simplest situation where a selection of villages is .
made directly within some country (or region), and estimates are obtained for
that country.
Selecting the Clusters
You need to have a list of all villages (or districts) in the region where the
survey is to take place. An estimate of the number of households in each village
is also necessary. If the average size of households does not vary greatly from
one village to another, then the number of households per village will serve as
an index of population size.
49
Table 3
Cumulating Village Populations
r
Village
Population
Size
Cumulative
Population
1
2
3
4
5
6
7
8
9
10
1000
400
200
300
1200
1000
1600
200
350
450
1000
1400
1600
1900
3100
4100
5700
5900
6250
6700
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
50
51
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Divide the total population of the villages (6700) by the number of
villages to be selected (3) to obtain the sampling interval (k)
(k=6700/3=2233). Choose a random number between 1 and 2233 (using a
calculator which generates random numbers or a table of random numbers).
Suppose this number is 1814. This should be located in the list to identify the
first village in the sample. Since 1814 lies between 1601 and 1900, village 4
will be chosen. Now add the sampling interval to the initial random number:
1814 + 2233 = 4047, and so village 6 is chosen. Add the sampling interval
again: 4047 + 2233 = 6280 and village 10 is chosen.
Sometimes a district may have a population which is larger than the sampling
interval in which case it may be selected twice. Two independent samples of
clusters will then be selected from the same district. If a large city is part of
the population to be sampled, then a further list of districts within the city may
be needed, with population estimates, so that the many clusters to be chosen
from the city can also be chosen randomly. It is not correct to simply select
another district when one district is selected twice, or to repeat the whole
sampling procedure until no villages are repeated. This would invalidate the
principle of selection with probability proportional to size.
If a constant number of households is selected within each chosen village, then,
overall, each household in the population will have an equal probability of
being in the sample. This kind of sample is said to be self-weighting and
»
K
‘
If this is impractical,
theni some other way must be used to ensure that the
sample selected is as random as possible. The standard EPI procedure for
choosing the first household as a starting point and for selecting succeeding
households after that is suitable (WHO, 1988; 1989). The procedure is to select
some central point in the town or village, choose a random direction from that
point, count the number of households between the central point and the edge of
town in that direction, and pick one of these houses at random as the starting
point of the survey.
&-
k
h. '•
I
■■
-r;
In large districts it would be best to spread the sample around by having more
than one starting point.
The remaining households in the sample after the first should be selected to give
as widespread coverage as is possible and practical. It is permissible, but not
desirable, to go from one door to the nearest neighbour, as is done in EPI
surveys. This is likely to be adequate for purposes of EPI coverage (where
children of the required age are found in only a small proportion of the
households visited). It is not ideal for estimating mortality, where most
households will have at least one and possibly more eligible women. It is more
important to try to distribute the sample more widely, taking say, every fifth
household, or selecting all the households completely at random.
Any method which achieves random or near-random selection of households,
spread as widely as possible across the village or district, is acceptable.
simplifies the analysis.
Selection of Households within the Clusters
The ideal procedure for the selection of households within the chosen clusters
would be to have a list of all households in the village and to choose a selection
from the list at random. If such a list does not exist, and if the village is small,
then a list can be created by carrying out a quick census or possibly by
consulting administrative records.
If the area from which you are sampling contains a few multiplerunit dwellings
then it is wise to count the entire structure as one bsu and interview all women
who usually reside in the structure, even if they are from different households.
This then eliminates the problem of deciding which household within the large,
structure to include: people living on the ground floor may be different fron
those living on the top floor, though the bias from this may be only slight. It
areas which consist entirely of high-rise or multi-occupancy dwellings, some
revisions to this approach will be necessary to assure PPS sampling.
52
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
If we can state the precision with which we want to estimate mortality, we can
use standard formulas to calculate the necessary size of our sample so that we
know how much chance error can influence the results. The precision of the
estimates will depend on the size of the sample and the amount of clustering
Determining Sample Size
In deciding on an appropriate sample size for a survey one is faced with the
need to strike a balance between precision and cost. Ideally, one should first
decide on the precision needed and calculate the sample size accordingly. In
practice, however, time and money are always limited and often the best one
can do is to calculate what sort of precision can be achieved with the resources
available. If the achievable precision is poor, then perhaps the survey should
of mortality.
The usual way to measure the precision of an estimate is by its standard
error. We construct a 95% confidence interval around the true value from:
not be carried out at all.
The drawback to using cluster sampling is that sampling error in the estimates
is likely to be larger than that from a strict random sample due to the violation
of the independence of respondents in the sample. The effect of this clustering
on the sampling error will depend on the variability of the proportions of
children dead among those in the cluster compared with its variability in the
whole population. If women sampled from the cluster are more like each other
in regard to their children’s survival than they are like women in the total
population (if they are more homogeneous than is the population at large),
then the error in the estimates of sampling from clusters will be increased.
This is called the design effect. The size of the sample can be increased to
the estimate ± 2 standard errors.
If we denote the proportion of positive responses expected to a question (say, is
your preceding-born child still alive?) by p and the total number of responses to
this question in the survey by n, then the sundard error(s) of am estimated
proportion p may be written as:
■•1
I
rh
compensate for this.
It is very hard for an investigator to make an estimate of the extent to which
clusters sampled will or will not be homogeneous with regard to child
mortality. In general, variation in the measures of mortality in the clusters are
not likely to be very different from that in the total population. This means that
the effects of cluster sampling on sampling error are not likely to be very large.
For the purposes of mortality estimates, increasing sample size by a factor of 2
will serve for most cases when calculating the necessary size for a cluster
sample (Dowd, 1986, Rothenberg, et al, 1985).
53
J
where D = the design effect.
The cost and amount of work involved in taking a sample of a given size can
then be weighed against the precision required for the estimate. In general, the
larger the absolute size of the sample, the smaller will be the error due to
chance in the estimate. The absolute size of the sample is more important than
the size of the fraction of the population sampled.
If a sample of a given size is distributed in numerous, small clusters, estimates
will be more precise than if the same sample were distributed in a few large
clusters.
The number of clusters to be used is determined by the precision required and
by the amount of variability between clusters expected to be encountered in the
variable(s) of interest, eg. numbers of dead children. If you have reason to
54
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
believe that levels of child mortality are distributed very unevenly in the
population, that is, that it is very likely that clusters will be quite different from
each other with regard to mortality, then the number of clusters used must
be increased. Increasing the number of clusters in the sample increases the
precision of the estimates even when the number of individuals in each cluster is
reduced. The 30 clusters used in EPI surveys is based on conventional practice
and practical considerations, and can be increased in light of your own survey s
needs. [If in doubt, increase the number of clusters sampled].
For a sub-group of the population, the error of the estimate is determined in
large part by the size of the group. If it is necessary to produce estimates by
region or by age groups, then for a given sample size the error in the estimates
of the sub-groups will be larger than for the total sample and will depend on the
size of the sample for that sub-group.
As an example, we will calculate our sample size for a survey based upon
estimating mortality using the Preceding Birth Technique, since this estimate
demands the largest numbers of women for mortality estimation.
We need to specify:
1. The anticipated proportion of preceding-born, children dead to
mothers with a last birth within two years (whom we will refer to as
"index" women). For Jordan, we anticipate this proportion to be .05,
based on a guess about the probable recent level of mortality.
I
Since we must measure proportions of children ever bom alive at the time of
the survey by five-year age-groups of mothers, proportion of preceding-bom
children dead, and also proportions of children surviving to various young
ages, we need to keep the size of these groups in mind when calculating sample
2. Number of preceding births to women with a more recent birth (say
within the last two years).
If there are multiple proportions of interest as in this survey, you should
determine your sample size based on the one which requires the largest
numbers or most precise estimate. In the kind of mortality survey we are
planning, this will usually be the proportion of preceding children dead to
mothers with a subsequent recent birth, since we know that we will need more
mothers overall to find the necessary sample of these mothers.
2. The level of confidence with which you wish to make your estimate. It
is common to set this to 95% in most investigations of this sort. (That is,
you can be confident that 95 samples of 100 taken from the same
population will yield estimates of the true population proportion within
the range specified).
< -
size:
1. Numbers of mothers in each five year age group, and their expected
number of children ever bom.
55
t
-
••
■.
■
■
-3. The absolute precision with which you want to make your estimate.
For example, we wish to estimate the proportion of preceding-born
children who died, which we guess to be about .05 or 5%, to within 1.5
percentage points. We want to say that our estimate of children who have
died lies between 3.5 and 6.5 percent of preceding births. (That is, that
our estimate would lie somewhere between 35 and 65 deaths per 1000).
4. Some estimate of the design effect, if you are using cluster sampling. This
depends on the homogeneity of mortality measures. From other surveys
of similar design and objectives, we estimate the design factor to be
about 2.
I
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
56
57
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Then:
The formula for calculating the sample size of a cluster sample then becomes:
c
e
n
4p (1 - p) D
dxd
n
T
For example, if the sample size is 1689 and 50 women can be interviewed in a
day, we would need to survey 34 clusters from the sampling frame.
Where p is the percentage to be estimated; D is an estimate of the design effect
and d is equal to the precision required (an estimate of how close to the true
proportion we want our estimate to lie). The factor 4 results from the use of
1689
= 34
"50"
the 95% confidence interval.
i
Using the formula above, we can calculate our sample size:
■
(
■
n
4 x .05 (.95) 2
.015 x .015
4 x .0475 x 2
.000225
1689
Sample size tables appear in Appendix HI for the 95% confidence level and a
range of levels of precision. The sample sizes in Table A are those appropriate
if your sample is a random one, and those in Table B are appropriate for a
i
i ■
1689
210
cluster sample with an estimated design effect of 2.
I
To determine the number of clusters to use in the survey, divide the total
sample size by the cluster sample size. It is best to set the cluster sample stze to
the number of women one team can interview in one day. In any case, do not
i-
use a cluster sample size greater than 100 women.
We can then calculate the number of clusters to use:
If c = the number of clusters to be used
b = cluster sample size (number of women to be interviewed in one day
by one team)
n = total sample size
Sometimes the number of clusters to survey has already been set. For example,
in Jordan the EPI survey consisted of 7 separate surveys of 30 clusters each, a
total of 210 clusters. With this number of clusters, 8 "index” women from
each cluster were required to reach the total sample size.
i
8 '
To estimate how many women in the age range will be required in order to
interview 8 women with a birth in the last 2 years, we need to know something
about the level of recent fertility. In high-fertility countries, such as most
countries in the Middle East, about 15% of women of reproductive age can be
expected to be giving birth in any given year.
So, if 8 = desired cluster sample size of "index" women
.15 = proportion of women giving birth in a year
Then:
8
= 27 women 15-49 will need to
.15 x 2
be interviewed in each cluster to
find 8 "index" women for the
PBT estimate.
The household is the basic sampling unit, so if each household averages one
woman of reproductive age, you will visit 27 households in each cluster.
58
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Chapter 3: Design of a Sample Survey to Measure Childhood Mortality
Then all women in the households visited (15-49 or ever-married) will be
interviewed.
If you expect to encounter many women in each household, you should consult
a statistician for advice in drawing the sample.
Requirements of a Good Sample
I
It is large enough to give estimates of desired precision.
1
Basic sampling units are selected in an unbiased manner.
In low fertility countries, it may be necessary to interview far more women to
find the number of "index" women needed.
In Lima, Peru, for instance, 7500 women between 15 and 50 were interviewed
to find 1300 index women for this estimate. In Jordan, about one-third of all
women in this age range were index women for the PBT estimates (that is, had
had a birth within the last two years and at least one previous birth). This was
almost one-half of the ever-married women interviewed.
In EPI surveys, the search for mothers of 7 children aged 12-23 months will
more than likely yield sufficient numbers of index women for this estimate if
enough clusters are to be included in the survey.
If the sample size is impractical within the constraints of your survey, then you
must be willing to allow a wider confidence interval, say precise to within 3 or
4 percentage points. If you are estimating what you think is a very small
proportion, this may not be precise enough. If you think that child mortality
has actually fallen to a level as low as, say, 50 per thousand by age two, then
estimating this proportion of preceding children dead, .05, to within 4
percentage points with 95% confidence, will only tell you that between 10 and
90 children per thousand die by age two. This range is unacceptable. Thus, it
may not make sense to carry out the survey unless you can afford to make your
estimate more precise by enlarging your sample.
Mi
1.
Make a random start, preferably using two different starting
point.
2.
Spread household selection as widely as possible through the
cluster ie. every 5th house rather than the next nearest door is
selected.
p
3.
Re-visit absent women or empty households.
te
I
i
•
•
Minimize absences by good timing of fieldwork.
Keep track of households and women missed.
Remember: Random selection of household does not mean haphazard
selection. More clusters of fewer women is better than
fewer, larger clusters.
59
►
WORLD
HEALTH ORGANIZATION
REGIONAL OFFICE FOR SOUTH EAST ASIA
ORGANISATION MONDIALE DE LA SANTE
BUREAU REGIONAL DE L'ASIE DU SUD-EST
Tel.; 331 7804 - 23 Telex: 31-65031 & 31-65095
RDOC(Docs)/89
18 December 1989
Your reference: Letter CM/5618/89 dt. 28.11.1989
Please find enclosed a copy each of the following
two documents, as desired:
1. Health Systems Research Training Course: How
to develop research proposals to solve priority
health problems, and
2. Health Systems Research Training Course:
Trainers Notes
Mr AS MOHAMMAD
Asst. Professor
Department of Community Medicine
St. John’s Medical College
Bangalore-560 034
This material is sent to you
with the compliments of the
Reports and Documents Unit of the World Health Organization
Regional Office for South-East Asia
World Health House, NEW DELHI 1.10.002, India Telegr.: WHO NEW DEHLI
World Health House, NEW DELHI 1.10.002. Inde Telegr.: WHO NEW DELHI
MM
HEALTH SYSTEMS
RESEARCH TRAINING
COURSE
TRAINERS NOTES
Adaptation for district level of the Health Services Research Course,
developed by the WHO Regional Office for Africa and the
Project for Strenghthening Health Delivery Systems in
Central andWest Africa
(WHO/AFRO-SHDS)
Joint WHO/DGIS/RTI Project on Health Systems Research
WHO Subregional Health Development Office III
Harare, Zimbabwe
PREFACE
The present volume contains directions for administrators and facilitators conducting the Health Sys
tems Research Training Course originally developed by WHO/AFRO/SHDS and adapted for provin
cial and district level at the initiative of the Joint HSR Project. The same multidisciplinary team of
researchers that developed the modules also developed the Trainers Notes. (For details please
consult Preface and Acknowledgements to the Health Systems Research Training Course).
The volume consists of three parts:
1.
Introduction to Trainers Notes
This part deals with management issues in the preparatory phase of the course, as well as
management issues to be considered during its implementation. Thirdly, it provides infor
mation on the training methodology underlying the course and on how to apply this meth
odology when presenting the modules.
2.
Introduction to the Course
Part 2 comprises of a questionnaire and two tests to assess participants’ knowledge of and
experience with HSR, as well as guidelines on how to use the tests. It further contains sug
gestions for a mutual introduction of participants and facilitators at the start of the course.
3.
Trainers Notes to the Modules
These entail more detailed information on how to present the modules and how to facilitate
the group work.
For the development of the HSR Modules (Vol I) as well as of the Trainers Notes (Vol. II), the Royal
Tropical Institute, Amsterdam, provided technical support. Both volumes were printed at WHO HQ
Geneva, with financial support from WHO AFRO and the Netherlands Ministry for Development Co
operation.
11 November 1988
Joint WHO/DGIS/RTI Health Systems Research Project
WHO Subregional Health Development Office
Harare, Zimbabwe
TABLE OF CONTENTS
Page
1
Introduction to Trainers Notes
Course Preparation
1
Selection of Course Management Team
1
Tasks of the Course Management Team
in Course Preparation...........................
2
B.
Course Management
6
C.
Training Methodology
9
A.
12
Introduction to the Course
1.
Questionnaire
13
2.
Mutual Introduction of Course
Participants and Facilitators ...
15
Assessing Course Participants’ Knowledge
on HSR Before and After the Course.........
15
3.
36
Trainers* Notes to the Modules
Orientation to the Course
37
2.
Introduction to Health Systems Research
39
3.
Selecting a Research Project
41
4.
Statement of the Research Problem
44
5.
Review of Available Literature and
Information..................................
45
6.
Formulation of Research Objectives
47
7.
Introduction to Health Systems
Research Methodology...........
49
Module 1.
ii
Page
8.
Variables
50
9.
Study Type
53
10.
Data Collection Techniques
54
11.
Sampling
60
12.
Plan for Data Collection
62
13.
Plan for Data Analysis and Interpretation
63
14.
Ethical Considerations in Health Systems
Research ................................................
67
15.
Pre-testing the Methodology
68
16.
Utilization and Active Dissemination
of Results.......................................
70
17.
Work Plan
71
18.
Administration, Monitoring and Evaluation
72
19.
Budget
73
20.
Finalizing the Research Proposal
74
Summary Orientation to HSR Training Course,
to Inform Authorities and Participants.............
75
Guidelines on Budget for Health Systems
Research Training Course.......................
77
3.
Sample of Course Schedule
79
4.
Evaluation Questions HSR Training
Workshop......................................
84
Annex 1.
2.
iii
INTRODUCTION TO TRAINER’S NOTES
A.
COURSE PREPARATION
Selection of Course Management Team
A course coordinator will, with support from four or five facilitators, make up a course management
team.
Course facilitators may be selected according to the following criteria:
Experience in health sytems research
Experience with participatory teaching
Availability for two workshops of two weeks and for some field visits during the 4-5
months of research implementation in between these two workshops, to provide
supervision and support
Preferably experience in previous HSR workshops as a participant or facilitator
Ideally, the team should comprise a variety of disciplines, such as: medical sociol
ogy; health management/public health; and epidemiology
An equitable mix of male and female facilitators is recommended
Though the course coordinator is responsible for the overall smooth running of the course it is
highly recommended that he delegates administrative tasks to a course manager. The course man
ager will, for example, make administrative arrangements, supervise support staff, (typists, drivers),
ensure that participants and facilitators receive the necessary support to travel to and from the
course site, and that financial obligations and support tasks during and after the workshop are car
ried out promptly. He/she will be a member of the course management team.
WHO/AFRO/SHDS
Joint HSR Project
1
Tasks of the Course Management Team in Course Preparation
The course management team will be responsible for the following tasks:
1.
Selection of Participants
The optimum number of participants for this course is 20-25.
working groups.
Participants will be divided into four
Criteria for selection of the participants include:
Multi-disciplinary representation (health managers, medical officers, nurses and
health inspectors, tutors from training institutes, representatives from health-related
sectors and junior researchers)
Each of the four groups should be from a geographical area that is limited in size
to enable a close working relationship during the implementation phase of the proj
ect
A good mix of males and females
Ability to actively participate in all three phases of this course over a 7 month pe
riod
Longer term availability to utilize the skills learned and to potentially be Sble to pro
vide training to others
2.
Discussion with Course Management Team on Training Methodology and Training Pro
cedures
It is extremely important that the Course Management Team as a whole takes time to discuss the
introduction on training methodology that follows later, and the training modules with trainers notes.
Consensus will have to be reached on how to introduce the modules and how to act as facilitator
during group work and plenaries. The collective capabilities of the team will have to be assessed in
relation to the training requirements.
3.
Selection of Additional Local Resource Persons
Additional local expertise in disciplines such as epidemiology, statistics, or of a librarian or someone
who is presently involved in an interesting HSR project may be required.
2
Outside resource persons should generally not be asked to present a module, unless they are very
familiar with the course and its methodology. However, it is good to invite them to familiarize them
with the course; to make their expertise available for the group work; to introduce them to the par
ticipants as valuable resource persons (both during the course and afterwards) and finally, to enlist
their support for the implementation of the proposals being developed.
4.
Invitation of Authorities to Open or Close the Course
Usually, a high official from the Ministry of Health, and - if possible - a representative from another
agency supportiqg the course are invited to open the course.
As a rule, the official openings of courses take place on the first morning. It might be worthwhile,
however, to officially open the course in the late afternoon of the first workshop day. This will save
time. It will also enable the authorities to be informed and to comment on the research topics cho
sen, which may sensitize them on the value of HSR and generate their support for the implementa
tion of research results.
5.
Invitation of Donors
If you consider inviting donors to explain what research projects they presently support and to pro
vide details on research priorities and funding procedures, it is advisable to invite them all together
one evening in a panel.
Each may give an individual presentation of 10-15 minutes. Then a general discussion may follow,
allowing course participants to ask questions.
6.
Selection of Support Staff
Support staff should include 2 typists (12 working days each) and 1 driver/messenger (15 working
days). For the last three days of the workshop 4 fulltime typists would be desirable. Typists may
have to work overtime to finish questionnaires before the pretest and to finalize research proposals.
Site Preparation
Space required:
Plenary space for 30 persons plus two small meeting rooms
Office facilities for 2 typists and space for a photocopy or duplicating ma
chine
Materials required:
Access to a vehicle for 12 working days; for the pre-test extra transport
may be needed
3
Access to a photocopying or duplicating machine plus supplies and paper
(2,000 sheets)
2 typewriters, overhead projector, calculators (1 per working group)
Flipcharts (6), paper pads (30)
Pens, pencils, marking pens for overhead and flipcharts, overhead transpar
ency sheets, hole puncher, stapler, tape, drawing pins, paper clips
Course library
The course management should make available to participants:
Country or area reports
Text books on:
Health management
Statistics
Health systems research
Social science research methods
Questionnaire design
As soon as the groups have selected their topics, considerable effort should go into
finding more specific and relevant literature and information on these topics.
8.
Pre-Course Contacts with and among Participants
It is strongly recommended that the questionnaires on experience with HSR (See pt 3,14, Question
naire) should be sent out to course participants at least a month before the beginning of the
course, so that this information can be included in the planning.
Also assessing the knowledge of the participants of HSR is most useful when done before the on
set of the course (see p15 etc). This can either take place on the day of arrival, or (if the partici
pants don’t live too scattered), during a special meeting some days before the workshop starts.
This information allows the Course Management Team to get a better idea of the strengths and
weaknesses of the participants. It may be possible to compensate for differences in level of the
participating working groups by allocating a “strong” participant who is not tied to one particular
geographical location (national level) to a weaker working group. Having working groups of ap
proximately equal strength will considerably facilitate the learning process of the group as a whole.
As geographical concentration of participants is one of the selection criteria, it may be pos
sible for facilitators to visit each working group to explain the purpose of the course and
assess the participants’ HSR abilities at the same time.
4
Such visits are highly recommended, as they have the additional advantage that working
groups can already brainstorm about possible research topics among themselves and bring
relevant local reports with them to the course for the literature review that forms part of
their proposal.
Note:
It is essential that health authorities and potential participants to the course understand
beforehand that their input will not only be required during the first course to develop a
research proposal, but also over a period of 5-6 months to actually carry it out, and for
a subsequent two-week workshop to analyze the data, make recommendations and
write the report.
9.
Essential Paperwork
(Required throughout the course preparation period)
Prepare a summary orientation to the course, including course objectives, content
and procedures (for example, see Annex 1) to inform authorities and participants.
Obtain official approval for the course from
Ministry of Health or other appropriate authority;
Additional sponsoring agency (if required).
Book a site for the training course.
Prepare a course budget (for guidelines, see Annex 2).
Develop Course Schedule (for example, see Annex 3).
Make administrative arrangements with the summary of the orientation to the
course, a course schedule and a questionnaire (requesting a reply within two
weeks) to all participants. The letter of invitation should include administrative de
tails.
Arrangements have to be made to speed up approval of the research proposals by
specifying the steps that have to be taken by all parties concerned (Ministry of
Health, Course Management Team, Donors) beforehand.
5
B. COURSE MANAGEMENT
Chairpersonship of Plenary Sessions
It may be useful to rotate the chairmanship according to the subject being discussed. For instance,
the person presenting the introduction and guiding the following discussion could be the chairper
son of that session. The course coordinator will, however, have the overall chairpersonship, conduct
opening and closing sessions and make general announcements (reading materials for the next
day; work in weekends).
Facilitators’ Meetings
To monitor progress of the course and give an opportunity to the facilitators to discuss possible
problems, it is desirable to have a daily meeting of facilitators. This meeting may be best held in the
evening and last between half an hour and an hour. The course coordinator is responsible for con
vening this meeting. It is probably helpful to have a secretary for each meeting and make a record
of at least the action points.
Allocating Facilitators to Working groups
When the participants have selected their research topic, a final decision will have to be made as to
which facilitator can, according to his interest and expertise, best be in charge of a particular group.
Facilitators will in principle stay with the same group throughout the course, in order to ensure con
tinuity and sufficient quality of the end product.
However, each facilitator may have an overall responsibility for certain technical aspects in which
he/she is specialized, and assist other groups as well. Also local resource persons may assist on
an ad hoc basis.
Though a course can be run by four facilitators (including the course coordinator) facilitation will be
more relaxed with five, one of whom (most likely the course coordinator) is ambulant.
Evaluations
It will be useful to have several evaluation sessions during the workshop, obtaining participants’
comments about, for example, course content, mode of conducting introductions, plenary discus
sions and group work.
6
This can be done by a group discussion in plenary or with a simple self-administered questionnaire,
or by combining both methods. At the end of the course, an overall evaluation should take place
in order to obtain suggestions for improvements (for an example of an evaluation form, see Annex
4).
Assessment of Participants’ knowledge of HSR before and after the
Course
It is a useful idea to have the course participants evaluate a proposal for a research project with
several omissions before and after the course, in order to enable the facilitators to judge a possible
increase in knowledge and understanding of HSR among the participants.
However, using the same test before and after the course has been shown to lead to certain diffi
culties. Especially the more experienced participants appeared bored when they had to answer the
same questions on the same proposal, and sometimes scored equal or even lower at the post-test
than during the initial test. Also it was difficult to get participants full attention for the post-test, as
they were very absorbed in preparing the final draft of their research proposals during the last 1-1 /2
days of the workshop.
The first problem may be solved by providing two research proposals on different topics but with
identical weaknesses and similar questions, (see Assessing Participants Knowledge on HSR, p.15).
If the same Course Management Team conducts different courses, it is advisable to alternate the
sequence of the tests to verify that the tests are equally difficult.
The second problem (to get participants full attention for the test at the end of the course) seems
more difficult to solve. Perhaps taking the test immediately after the presentation of module 20 will
somewhat alleviate this problem.
Pre-test of Research Methodology
A pre-test of data collection techniques and other methodological or organizational aspects of the
newly developed research proposals will be carried out in the course of the second week. The
Course Management Team should make the arrangements for the pre-test during the first week,
when the four research problems have been selected, and some thought has been given to the
methodology. Arrangements may include obtaining consent from the relevant authorities to do the
field work and organizing the extra transport and secretarial support needed by each group.
In order to involve the participants as much as possible in the preparations and minimize stress on
all parties concerned, it is highly recommendable to have one full course day between the presenta
tion of the module on the pre-test and its actual implementation (see course schedule Annex 2).
7
Finalizing the Research Proposals
As the intention is to start implementing the research proposals that have been developed immedi
ately after the workshop, the proposals have to be in such a shape by the end of the workshop
that they can be presented for consent and carried out.
This implies that one full day has to be reserved for revising the various parts of the proposal and
putting them together. Most likely this will take place the Friday afternoon and Saturday of the sec
ond week.
Participants, typists (preferably four at this time) and facilitators will have to be aware that they will
probably have to do overtime during these days.
Workshop Report
The official report of the workshop should be as brief as possible. After a one page introduction
(when, where, why, organizers, sponsors, type of participants of the course), a summary report of
2-4 pages could follow, describing the training process, starting with the topics chosen and ending
with evaluation results. A list of participants and course facilitators for research by the working
groups could be annexed to the report, as well as (summaries of) opening speeches.
The main body of the report should consist of the final draft of the four research proposals. It is
highly recommended that the course facilitators as a group screen the proposals once again imme
diately after the workshop, because some items may have dropped out or entered the proposal in a
way that would need some clarification. The final brushing up on minor points can be done in the
month following the workshop. Usually this month is taken up by procedures to obtain consent for
implementing the proposals so that there is some spare time.
Preparation for Research Implementation
Potential sponsorship for the implementation of the research proposals developed during the work
shop must have been agreed upon before the workshop, as well as procedures on how to get final
consent and ensure rapid availability of funds.
The focal point for HSR in the Ministry of Health will (together with the course coordinator, if this is
a different person) usually be responsible for getting the proposals to the donor immediately after
the workshop, with an official request for funding and an account number on which the money can
be transmitted. The donor (if Joint HSR Project or WHO) will prepare a Technical Agreement (TA)
which has to be signed by the principal investigators of the projects and the principal administrator
within the Ministry, and be returned straight away for final signing by the donor. It may take 1-1/2 2 months before the money is actually on the spot. Therefore procedures may have to be devel
oped within the MOH to advance some funds for those working groups that would like to start the
actual field work earlier.
8
The national channels to endorse the research proposals, (National Research Council, for example)
will also have to be mobilized before as well as after the workshop, to speed up procedures.
Finally, the research teams themselves will have to carry out preparatory activities, such as final
testing of the data collection tools, and looking for assistance in data collection and data process
ing, if required. They have stated these activities in detail in the work plan that they developed as
part of the research proposal.
Meeting the Working Groups’ needs for Technical Support during
Research Implementation
The local facilitators that have assisted in the development of the proposals will also assist the
groups in the implementation of the proposals. However, sometimes additional support may be re
quired, e.g. the assistance of an experienced sociologist or statistician with respect to data collec
tion and data processing.
The working groups have to state this in their proposal (section budget). The Course Management
Team will have to assist the groups, if necessary, in finding the right persons.
All groups will need assistance when they start sorting and processing their research data.
C.
TRAINING METHODOLOGY
Sessions in this training course on health systems research contain the following components:
Introduction and discussion
Group work
Exercise
Plenary
1.
Introduction and Discussion
An introduction and discussion is meant to briefly explain new concepts and their application. Invit
ing responses and suggestions from participants and listing these on a flipchart or using them as a
starting point for discussion is an essential element of the training method.
9
This increases the interest of participants and may bring up valuable points of view that would be
missed in classical classroom teaching. Do not allow the discussion to be dominated by a few
participants.
Depending on the level of the participants the facilitator can delete or add details in the introduction
Note:
The text of the modules is therefore not meant to be followed word by word.
introduction should not last longer than 30, at most 45 minutes.
The
Participants should have ample time to become familiar with the newly introduced concepts when
they apply them to their own research proposal during the subsequent group work session. In
case there are questions, they can consult their facilitator and the text of the modules.
It is important to prepare the presentations very well preferably on transparencies, as this will
help you to limit yourself to the major concepts that have to be explained. A thorough preparation
is particularly important to focus the discussion on the main points. Use the discussion also to
check that participants have grasped the topic well.
2.
Group Work
The group work aims at developing four research proposals (one per working group) that should be
ready for implementation by the end of this two-week workshop. It should therefore always be
kept in mind that the proposal being developed needs to be feasible and of good quality.
To increase the efficiency of the groups, it is desirable that for each group work session a chair
person and rapporteur are appointed. The chairman is not only responsible for leading the discus
sion, but also for dividing the work between group members. It is recommended that after discus
sion within the working group, it splits up in groups of two or three persons to elaborate separate
parts of the work to be done. These can then be discussed and amended before presentation in
the plenary.
Each facilitator will be responsible for one group throughout the course to ensure continuity. Facili
tators should only change groups if they have major problems in assisting their own group. Of
course, other facilitators and resource persons can be consulted at all times on technical issues.
The amount of time the facilitator spends with his/her group will depend on the needs and de
mands of that group. In the beginning of the course the needs may be greater than towards the
end. In principle facilitation is a full-time activity. Even if a facilitator is not participating in the
group work he should be available all the time for consultation.
10
His/her role in discussion is in the first place to stimulate the group to find its own solutions. How
ever, if the group is clearly going in the wrong direction he should provide more direct guidance.
At the beginning he may have to prevent groups wasting time on less relevant issues, or prevent
relevant issues brought up by group members from being dropped because not everyone sees their
importance.
Groups should be encouraged to write summaries of each session on flipcharts or transparencies
for presentation in plenary and for future reference.
3.
Exercise
There are two types of exercise. In some exercises groups practice the use of new concepts in
case studies prepared in advance. For these it is probably a good idea to make groups of a differ
ent composition to those in the group work, so that all participants get to know each other well. In
the second type of exercise a group will examine a component of the proposal that is being devel
oped by another group and provide constructive criticism. Groups should be encouraged to put a
summary of their comments on a flipchart or transparency for presentation is plenary and for refer
ence by the group developing the proposal.
4.
Plenary
Presentations of the results of group work or exercises in plenary require special skills. Before the
first plenary (in which the research topics considered for the development of research proposals are
presented) visibility and audibility of the presentations should be discussed with the participants.
The working groups can either use flipcharts or transparencies for presentation. Flipcharts have the
advantage that they can be easily used in the working groups for referral and further elaboration.
However if the plenary exceeds 25 persons, it may become difficult for all to read them. The use of
transparencies and an overhead projector may in that case be indicated.
It should be stressed that there are limits to what one can put on a transparency as well as a
flipchart. Have two examples, one of a readable and one of an unreadable transparency, and let
the participants give suggestions as how much information a transparency can optimally contain.
Stress also that one should never turn one’s back to the audience when presenting.
In general, the presentation of one working group should not exceed 15 minutes, discussion in
eluded. Sometimes even less time is required.
11
Trainers Notes
INTRODUCTION TO THE COURSE
1.
Questionnaire
The questionnaire that follows is meant to give an impression about the experience with and
interest in HSR of the course participants. This information will be very useful when consid
ering the final composition of the working groups, each of which will develop a research
proposal. Minor adjustments may be necessary to ensure that all working groups are of
approximately equal strength.
In order to receive this information in good time, the questionnaire has to be sent to the
participants some 6 weeks before the course starts, together with more information about
the course (content, location, administrative details).
The completed questionnaires should reach the Course Management Team at least two
weeks before the course starts.
Action required
Photocopy the questionnaire on pp 13-14 in 30 fold or have it stencilled, and mail
it 6 weeks before the course starts.
WHO/AFRO/SHDS
Joint HSR Project
12
Health Systems Research Course
PARTICIPANTS’ BACKGROUND, EXPERIENCE AND INTERESTS
Please complete and send back within 14 days.
Last name:
Dr./Mr./Ms.
(Circle correct
title)
First name:
Commonly used first name:
(if different)
Complete mailing address :
Telephone number:
(work)
Qualification/T raining:
Present occupation or profession:
Present position and/or title:
Employer:
Present professional responsibilities:
Past work experience:
WHO/AFRO/SHDS
Joint HSR Project
13
(home)
Have you already had experience in:
yes
yes
yes
yes
designing a research project
executing a research project
monitoring a research project
evaluating a research project
no
no
no
no
If yes, please describe in more detail the types of experience you’ve had related to research:
What are in your opinion the problems related to health services and primary health care for which
applied research might be useful in finding appropriate solutions:
Are you interested in applied research?
If yes, why, and in what subject area would you like to do research?
What are the constraints in undertaking applied research where you work?
What would you like to gain from this course?
14
2.
Mutual Introduction of Course Participants and Facilitators
Participants and facilitators should get to know each other as soon as possible. One and a
half hours should be reserved for a mutual introduction on the day of arrival (Day 0). This
could be done by interviewing each other in pairs (15 minutes) on:
Name
Professional education
Professional experience
Experience with HSR
Topics on which they would like to develop a research proposal
The main points can be written on flipcharts (one per pair) and stuck to the walls of the
plenary room. Each person then briefly introduces his/her partner. (1 hour in total)
If the presentation cannot take place on Day 0, it has to be included in Session 1, Orienta
tion to the Course.
Name tags may be handed out at the beginning of this session.
3.
Assessing Course Participants’ Knowledge on HSR Before
and After the Course
Teaching Methods and Timing
5 minutes
1 hour
Introduction
Accomplishing the test
65 minutes
TOTAL TIME
One possible (and non-threatening) way of assessing course participants’ knowledge on
HSR is to let them evaluate research proposals which have obvious omissions. Two such
defective research proposals are included in this section, one originally presented in the
SHDS modules, and the other developed along similar lines. You may use either to start
with, or use the same twice, or develop other case-studies as testing materials.
You should administer the first test (pre-test) before any detailed information about the
course has been given (preferably on Day 0). The second assessment of participants
knowledge on HSR (post-test) may best take place just before the research proposals are
finalized (Day 12). When comparing the scores of both tests it will be possible to measure
any change in knowledge of individual participants and of the group as a whole.
15
Guidelines on administering the tests
When distributing the test to the participants they should be advised to first thoroughly read
the questions and the research proposal. Only then should they answer the questions, for
which they may take an hour at the most.
Both tests should be administered in the came manners and for an identical period of time
(1 hour) so that results will be comparable.
Do not mention, when the first test is being given, that an identical test will be administered
at the end of the course.
When grading both tests and comparing results, some systematic method should be used.
We have developed some questions for the participants as well as a proposal for grading
the answers. This schedule follows immediately after the test. The grading system can be
adjusted as long as the same system is used in both tests.
The following points could be helpful in grading the tests:
Two facilitators should be rating the answers of each test independently. Compare and dis
cuss differences in rating, and develop a uniform system. The same two facilitators should
rate both tests.
Sometimes participants mention relevant points, but which are not under the appropriate
question. These points should still be counted.
Scores of all individual participants can then be added. The total divided by the number of
participants gives the average score.
The scores should be calculated immediately after conducting the tests. The results of the
first test may assist the Course Management Team in adjusting the composition of the
working groups. The participants should be informed that the content and results of the
test will be discussed at the end of the course (to prevent discussion that could influence
the post-test).
The post-test should be graded before the participants leave, so that the differences in re
sults can be communicated to the participants. Now also the answers to the test(s) can be
discussed.
16
-
Action required
Photocopy the first test that you have decided to use in 30 fold or have it
stencilled before the course starts (Day O).
Provide enough blank paper for the participants to answer the questions.
Photocopy the Score Sheet for trainers in 4 or 8 fold
Photocopy or stencil the post-test that you have decided to use in 30 fold
before Day 12.
t
Photocopy or stencil the Answer Sheet(s) to the test(s) in 30 fold and dis
cuss these together with the results of the tests, on Day 13.
17
TEST A
Kibara Case Study
Introduction
Imagine that you are the chairman of a department in a health training institution or that you are the
Director of the Research Unit in the Ministry of Health.
You have received the research proposal outlined on the following pages and you have to decide
whether it should be supported with the research funds you have at your disposal.
As you read the proposal, think about the following questions:
Statement of the Problem
1.
Are the reasons for carrying out the study clearly stated in the background information?
Explain your opinion.
Literature Review
2.
Is the literature review satisfactory? If not, what else could have been considered?
Methodology
3.
Are the data to be collected under the stated methodology, enough to provide answers to
the problem? If yes: why? If not, what major point is lacking?
4.
Are you satisfied with the methods proposed for data collection? Give your reasons.
5.
Is the process for selecting people to be interviewed stated clearly enough? Please ex
plain your opinion.
Staffing and Timing
6.
What is your opinion of the personnel and the time allocated for the study?
Other Points
7.
Can you suggest any other points which should be included or changed to make this pro
posal acceptable for funding?
You now have 60 minutes to give your answers to each of the above questions on the study
provided.
WHO/AFRO/SHDS
Joint HSR Project
18
TEST
Summary of a Research Proposal
1.
Title
A study of the Training and Utilization of Health Personnel in Kibara.
2.
The Problem
2.1
Background situation
Kibara is a small country of five million inhabitants situated in the developing world.
The health care policy of the government aims at providing all forms of health care
(preventive, curative and educative) for the country’s entire population on a total
coverage basis. To achieve these lofty aims, efforts will be made:
to train health personnel adequately and in suffient numbers to cover all
sections of the service at the centre and periphery.
to establish an adequate number of basic health services to serve the pre
dominantly rural population and support the progressive development of pri
mary health care activities in the villages.
There is a Medical School attached to the only University in the country which only
trains doctors. In addition, there are four other polyvalent training institutions in the
country which train State Registered Nurses, State Certified Midwives, Medical As
sistants, as well as other middle and lower level health personnel. There is a good
working relationship between the Medical School and the Ministry of Health, with
the latter using the former in a consultant capacity.
3.
Review of the Literature
A review of the literature shows that while some studies have been carried out in North
America on the use of nurse practitioners, no study of this type has been carried out in the
region where Kibara is located.
WHO/AFRO/SHDS
Joint HSR Project
19
4.
Methodology
An assessment of the health needs of the population will be made, using reports and other
data already available.
Descriptions of student terminal competence for the three categories of health personnel will
continue to be collected or prepared at the institutions where the students are trained.
A suitable questionnaire will be elaborated, and dispatched to:
all the doctors in the country (a total of 120 doctors); and
all the nurses, midwives and medical assistants working in a random sample of 25%
of the administrative districts of the country.
A suitable sample procedure will be used. The questionnaire will explore the health work
ers’ job descriptions, the tasks the workers are actually performing, the conditions in which
they work and the amount and type of supervision they receive.
Completed questionnnaires will either be retrieved personally or the respondents will be
asked to mail them. The data processing will be carried out manually. Analysis will be
made of the extent to which the health personnel in the three categories are using the train
ing they received, and recommendations will be prepared concerning ways in which train
ing, placement and job description, and/or supervision could be improved so that workers
can better serve the needs of the population.
5.
Project Personnel and Consultants
The head of the Department of Community Health in the medical school will be director of
the research project, devoting 20% of his time to the study. He will be assisted in the exe
cution of the research by a small team made up of:
The Lecturer in Biostatistics in the Medical School (10% time)
The Chief of the Division of Training in the Ministry of Health (10% time)
The Principal of the Health Personnel Trainig School in the capital (10% time)
A staff member principally responsible for training medical assistants at the Health
Personnel Training School. (15% time)
20
In addition, an outside consultant will be engaged for 30 days during the final phase of the
project to assist the project director and other members of the research team in data analy
sis and preparation of the research report.
6.
Plan for Project Administration and Monitoring
The director of the research project will be responsible for managing the research grant,
which will be operated through the University Research Account. He will be responsible for
supervision of the work of the rest of the research team and will hire and supervise the out
side consultant.
Both the University and donor agency will monitor the project to see that it is on schedule
and achieving its goals, through study of the questionnaires, interim report, financial records
and final report.
7.
Financing the Project
The project will be jointly financed by the government of Kibara through the University of
Kibara and a donor agency. Those aspects of the budget to be financed by either party
are shown in the detailed budgetary provisions.
8.
Budget
A.
YEAR 1
YEAR 2
TOTAL
Project Director
(20% of time)
$ 3000
$ 3000
$6000
Biostatistician
(10% of time)
$ 1500
$ 1500
$ 3000
Chief of Division,
Ministry of Health
(10% of time)
$ 2000
$ 2000
$ 4000
Contribution of the host
Government
Personnel
21
YEAR 1
YEAR 2
TOTAL
Principal of the Health
Personnel Training School
(10% of time)
$ 1500
$ 1500
$ 3000
Staff Member of Health
Personnel Training School
(15% of time)
$ 1500
$ 1500
$ 3000
Subtotal
$ 9500
$ 9500
$19000
Office space, furniture,
telephone, electricity
$ 2500
$ 2500
$ 5000
Subtotal
$ 2500
$ 2500
$ 5000
TOTAL FOR A
$12000
$12000
$24000
$ 3000
$ 3000
$ 6000
$-
$ 8000
$ 8000
$ 3000
$11000
$14000
Duplicating Machine
$ 1000
$ -
$ WOO
Resource Material (books, etc.)
$ 500
$ 250
$ 750
Subtotal
$ 1500
$ 250
$ 1750
Data analysis
$ 500
$ 1500
$ 2000
Subtotal
$ 500
$ 1500
$ 2000
TOTAL FOR B
$ 5000
$12700
$17750
GRAND TOTAL
$17000
$24750
$41750
Other Direct Costs
B.
Contribution of the Donor Agency
Personnel
Project Secretary (100%)
Consultant (travel fees and per diem,
(30 days)
Subtotal
Equipment and Supplies
Data Analysis and Reporting
22
Trainers Notes
TESTA
Kibara Case Study
(Answer Sheet)
QUESTION 1: Are the reasons for carrying out the study clearly stated in the background in
formation?
The reasons for carrying out the study are not clear as the problem has been poorly defined.
1.
Indications should be given as to what “sufficient numbers” are for various categories of
staff, to which extent these numbers have been reached, and a question for additional infor
mation (for reasons why these targets are not met, or even the question whether these tar
gets are appropriate) should have been developed.
2.
Indications should be given as to what they consider by “adequate” training, and, for ex
ample, why they fear that the training is less adequate than they would wish.
3.
No objectives have been developed, thus we still do not have a clear idea what the re
searchers hope to achieve with the study.
(Any of these three points raised can count for one point 3 in total).
QUESTION 2: Is the literature review satisfactory?
The literature review is insufficient in various ways.
1.
It is unlikely that there would not have been any study in this field carried cut in SubSaharan Africa. It appears that the proposal is based on an incomplete review.
2.
Literature should never be just stated, but rather analyzed carefully, and if appropriate, used
to justify the study.
If the American study would have been the only one of this type, the conclusion should
have been that such a study in Kibara would be timely.
3.
No references have been included of the studies mentioned.
WHO/AFRO/SHDS
Joint HSR Project
23
4.
An additional point, (made here or under methodology), could be that key informants can
provide information on the problem even if it has not been written down, and that local
MOH reports and evaluations related to the problem should be consulted. These reports
exist (see methodology).
QUESTION 3:
Are the data to be collected under the stated methodology enough to provide
answers to the problem?
1.
' I be how the content of the
A major question with respect to adequacy of the training would
training relates to the job description and actual tasks carried out by the four types of
health personnel mentioned.
There is no mention of an analysis of the training
programme’s content.
2.
Nowhere is it mentioned that the population has been involved in the needs assessment,
The reports on which the assessment of the health needs of the population is based are
neither discussed not included.
QUESTION 4: Are you satisfied with the method proposed for data collection? Give your rea
sons.
Only one method (or technique) for data collection is mentioned; (questionnaires for health staff)
One could also consider:
1.
Interviews with key figures (staff, health managers, policy makers, tutors of training
schools) about how they think the training could be improved, in relation to tasks now per
formed and in relation to needs of the population (see question 2, point 4). These inter
views should preferably be done before the drafting of the questionnaire.
2.
Observations of how staff actually perform the duties in which they were trained (to judge
quality of training).
3.
Analysis of curriculum (if not already mentioned under question 3).
(1 point for each valuable addition made).
QUESTION 5: Is the process for selecting people to be interviewed stated clearly enough?
1.
The sampling procedure has not been defined.
2.
The sample size (number of nurses, midwives and medical assistants to be interviewed) is
unknown. In this way, planning of manpower, time and money required for the study be
comes impossible. (Each of these issues raised can count for a point.)
24
QUESTION 6: What is your opinion on the personnel and time allocated for the study?
Personnel and time were poorly allocated:
1.
The professional personnel mentioned have only a small fraction of their time allocated.
Without research assistants it seems unlikely that the investigators will accomplish the study
in a reasonable amount of time.
2.
All researchers are health trainers. Health staff should also be involved in the study.
3.
The project period has not been stated in the text.
clear that the project will last for two years.
4.
No work plan has been made!
5.
Consultants time seems poorly planned. If a consultant is to be used, that person should
come in at the beginning, not at the end (when possible shortcomings which become ap
parent during the data analysis cannot be corrected anymore).
(Any major point made can count as a point).
Only from the budget does it become
QUESTION 7: Can you suggest any other points which should be included or changed to
make this proposal acceptable for funding?
There are a number of other shortcomings, for example
1.
There is no mention made of how the results will be disseminated or utilized.
2.
The budget has some major omissions (costs for transport,
contingencies).
3.
The principal investigator cannot at the same time be the principal administrator.
4.
It is unclear who is requesting support for this research project.
5.
A decision on how the questionnaires will be retrieved should be made beforehand, and
should be communicated to the respondents. If they know that the questionnaires will be
collected, the response rate may be higher.
(Any major issue can count for a point.)
field costs, reporting costs,
If additional major points are brought forward, they can count for a point provided
the facilitators who grade the answers agree on these points
25
Trainers Notes
PRETEST AND POST-TEST
(Score sheet for trainers)
Total
Question Number
Name of
Participant
1
2
4
3
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
26
5
6
7
Points
TEST B
Compliance to Diet Study
Introduction
Imagine that you are the Director of the Research Unit in the Ministry of Health.
You have received the research proposal outlined on the following pages, and you have to decide
whether it should be supported with the research funds you have at your disposal.
As you read the proposal, think about the following questions:
Statement of the problem
1.
Are the reasons for carrying out the study clearly stated in the background information?
Explain your opinion.
Literature review
2.
Is the literature review satisfactory? If open to improvement, what would you suggest?
Methodology
3.
Are the data to be collected under the stated methodology enough to provide answers to
the problem (non-compliance to diet)?
If yes: why? If not, what major point is lacking?
4.
Are you satisfied with the methods proposed for data collection? Give your reasons.
5.
Is the process for selecting people to be interviewed stated clearly enough? Please ex
plain your opinion.
Staffing and time
6.
What is your opinion of the personnel and time allocated for the study?
Other points
7.
Can you suggest any other points which should be included or changed to make this pro
posal acceptable for funding?
You now have 60 minutes to give your answers to each of the above questions on the study pro
vided.
WHO/AFRO/SHDS
Joint HSR Project
27
TEST B
Summary of a Research Proposal
1.
Title
A Study of Factors Influencing Compliance to Diet by Diabetic Patients Referred to
Dietetic Clinics in Matete.
2.
The Problem
Matete is a small country of one and a half million inhabitants situated in the tropics. The
country is doing well economically, with an annual growth of the GNP by 4%. Its main for
eign currency generating activities are the production of sugar cane and cotton, and light
industry.
The IMR has over the past 30 years gone down from 80 to 25 per 1000 live births. The
health care policy of the country aims at providing all forms of health care (preventive, cura
tive and educative) for the country’s entire population on a total coverage basis.
The Ministry of Health has recently started dietetic clinics in three district hospitals, to which
dispensaries and health centres as well as out-patient clinics of the hospitals and private
physicians are supposed to refer their patients. One of the activities of these clinics is the
provision of advice on diet. However, the attendance of patients to these clinics is not yet
satisfactory.
3.
Review of the Literature
A review of the literature shows that dietetic counselling can to a large degree control the
maturity-onset type of diabetes that is associated with obesity.
Few of these studies have, however, been carried out in developing countries, and in
Matete no such study has been conducted.
WHO/AFRO/SHDS
Joint HSR Project
28
4.
Methodology
A large national survey was conducted one year ago, assessing the prevalence and severity
of non-communicable diseases in the country. The preliminary results of this study indicate
high incidence of diabetes, hypertension and coronary heart diseases, underlining the im
portance of diet counselling. Final results of this survey will be made use of in the present
study.
To measure how health staff and patients evaluate the newly established dietetic clinics,
suitable questionnaires will be sent out to
all doctors in the three districts, and
a sample of 25% of all medical assistants and nurses involved in out-patient care at
the three service levels in the districts (dispensaries, health centres and district hos
pital).
Moreover, 10% of all diabetic patients registered in the three dietetic clinics will be
interviewed.
Suitable sampling procedures will be used.
The questionnaires will explore to what extent health staff are aware of the newly estab
lished dietetic services, to what extent they refer patients, and how patients evaluate the
dietetic services and the advice provided. Analysis will be done by hand. Recommenda
tions will concentrate on how the newly established diet counselling services can be made
better known to health staff and the public, and how the content of the counselling can be
best adapted to the patient’s needs.
5.
Project Personnel and Consultants
The Head of the Department of Community Health in the medical school will be the princi
pal investigator, devoting 20% of his time to the study. He will be assisted in the execution
of the research by a small team made up of:
the Head of the Nutrition Unit within the Ministry of Health (10% time)
the dietician attached to the MOH Nutritional Unit (20% time)
the lecturer medical sociology of the Department of Community Health (10% time).
An outside consultant, a statistician, will be engaged for 30 days during the final phase of
the project to assist in data analysis.
29
6.
Plan for Project Administration and Monitoring
The Principal Investigator will be responsible for managing the research grant, which will be
operated through the University Research Account. He will be responsible for supervision
of the work of the rest of the research team and will hire and supervise the outside consult
ant.
Both the University and the donor agency will monitor the project to see that it is on sched
ule and achieving its goals, through study of the questionnaires, interim report, financial rec
ords and final report.
7.
Financing the Project
The project will be jointly financed by the government of Matete through the University of
Matete, and a donor agency. Those aspects of the budget to be financed by either party
are shown in the detailed budgetary provisions.
8.
Budget
A.
Contribution of the
host government
Year 1
Year 2
TOTAL
Project Director
(20% of time)
$ 3000
$ 3000
$ 6000
Head Nutrition Unit MOH
(10% of time)
$ 2000
$ 2000
$ 4000
Dietitian Nutrition Unit MOH
(20% of time)
$ 2500
$ 2500
$ 5000
Lecturer Medical Sociology
(10% of time)
$ 1250
$ 1250
$ 2500
$ 8750
$ 8750
$17500
Personnel
Subtotal
30
Other Direct Costs
B.
Office space, furniture,
telephone, electricity
$ 2500
$ 2500
$ 5000
TOTAL FOR A
$11250
$11250
$22500
Year 1
Year 2
TOTAL
$ 3000
$ 3000
$ 6000
$ 8000
$ 8000
$11000
$14000
Contribution of the
Donor Agency
Personnel
Project Secretary
(100%)
Consultant (travel fees and
per diem)
(30 days)
$ 3000
Subtotal
Equipment and Supplies
$ 1000
Duplicating Machine
$ 1000$
Resource Material
(books etc.)
$ 500
$ 250
$ 750
$ 1500
$ 250
$ 1750
$ 500
$ 1500
$ 2000
Subtotal
$ 500
$ 1500
$ 2000
TOTAL FOR B
$ 5000
$12750
$17750
GRAND TOTAL
$ 16250
$ 124000
$ 140250
Subtotal
Data Analysis and Reporting
Data analysis
31
Trainers Notes
PRETEST AND POST-TEST
(Score sheet for trainers)
Total
Question Number
Name of
Participant
1
2
4
3
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
32
5
6
7
Points
Trainer’s Notes
TEST B
Compliance to Diet Study
(Answer Sheet)
QUESTION 1: Are the reasons for carrying out the study clearly stated in the background in
formation?
The reasons for carrying out the study are not clear, as the problem has been poorly defined.
1.
In no way has it been explained why the authors of the research proposal suspect that noncompliance to diet is a problem among diabetic patients, and what the possible causes
could be.
2.
Obviously, irregular or non-attendance to the dietetic clinic could be one of the reasons, but
also here neither magnitude nor possible causes of non-attendance have been indicated.
3.
No objectives have been developed, therefore we have little idea what the researchers are
going to investigate and what they hope to achieve with the study results.
(Any of these points raised can be counted for one point, 3 in total).
QUESTION 2: Is the literature review satisfactory?
The literature review is unsatisfactory in several respects:
1.
No references have been included.
2.
No efforts seem to have been made to find similar studies carried out in developing coun
tries. Even if these would have been rare with respect to compliance to diet or regularity inattendance at dietetic clinics, studies in related fields are most likely available (leprosy, TB
services).
3.
More information should have been provided about the content of the studies referred to,
especially revealing reasons for compliance or non-compliance of patients to services and
to diet. Such a literature analysis could, among others, provide arguments to justify the
study.
WHO/AFRO/SHDS
Joint HSR Project
33
4.
Relevant local Ministry of Health reports and studies, for example on estimates of the size
of the problem of diabetes (which exist, see methodology), should have been included and
analyzed here. Interviews with key figures (the dietitians, for example) could have been
mentioned here or under methodology.
QUESTION 3: Are the data collected under the stated methodology sufficient to provide an
swers to the problem (non-compliance to diet)?
Not mentioned is how patients’ compliance to diet will be evaluated. This is the most important
omission of the study design: The title does not cover the content. (1 point)
QUESTION 4: Are you satisfied with the methods proposed for data collection? Give reasons
for your answer.
Only one method (or technique) for data collection is mentioned (questionnaire).
One could also consider:
1.
Interviews with key figures such as the dietitians (if not already mentioned under question
3)
2.
The attendance records should be analyzed in order to evaluate the size of the problem of
non-attendance of diabetic patients to dietetic clinics, and to distinguish different attendance
categories of patients.
3.
The medical records could give information on the seriousness of the diseases, which is
most likely influencing patients’ compliance to diet. (1 point for each valuable addition
made)
QUESTION 5: Is the process for selecting people to be interviewed stated clearly enough?
1.
No, the sampling procedure has not been defined.
2.
Also the sample size has not been calculated. Therefore it seems impossible to plan the
research (time and other resources needed) (Each of these issues can count for a point).
34
QUESTION 6: What is your opinion on the personnel and time allocated for the study?
Personnel and time were poorly allocated.
1.
The professional personnel mentioned have only a small fraction of their time allocated.
Without research assistants it seems unlikely that the investigators will accomplish the study
in a reasonable period of time.
2.
The staff providing dietetic counselling should also be involved.
3.
The project period has not been stated in the text. Only the budget provides an idea about
the duration of the study.
4.
No work plan has been made.
5.
If the support of an external consultant will be needed, that person should be involved in
the beginning, not at the end (when possible shortcomings which become apparent during
the data analysis cannot be corrected anymore). (Any major point made can count as a
point).
QUESTION 7: Can you suggest any other point which should be included or changed to
make this proposal acceptable for funding?
There are a number of other shortcomings, for example:
1.
There is no mentioning of how the results will be disseminated or utilized.
2.
The budget has some major omissions (costs for transport, other field costs, reporting, con
tingencies).
3.
The principal investigator cannot at the same time be the principal administrator.
4.
It is unclear who is requesting support for this research project.
(Any major issue can count for a point).
Note:
If additional major issues are brought forward they can count for a point provided
that the two facilitators who rate the answers agree on these points.
35
TRAINER’S NOTES
TO THE
MODULES
36
Trainers Notes
Module 1 - ORIENTATION TO THE COURSE
Training Methods and Timing:
3/4 hour
1/4 hour
1 hour
Description of the course
Administrative remarks
Introductions among course participants and listing of health re
search interests.
(Ideally, this mutual introduction should take
place the afternoon or evening before the course begins)
2 hours
TOTAL TIME
Materials:
Name tags for participants and trainers
Course training materials for participants
Training Guidelines
1.
Summarize briefly the history of the course including why and how it was developed, for
whom, and where it has been held.
2.
If you were unable to give the pre-test the evening or afternoon before the course began,
allow 1-1/4 additional hours at this point (see pre-test trainers notes).
3.
Present the major objectives of the course, and stress its practical orientation.
4.
Distribute the course training document to the participants. Describe how the course will
be structured and how the training document will be used. Outline that the annexes pres
ent more details on research methodology for those who are interested.
5.
Emphasize the uniqueness of each participant’s background and experiences, pointing out
how important it will be for everyone to contribute as much as they feel comfortable.
WHO/AFRO/SHDS
Joint HSR Project
37
6.
If you were unable to do the mutual introduction of participants on the evening before the
course began have each participant (including the facilitators), introduce themselves. Make
certain everyone indicates their profession, title, major activities, and benefits they hope to
obtain from the course. This may be done by having participants interview each other and
then introduce the person they interviewed.
7.
While introducing themselves, have each participant state what she/he feels are the priority
research needs within the context of their work and list these on a flipchart. Utilize this list
to emphasize that there is a wide variety of possible research topics and to have partici
pants begin thinking about areas they might be interested in exploring further.
8.
Present any other information concerning the course and administrative arrangements that
may be necessary and ask for final questions.
38
Trainers Notes
Module 2 - INTRODUCTION TO HEALTH SYSTEMS
RESEARCH
Teaching Methods and Timing:
25 minutes
1/2 hour
20 minutes
Introduction
Brainstorming session on guidelines to HSR
Discussion on country approaches to HSR
1-1 /4 hours
TOTAL TIME
Introduction
1.
Hold a discussion on the definition of health systems research and important problem areas
for possible investigation.
2.
Stress that health staff and health managers are important categories of researchers in
health systems research.
3.
Encourage participants to give examples of HSR topics that might address certain priority
problem areas and ask for their own experience in HSR. One or two projects could be
presented in more detail.
4.
Discuss reasons for increased interest in health systems research.
Lead a Brainstorming Session on Guidelines for Health Systems Research.
1.
Ask the participants to contribute ideas on the guidelines for health systems research that
are necessary to make a real and practical contribution to the improvement of health.
2.
Put the participants’ suggestions on a blackboard or flip chart and discuss some of the
possible guidelines.
WHO/AFRO/SHDS
Joint HSR Project
39
Plenary Discussion on Country Approaches to HSR
1.
Discuss the points on page 6 module 2 in plenary
2.
Make certain that you as a facilitator know what mechanisms and organizational structure
are in place in your country.
3.
It might be useful to invite the head of the national research council or a member of the
Health Research Unit of the Ministry of Health to provide the above mentioned information.
For the Following Session
1.
Introduce module 3.
2.
Encourage participants to think of one or more topics they would be interested to examine
in detail.
40
Trainers Notes
Module 3 - SELECTING A RESEARCH PROJECT
Teaching Methods and Timing
1 /2 hour
20 minutes
10 minutes
2-1/2 hours
1/2 hour
Introduction and discussion
Exercise: Chobe District
Explanation of group discussion method
Group work
Group reporting
4 hours
TOTAL TIME
Introduction and Discussion
1.
In this session each sub-group (working group) will select the project on which it will work
for the rest of the course.
2.
Before the session, read the instructions for the various approaches to the group work thor
oughly. Be sure you have a clear idea of how the three alternative approaches to be con
sidered in selecting a research topic can be applied and what their advantages and disad
vantages are. If, to save time, you as facilitators have decided beforehand for the third op
tion, it is still useful to present all options to the participants so that they are aware of the
reasons for your choice.
3.
Make certain that participants have been familiarized with the rating technique using the ex
ample of the two research topics given in the exercise.
WHO/AFRO/SHDS
Joint HSR Project
41
EXERCISE 1:
The Chobe District Health Team Selecting a Re
search Project
1.
Discuss briefly the background explanation for the two topics.
2.
Discuss the example of a score card given in the module.
3.
Before going into detail on the Chobe District Health Team scoring, ask each participant to
sketch their own score card on the blank score card in the module.
4.
Ask each participant to rate the two research topics against the 8 criteria discussed earlier.
5.
Collect the scores of each two groups of participants and total them on a transparency or
flipchart.
6.
Have participants discuss the reasons for rating each topic, looking at one criteria at a time.
7.
Have a general discussion on the total rating scores for the two topics. A debate is likely
to develop. This could serve as a useful moment to illustrate the criteria discussed earlier.
Participants should be made to see the importance of looking at all dimensions of a prob
lem before moving ahead to select their own topics.
Group Work for Selection of Research Topics
1.
When the selection process for choosing group topics is introduced, make sure that the
participants realize they are involved in more than a “hypothetical exercise”. Participants
should be made aware that they will be developing the topics they select throughout the
course and that they will carry out this project on their return home.
2.
Review with the other trainers the logistics of supervising the group work selected. Choose
meeting places for the four groups and be sure flip chart papers, tape or thumb tacks and
markers are available.
3.
The alternative ways of organizing this group work for selecting projects take 2-1/2 hours
(alternative 1 and 3) to 3-1 /2 hours (alternative 2). Experience has shown that it is impor
tant to allow as much time as possible for this critical step in the course. If necessary this
session may even spill over into the evening for some groups. Make certain to encourage
participants in their thorough discussion and debate so that the final choice is clear to ev
eryone.
4.
As facilitators, you should bo present in the sub-groups but not dominate the discussion.
You should only make sure that the procedures run smoothly and that overlaps or unfea
sible projects are dropped before the rating starts. Care should be taken as well that no
important proposals or initiatives of certain group members are dropped because the sub
group is not yet familiar with handling the criteria.
42
5.
Once a “short list” of topics has been developed (i.e. the top 2 or 3 after prioritization), a
final criterion should be used. Have each group participant rate each of the top 3 projects
according to whether or not he/she feels it can be realistically carried out by their own
group within the 4-5 months allocated.
6.
At the end of the group work for selecting projects, assist the rapporteur in editing and writ
ing on a flip chart the list of topics debated by the group, along with the record of the crite
ria rating. Ask the secretarial staff to type out lists and voting results for possible inclusion
as an Annex of the final course report.
7.
Have the groups reconvene to present their list of topics and their final choice. (1/2 hour)
43
Trainers Notes
Module 4 - STATEMENT OF THE RESEARCH
PROBLEM
Teaching Methods and Timing
1/2 hour
2-1/2 hours
1 hour
Introduction and discussion
Group work
Plenary
4 hours
TOTAL TIME
Introduction and Discussion
Hold a discussion concerning the development of the first part of a research proposal. En
courage participants to brainstorm the Why, How, Who, and What’s of Problem Definition as
it is appropriate within their context . Use a flip-chart to list their comments.
Discuss the importance of defining difficult terms and abbreviations which will be used in
the proposal.
Group Work
Ask the participants to work in their sub-groups, preparing a statement of the problem for their re
search proposal and defining the terms which need clarification. The groups should try to present
the statement of the problem clearly, supplemented by a diagram of major influencing factors.
Plenary Discussion
When the groups have finished, reconvene. Each working group will then present their
statement of the problem followed by plenary discussion (15 minutes per group).
The facilitators must ensure that the statement of the problem for each group, is clear and
provides the essential elements that will be necessary to develop an appropriate set of ob
jectives in Module 6.
WHO/AFRO/SHDS
Joint HSR Project
44
Trainers Notes
Module 5 -
REVIEW OF AVAILABLE LITERATURE
AND INFORMATION
Teaching Methods and Timing:
1/2 hour
3 hours
Introduction and discussion
Group work
3-1 /2 hours
TOTAL TIME
Materials
Annex 5.1, 5.2, 5.3, 5.4; Examples of:
Abstract
Index Medicus
Index card
Computer card
2 blank index cards or blank sheets of paper for each participant
Obtain copies of “Current Health Information Zimbabwe” (available from Box A178, Avon
dale, Harare, Zimbabwe) or any other abstracting journal for distribution or display for
participants
Ensure that course library is ready (see page 4 Course Preparation)
Prior to this session, facilitators should look through their own resources to find relevant
additional articles for each research topic.
WHO/AFRO/SHDS
Joint HSR Project
45
Introduction and Discussion
Discuss why and how to do a review of the literature. Have participants suggest answers
to the questions, but provide additional information when necessary.
Refer to Annex 5.1, a book index and reference list to explain ways of finding infor
mation relevant to a specific research topic
It may be useful to have the assistance of a librarian in this session
Provide information of national library facilities
Stress the importance of developing libraries at all management levels in organiza
tions and Ministries concerned with solving health problems
Present the points concerning preparation and use of index cards.
Discuss possible biases in documents and literature reviews
Ask for comments or questions concerning the review of the literature and problems partici
pants are likely to face.
Group Work
Ask the group to carry out a review of literature and information relevant to the proposal which they
will complete when they return home.
As a first step, each participant should review at least two articles or reports/books, using index
cards or blank sheets of paper then the information should be put together in a review of 1 -2
pages.
Emphasize that the information should be used to support the proposed research (or not), and be a
thorough and critical review.
46
Trainers Notes
Module 6 - FORMULATION OF RESEARCH
OBJECTIVES
Teaching Methods and Timing
1/2 hour
Introduction and discussion
Group work
Exercise 2: Assess the objectives of one of the other groups
Presentation by each group, followed by comments of the group
that did the exercise and general discussion
Adjustments
4 hours
TOTAL TIME
1/2 hour
1-1 /2 hours
1 /2 hour
1 hour
Introduction and Discussion
The formulation of clear and appropriate objectives is critical to the development of all the
other components of a research design and proposal. This should be emphasized.
Formulation of good objectives is a skill with which many participants have difficulty. Two
types of problems come up quite often:
Difficulties with developing concise, measurable objectives that focus clearly on
what the study hopes to accomplish;
Difficulties in understanding the difference between programme objectives and re
search objectives. For example, many participants may not, in the beginning, see
the distinction between a programme objective, such as, "To improve the Primary
Health Care programme in District X” and a research objective, such as "To com
pare two methods of providing Primary Health in District X”.
Efforts should be made during the session to assist participants in clarifying their
understanding of these problems.
WHO/AFRO/SHDS
Joint HSR Project
47
Group Work
Plan on giving ample time for the groups to formulate objectives for their chosen projects.
trainers should be present during group work to provide advice and guidance, as needed.
The
EXERCISE 2:Formulation of Research Objectives
Hold an exercise in which groups evaluate the objectives prepared by another group, using the cri
teria set out on the exercise sheet.
Plenary Session
Have each sub-group present the objectives they have developed, immediately followed by com
ments of the group that analyzed them during the exercise, and then by general discussion (15
minutes per topic).
By now, also the title of the research project should be formulated, matching with the objectives.
When the results of the group work are presented, it is important that each group receive good
feedback on the quality of the objectives they have developed thus far, as well as practical sugges
tions for improvement. When providing feedback, ask yourself.
1.
Do the group’s objectives really measure what they want them to?
2.
If the objectives were obtained, would the study really provide the results needed to solve
the problem posed in the statement of the problem?
3.
Are the objectives not too ambitious? If so, could the scope of the study be reduced?
4.
Is the title specific enough, and does it cover the objectives?
Adjustments
In past courses it has been found that it is often useful to provide a second group work session for
participants to finalize their objectives and title of the research project, after they have received
feedback from the group as a whole.
48
Trainers Notes
Module 7 - INTRODUCTION TO METHODOLOGY
Teaching Methods and Timing:
1 /2 hour
Introduction and discussion
1/2 hour
TOTAL TIME
Guidelines for Trainers
List and explain the components of a good research design as outlined in the presentation.
WHO/AFRO/SHDS
49
Trainers Notes
Module 8 - VARIABLES
Teaching Methods and Timing:
Introduction and discussion
Identification of variables in research, and discussion
Exercise 3 :
1 /2 hour
1 /4 hour
1-3/4 hours
Group work
2-3/4 hours
TOTAL TIME
Introduction and Discussion
Give an introduction to variables and their use in research, followed by discussion.
Exercise
Carry out the exercise on “Identification of Variables in Research,” in plenary and ask the
participants to read and fill out the exercise individually or in small groups of 2 or 3 people.
When they have finished hold a discussion on the answers. (Suggested answers are on the
following 2 pages.)
Group Work
Ask the participants to meet in their working groups to select the variables that will be involved in
the study being designed.
Participants should determine what variables will be measured in their study. The factors
they identified as influencing the problem in Module 4: statement of the research problem
can serve as a point of departure. Now these factors should be translated into variables
that are measurable.
They should be encouraged to elaborate a diagram, indicating how the variables are related
to the problem, and how they are interrelated.
They should then prepare a short description of the variables involved in the study for inclu
sion in the methodology section of their proposal.
WHO/AFRO/SHDS
Joint HSR Project
50
EXERCISE 3: Identification of Variables in Research
Answer Sheet
The following answers are by no means exhaustive:
1.
Dependent variables
Presence or absence of malaria
Presence or absence of anaemia
Presence or absence of malnutrition
qualitative
Independent variables
age
(quantitative)
occupation
(qualitative)
educational background
2.
(quantitative or qualitative)
Dependent variables
Availability of specific drugs for village health workers
Independent variables
Number of meetings attended (again influenced by distance to Health Centre)
Availability of drugs at Health Centre (again influenced by frequency of ordering and
frequency of supply)
Willingness of staff to supply drugs to VHWs
Amount of drugs monthly supplied to VHWs
3.
Important independent variables that should be taken into account include:
Division of labour
Activities according to age group
Sex
Contact with water
Season
Location in the village
51
Closer study revealed that among younger farmers between 20-25 years of age schistosomiasis was
present in 70%, while it was almost entirely absent in older farmers above 50 years. It turned out
that younger farmers tended to have farms much farther away from the village where the land was
more fertile and had to cross a river where they bathed on their way home in the evening. The
older farmers on the other hand had always had their farms close to the village and obtained water
from wells.
52
Trainers Notes
Module 9 - STUDY TYPE
Teaching Method and Timing
1 hour Introduction and discussion
1 hour TOTAL TIME
Introduction and Discussion
It is helpful if participants can be asked to study the session on “Selection of Study Type” the night
before, so that there will be no need to spend a great deal of time describing the types of study. If
participants have adequately reviewed the material beforehand, the session can focus on clarifica
tion of remaining questions and on how to use the studies in typical field situations. An exercise on
selection of study type is given in Module 10 (combined with selection of data collection tech
niques).
WHO/AFRO/SHDS
Joint HSR Project
53
Trainers Notes
Module 10 - DATA COLLECTION TECHNIQUES
The sessions “Types of Study” and “Development of Data Collection Techniques” will take a whole
day, and may even spill over in the evening or Saturday morning to finalize the data collection
tools.
Teaching methods and Timing
2 hours
2 hours
Introduction to data collection techniques, with discussion
Exercise 4:
Selection of study type(s) and data collection
technique(s)
Introduction to questionnaire design and bias, with discussion
Group Work: selection of study type and data collection tech
niques; Development of data collection tools
Exercise 5: commenting on data collection tools of other groups
Plenary discussion of data collection tools
9 hours
TOTAL TIME
1/2 hour
1 hour
1/2 hour
3 hours
Introduction Data Collection Research Techniques and Discussion
1.
Give an overview of the various data collection techniques
2.
Make the connection clear between type of study (qualitative or quantitative for example)
and data collection techniques selected
3.
Explain the difference between data collection techniques and data collection tools
4.
Let the participants mention possible advantages and disadvantages of the various data col
lection techniques. Stress the advantages of combining various techniques
WHO/AFRO/SHDS
Joint HSR Project
54
EXERCISE 4: Selecting Types of Study and Data Collection Tech
niques
1.
Exercise 4 is designed to give participants some experience in choosing types of study ap
propriate to typical research situations before they have to select types of studies for their
own proposals.
2.
Stress that objectives, if well formulated, should help to determine the appropriate study
type(s).
3.
The exercise can be best conducted after the first part of the introduction to data collection
techniques as the participants are asked to select study types and data collection tech
niques for different research problems.
4.
Ask participants to go into sub-groups of 4-5 persons to do the exercises. Each sub-group
may take 2 topics. Allow 15 minutes for this. Then, ask the groups to report and present
the answers on flipcharts, facilitating plenary discussion.
ANSWER SHEET TO EXERCISE 4:
Selection of Study Type and
Development
of
Data
Collection Techniques
Selection of Study Type and Development of Data Collection Techniques
1.
Study types:
Descriptive (qualitative & quantitative) and Analytical (case control)
Quantitative study, to establish magnitude of the problem;
Qualitative study, to find out opinions about goitre, explore possible causes and get
opinions on possible interventions;
Case-control study, on a sample of cases and a control group selected from the
quantitative study to test hypotheses on possible causal relationships.
Data collection techniques:
questionnaires and medical examination for the quantitative study;
focussed group discussions and a number of interviews with key informants (cases,
non-cases) for the qualitative study;
questionnaires and observations on behaviour for the case-control study.
55
2.
Type of study:
Descriptive, qualitative
Data collection technique:
Participant observation, concealed. A number of researchers receive a short train
ing in spraying procedures, and mix among the spraying teams. They find out that
the sprayers dump most of the insecticide in the morning, so that their load is
lighter in the afternoon. The villages sprayed in the afternoon are underserved.
(Foster GM. World Health Organization behavioral science research: problems and
prospects. Social Science and Medicine 1987; 24:709-717.)
3.
Types of study:
Descriptive, quantitative study:
(analyzing risk factors recorded on antenatal cards, as far as possible, comple
mented by risk factors known from literature review);
Descriptive, qualitative study:
(concentrating on possible socio-economic factors contributing to low birth weight).
Prospective cohort study:
(analyzing a group of pregnant mothers recruited at the antenatal clinic, who are
characterized by certain risk factors identified above of giving birth to underweight
babies, and a control group of mothers among whom those risk factors are absent).
Data collection techniques:
Thorough history taking; measuring of mothers’ body mass index W/H2 and growth
during pregnancy; lab tests on Hb, sugar, protein, bloodsmear for malaria; request
to mothers who deliver at home to have babies weighed and examined one week
after birth (if not coming: follow up);
Focused group discussions with women in the villages about possible socio-eco
nomic factors contributing to low birth weights;
Interviews with a group of mothers at high risk and a group at low risk to give birth
to low birth weight babies on possible socio-economic factors;
Analysis of (now well filled in) antenatal clinic cards on mothers included in the co
hort.
56
4.
Study types:
descriptive study, with qualitative as well as quantitative parts, to examine the health
needs of the community;
descriptive study of the functioning of the health services, including the quality of
the facilities.
Data collection techniques:
group interviews with key figures in the community;
individual interviews with community members;
individual and group interviews with staff members;
observations, both in the community and in clinics;
analysis of clinic records.
5.
Study type:
descriptive study
Data collection techniques:
observation of time, magnitude and duration of queuing in various sections of the
O.P. department;
interviews with staff to explore reasons and possible solutions;
interviews with patients to establish the duration of waiting time in relation to the
time diagnoses/ treatment takes, and their opinion on causes and possible solu
tions.
6.
Study type:
descriptive study;
case-control study to compare possible causes of irregular attendance among regu
larly and irregularly attending patients.
Data collection techniques
analysis of records to identify size of the problem and point of time in which pa
tients are most likely to default (after diagnosis, after admission, or at end of treat
ment);
57
group interviews with key figures in the community to get insight in perceptions of
causes, symptoms and treatment;
individual interviews with
staff members (to get their opinion on bottlenecks in services and causes
for irregular patient attendance);
patients attending services with varying regularity;
community members (neighbours of patients, e.g.).
observation of behaviour of relatives and neighbours towards TB patients (fear or
not).
Introduction to Questionnaire Design and Bias, and Discussion
1.
Explain the successive steps in questionnaire design.
2.
Discuss advantages and disadvantages of open and closed questions.
3.
Discuss various types of bias and let the participants come up with some examples, and
propose ways of reducing possible bias.
Group Work
1.
Have the groups select the type(s) of study and data collection techniques appropriate for
their proposal.
2.
Let them develop their data collection tools.
The data collection tools and techniques will most likely have to be revised after this exercise. They
will be checked in the filed during the pretest, and will thereafter have to be finished before that.
This may imply overwork. In their final form, they will have to be attached to the research proposal.
58
EXERCISE 5:
Commenting on Data Collection Tools of Other
Groups
Have each working group review and critically comment on the data collection tools of the others
groups. Each group should be allocated one other group to give particular attention to.
Plenary
The plenary discussion has to be well structured to allow for as many useful comments as possible
within the time limit of 1 /2 hour per group. First ask the rapporteur of one group to summarize
their comments on the data collection tools of the group they have been concentrating on. There
after one spokesman from each of the other groups may comment. Instruct the spokesman to
avoid duplication of comments. Finally the group may respond briefly to some of the comments.
The same procedure is followed for all four groups. It should be made clear before starting the ple
nary, that the aim of this session is to improve the data collection tools and not to provoke an “at
tack and defense” discussion.
59
Trainers Notes
Module 11 - SAMPLING
Teaching and Timing
The topic on sampling has two components which may be presented in two sepa
rate sessions. These sessions will require 5 hours in total.
Materials
Calculators
Introduction to Sampling Procedures, and Bias in Sampling
1 hour
1 hour
1/2 hour
Introduction and discussion
Group work
Exercise 6
Give an introductory lecture on sampling methods and procedures and bias in sampling,
using examples, when possible, and allowing time at the end for questions and discussion.
Have the participants in their working groups determine the sampling procedures to be
used for their proposals.
Ask each group, as an exercise, to critically examine another group’s chosen sampling pro
cedures and look for possible sources of bias. Encourage the groups to make suggestions
for reducing the bias.
WHO/AFRO/SHDS
Joint HSR Project
60
Introduction to Sample Size
1 /2 hour
3/4 hour
1-1/4 hours
Introduction and discussion
Group work
Group reports in plenary
Give an introduction to factors that must be considered when deciding on sample size. Try
tq use specific examples that you are familiar with.
Let each group determine the sample size for the proposal they are working on. The
groups preparing qualitative research will need less time for calculating their sample size
than groups doing quantitative research. However, the former most probably will need
more time for developing their data collection tools, so that all groups will be ready at
about the same time for conducting the Pre-test.
Formulae for calculating the sample size are provided in Annex 11.2 to help you in case
you need them during group work. You are not expected to go into technical details of
sample-size calculation during your lecture. Participants should be told to consult people
with statistical training when they think they will need to calculate sample size (depending
on the type of analysis and tests they wish to undertake) but do not know how to go about
it. Ensure for this reason that there is a statistician present who can be consulted during
group work and plenary presentations.
Problems of calculating certain statistics (for example infant mortality rates or maternal mor
tality rates) should be discussed. When such rates are envisaged, participants should be
told to always consult a person with statistical training.
Plenary
Have each group present their sampling methods and sample size, immediately followed by the
comments of the group that examined the sampling methods on bias, followed by plenary discus
sion (15 minutes for each topic).
61
Trainers Notes
Module 12 - PLAN FOR DATA COLLECTION
Teaching Methods and Timing:
1 /2 hour
1-1/2 hrs
Introduction and discussion
Group work
3-1 /4 hrs
TOTAL TIME
Introduction and Discussion
Hold a discussion on issues to be considered when planning data collection.
Group Work
Ask the participants to divide into their working groups and prepare a plan for data collection for
the research proposal they are developing, using the outline of points given in the module.
WHO/AFRO/SHDS
Joint HSR Project
62
Trainers Notes
Module 13 - PLAN FOR DATA ANALYSIS AND
INTERPRETATION OF RESULTS
Teaching Methods and Timing
1 /2 hour
1 hour
2-1 /2 hours
Introduction and discussion
Exercise (in plenary)
Group Work
4 hours
TOTAL TIME
Introduction and Discussion
1.
At the start of the session ask the participants to fill in the questionnaire to be used in the
exercise on data coding and analysis.
2.
Give a brief introduction to the topic.
3.
Explain and clarify terms such as sorting, quality control, tallying, and analysis.
4.
Discuss aspects of a plan for organization and analysis of data.
WHO/AFRO/SHDS
Joint HSR Project
63
Exercise 7: Smoking (Data Analysis and Interpretation)
An exercise is presented here, to give participants a “feeling” for data analysis. It may also help
them in the preliminary analysis of their research project before the second two-week workshop.
Ensure that you know what is required of the participants in each step of the exercise before start
ing on that part.
Part 1. Divide the participants in sub-groups of 6, while keeping them seated in plenary.
Collect all questionnaires and divide them equally over the sub-groups. Ask each
sub-group to number their 6 questionnaires and to fill in the data mastersheet.
Part 2. Let the sub-groups prepare frequency distributions for the data, by deciding on
categories followed by tallying.
Part 3. Instruct the participants on how to make cross-tabulations. The facilitator should
explain every table so that the participants understand what they are doing. After
filling in the tables, ask the participants to answer questions one to six.
Part 4. After having done part 1 to 3 in plenary, ask participants to divide into their working
groups. Before starting on the group work they may complete part 4 of the exer
cise. Assist the participants in carrying out part the analysis of answers to an openended question. To ensure a wide range of responses, these have been provided
in the exercise, rather than using the real responses. One way you might catego
rize and interpret the answers is given below. For analysis of a large number of
questionnaires each category could be given a code (number or letter).
Answer Sheet to Part 4
1.
Pleasure
Because it gives me pleasure
I like the feel of the cigarette in my hand
Because I like to blow the smoke through my
mouth and nose
I like to blow smoke rings
I like the taste
64
2.
Cannot give up
I have tried to give up so many times but I have
been unable to
It is too difficult to give up
3.
Being sociable
All my friends are smokers
It helps to make people more friendly and comfort
able, especially when offering a cigarette
4.
Status/confidence/respect
Because I feel confident and in charge when I am
smoking
I like the image that comes with smoking
I feel that people respect me more as a smoker
Smoking makes me feel like a man
My wife likes a man who smokes
5.
Reduction of tension
It helps me to relax
It helps me to reduce the work pressure and ten
sion
It helps me to think better
6.
No reason/defiance
I do not see why I should give up smoking
Why not
The above statements may be interpreted as follows:
Smoking seems to be very much a social activity as most of the responses were concerned with
status and being sociable.
In terms of possibility to influence the respondents to change their habits, categories 1 and 2 might
be the easiest groups: category 1 because there are many other things in life that give pleasure
and are not harmful for one’s health; category 2 because it is already motivated to stop smoking
and, with an extra encouragement, might succeed.
Category 3, might also be convinced that there are many other ways of being sociable which are
less dangerous.
65
Category 4, who need smoking to give them more security, might be more difficult to convince.
One would perhaps have to find out why they are insecure, or in what other way they could com
bat their insecurity (which might differ from person to person).
Category 5 may even have more personal problems and will probably need repeated group or indi
vidual counselling.
Category 6 is defensive, and therefore most likely to be the least approachable.
Facilitators have to stress that, by categorizing the answers this way, they do not only quan
tify the more common reasons but at the same time acquire an insight in how they could at
tack the problem.
As a final aspect of this exercise, have the participants utilize their findings in making a list of rec
ommendations or action steps.
Group Work
Let each group prepare a plan for data analysis and interpretation for their research proposal, using
the outline given in the module.
66
Trainers Notes
Module 14 - ETHICAL CONSIDERATIONS IN HEALTH
SYSTEMS RESEARCH
Teaching Methods & Timing
1/2 hour
3/4 hour
Introduction and discussion
Group work
1-1/4 hour
TOTAL TIME
Introduction and Discussion
Introduce the subject of ethical considerations in health systems research and let the participants
give examples of ethical considerations in each phase of research. This can be done briefly by
writing some examples on the flipchart as you go through each phase of the research process.
Group Work
1.
Ask the participants to go over their proposal and discuss whether they have given enough
consideration to ethical issues in the development of their research project.
2.
Although the topics have not yet been covered, encourage them to think of ethical issues
that might come up during data analysis and dissemination of results.
3.
The end results of this session should be a paragraph or two on ethical considerations.
4.
Provide the necessary information about local procedures to obtain consent for the implem
entation of their research proposals. (Include names and addresses of relevant national au
thorities, etc.)
Obtain specific forms used in the country (if required), so that they can be distributed to the
participants and filled out with assistance from the facilitators.
WHO/AFRO/SHDS
Joint HSR Project
67
Trainers Notes
Module 15 - PRE-TESTING THE METHODOLOGY
Teaching Methods and Timing
3/4 hour
1-1/2 hours
3 to 4 hours
1 hour
2 hours
1 hour
1 hour
Introduction and discussion
Group work to prepare field exercise
Field exercise
Discussion of results
Revision of instruments
Reporting:
- writing
- in plenary
10 to 11 hours TOTAL TIME
Guidelines for Trainers
Discuss the concept and process of pre-testing or conducting a pilot study of the methodol
ogy, covering the questions listed in the introduction.
Refer to annex 15.1 “Summary of Points to be Checked During a Pre-test or Pilot Study”.
Using the annex, review quickly the important aspects of pre-testing covered in the session.
Discuss the field exercise that will be undertaken, so that each working group can pre-test
one or more of their data collection techniques.
Group Work
Ask the participants to meet in their working groups to design a pre-test for their project.
WHO/AFRO/SHDS
Joint HSR Project
68
Pre-test
Arrange a field exercise during which each working group can pre-test some research components
(preferably including their data collection techniques/tools).
Undertake the pre-test exercise, during which members of each working group will pre-test
some research components in an appropriate study area (3 to 4 hours). Make certain con
sent has been arranged beforehand from the relevant authorities/study population.
Allow each working group after the field exercise, to meet separately and to discuss and
analyze their experience. Finally ask each group to prepare a short report on what they
have learned. They may then make any needed changes in their data collection tools and
other aspects of the methodology, based on their pre-test experience.
Ask each group to present their report in plenary and allow some time for discussion (15
minutes per group).
69
Trainers Notes
Module 16 - UTILIZATION AND ACTIVE
DISSEMINATION OF RESULTS
Teaching Methods and Timing
1 /2 hour
1-1/2 hour
1 hour
Introduction and discussion
Group work
Plenary session
3 hours
TOTAL TIME
Introduction and Discussion
1.
Give a brief introduction into the reasons why health systems research should be concerned
with utilization and dissemination of results, to whom the results should be fed back, and
how.
2.
Lead a discussion on the utilization and active dissemination of results for each phase of
the project.
3.
Discuss the format of Annex 16.1 “Outline of a sample research report’’.
only elements of the outline may be suitable for a particular report.
Emphasize that
Group Work
Ask the participants to divide into their working groups and prepare a plan for disseminating and
promoting utilization of the results of their research using the checklist in annex 16.2. The plan
should be included in the project proposal they are developing.
Plenary Session
Have each group present their plan for utilization and active dissemination of results followed by
discussion in plenary.
WHO/AFRO/SHDS
Joint HSR Project
70
Trainers Notes
Module 17 - PROJECT WORK PLAN
Teaching Methods and Timing
3/4 hour
3-1 /2 hours
1 hour
Introduction and discussion
Group work
Plenary
5-1 /4 hours
TOTAL TIME
Introduction and Discussion
1.
Introduce and discuss the aims and uses of a work plan including a plan on selection of
project staff.
2.
Encourage participants who have had experience in these areas to contribute actively dur
ing the session.
Group Work
Ask the participants to prepare a work plan including an outline on the selection and training of
project staff for their research proposal. The importance of having a feasible, realistic plan should
be stressed.
Plenary
Have each group present its work plan followed by and a short plenary discussion (5 minutes per
group).
WHO/AFRO/SHDS
Joint HSR Project
71
Trainers Notes
Module 18 - ADMINISTRATION, MONITORING AND
EVALUATION
Teaching Methods and Timing:
Presentation and discussion on:
1 /2 hour
1 hour
Introduction and discussion
Group work
1-1/2 hours
TOTAL TIME
Introduction and Discussion
Administration
After reviewing the topics that will be covered in the session, give a brief introductory pres
entation on project administration and its importance. Discuss administrative activities that
should be undertaken by the principal investigator before, during and at the end of a proj
ect. (If there is a participant with experience in this area, he may lead the discussion cover
ing the major points in the presentation).
Monitoring
Discuss briefly what is involved in monitoring a research project and addressing the ques
tions in the presentation.
Evaluation
Discuss briefly what is involved in final evaluation of a research project.
Group Work
Ask the participants to meet in their working .groups and develop a plan for the administration,
monitoring and evaluation of their project, preparing a brief summary for inclusion in their proposal.
WHO/AFRO/SHDS
Joint HSR Project
72
Trainers Notes
Module 19 - PROJECT BUDGET
Teaching Methods and Timing
3/4 hour
2-1/2 hours
1 hour
Introduction and discussion
Group work
Plenary
4-1/4 hours
TOTAL TIME
Introduction and Discussion
Introduce and discuss issues in project budgeting.
Ask the participants to read the sample budget format (annex 19.1) and suggest any typical
items that they feel have been left off. Have everyone write in any new suggestions on their
copies.
Ask the participants to give advice, based on their own experience, on how to prepare a
realistic budget and avoid pitfalls. Include the points in the discussion.
Discuss the importance of budget justification.
Discuss how to realistically reduce budgets.
Group Work on Budget Preparation
Ask the working groups to meet and prepare their project budgets. Emphasize the importance of
developing a realistic budget, that uses as much as possible available resources.
If an external donor is asked for assistance, let the participants make clear in the budget what the
contribution of their own institution/ministry is.
Plenary
Ask each group to present budget in plenary. Allow some time for discussion after each presenta
tion.
WHO/AFRO/SHDS
Joint HSR Project
73
Trainers Notes
Module 20 - FINALIZING THE RESEARCH
PROPOSAL
Training Methods and Timing
1/4 hour
Introduction and discussion
8 hours*
Group Work
Introduction and Discussion
The participants have to be made aware that it is of utmost importance to submit their proposal in
such a way that it reads well for an outsider. It should be comprehensive, to the point and coher
ent.
A brief summary is required for decision makers with interest but little time available. This summary
can best be written when the research proposal is more or less finalized. It should be stressed that
the group invests time in writing the summary, as it is the eye catcher of their proposal.
Group Work
The major task of the facilitator is to assist the principal investigator in distributing writing tasks
among group members, in editing, and in organizing the typing and the making of corrections in
the final draft. All group members should be involved optimally. They should read all sections; the
summary should be discussed by the whole group.
WHO/AFRO/SHDS
Joint HSR Project
74
Annex 1
SUMMARY ORIENTATION TO HSR TRAINING COURSE TO
INFORM AUTHORITIES AND PARTICIPANTS
To All Provincial Medical Directors
The Ministry of Health would like to organize a course in Health Systems Research Methodology,
with support from the Joint WHO/Dutch Project on HSR.
1.
Duration of the Course
The course consists of three parts:
(i)
A two-week workshop during which course participants in small groups will develop
step by step a research proposal on a priority problem in their own working situ
ation that they would like to address.
00
A period of 5-6 months during which the same groups implement the research pro
posal.
(iii)
Another two-week workshop to assist in data analysis, report writing and formulation
of recommendations to help solve the problem.
The first part of the workshop has been planned from 9-22 January 19.. but one week later would
also be possible. The second two-week workshop could take place the last two weeks of July (20
July - 1 August).
During the interlying 6 months, some 4-5 weeks is usually attributed to the implementation of the
research, scattered over the period.
2.
Participants
The first course will take place in the four southern provinces.
Participants should be selected in such a way that they can work together when carrying
out the research, for example: two representatives from the Provincial Health Team, two
from another sector related to health (agriculture or education) from the same provincial
level.
WHO/AFRO/SHDS
Joint HSR Project
75
During the first course, also some representatives from the national level (MOH) and from
health training institutes will participate. At least two of the course participants should be
willing and be able to act as facilitator in subsequent courses.
The idea is to repeat this course next year. In 1990 the four Northern Provinces can send
participants.
3.
Place: Most probably ....
Note:
The dates for the workshops should be fixed at the next meeting of Provincial Medical Di
rectors.
Course Coordinator
76
Annex 2
GUIDELINES ON BUDGET FOR HEALTH SYSTEMS RESEARCH
TRAINING COURSE
The following items will probably have to be budgeted for. Indicate for each item who will cover
the cost (Ministry of Health or donor). For instance salaries of local participants and transport is
usually provided by the Ministry of Health, whereas accommodation and meals are usually covered
by the donor.
1.
Accommodation and meals
Block booking board and lodging for:
24
participants
5
facilitators
2
typists (4 at the end of the workshop)
Make sure to include:
a large conference room
two small meeting rooms
a room for the typists
Also include:
Coffee/tea for 13 days, twice a day, for 31 people.
Consider inclusion of a:
Cocktail for 50 people, after official opening
2.
Salaries and Allowances
Facilitators
salary
allowance
at $
at $
/day x 14 days
/day x 14 days
Participants
salary
allowance if
applicable
at $
/day x 13 days
at $
/day x 13 days
WHO/AFRO/SHDS
Joint HSR Project
77
3.
Transport
For facilitators and participants to come to workshop and return home
For pre-test of methodology: field visits of four working groups of participants.
4.
Stationery
If all duplicating is done using stencils:
500 stencils for use during workshop
200 stencils for the final report
34 rims of duplicating paper (500 sheets each)
1 rim of typing paper
ink for stencilling
If photocopying during the workshop but stencilling the final report:
10 rims of photocopying paper
24 rims of duplicating paper
200 stencils
2 rims of typing paper
ink for stencilling and toner for photocopying
40 A4 note pads, 40 pens, 40 pencils, 40 rubbers, 40 file holders
1 box carbon paper
35 name tags
paper clips, staplers, staples, paper hole punchers, scissors, chalk
200 overhead sheets, markers
5 flipcharts, markers.
78
Annex 3
SAMPLE COURSE SCHEDULE
The course schedule that follows is for a 2 weeks course, but all 14 days will be more or less taken
up by activities, except Sunday (Day 7).
In our experience, it is difficult to have health decision makers abandon their duties for longer than
two weeks. However, it might be desirable to stretch the course over three weeks if at district level
the participants have no previous experience with research at all.
In the present schedule it is taken for granted that there will be a Day 0, during which the HSR
knowledge of the participants can be tested (1 hour) and they introduce themselves (1 hour). This
can be the day of arrival, or even earlier, if feasible.
The official opening is postponed till the evening of the first day, so that health authorities can par
ticipate in the plenary where the working groups will present the topics they selected to develop a
research proposal.
WHO/AFRO/SHDS
Joint HSR Project
79
SAMPLE COURSE SCHEDULE
Responsible
Person(s)
Activity
Time
Day 0
1645 - 1800
1800 - 1900
Test Knowledge of participants of HSR.
Introduction of participants and facilitators (in pairs).
Day 1
0800 - 0845
0845 - 0900
-0900 - 1015
1015 - 1045
1045 - 1145
1145 - 1330
1330 - 1600
1600 - 1630
1630 - 1700
1900 - 2100
Orientation to the Course.
Announcement of Administrative Matters.
Introduction to HSR with Brainstorming
on Guidelines for HSR.
TEA BREAK
Selection of Research Topic, (introduction
and exercise).
LUNCH
and Administrative Matters.
Group Work (including
TEA BREAK)
Official Opening Speeches.
Plenary on Selected Research Topics with
Group Reports.
Forum with Donors sponsoring HSR (optional)
(if not done on Day 4)
Day 2
0800
0830
1030
1050
1130
1230
1330
0830
1030
1050
1130
1230
1330
1400
1400 - 1515
1515 - 1530
1530 - 1730
Evening (if
required)
Introduction to Statement of the Research Problem.
Group Work.
TEA BREAK
Group Work (Contd.)
Plenary
LUNCH
Introduction to Review of Available
Literature and Information.
Group Work
TEA BREAK
Group Work (Contd.)
Finishing the statement of the problem and the
literature review.
80
Activity
Time
Responsible
person (s)
Day 3
0800
0830
1030
1050
1145
1230
1330
1345
1430
1500
1515
0830
1030
1050
1145
1230
1330
1345
1430
1500
1515
1800
Introduction to Formulation of Research Objectives.
Group Work and Exercise
TEA BREAK
Plenary
Revision of Objectives
LUNCH
Introduction to Research Methodology.
Introduction to Variables.
Exercise
TEA BREAK
Group Work
0915
1000
1020
1050
1200
1300
1445
Plenary Presentation of Revised Objectives and Variables.
Introduction to Study Type
TEA BREAK
Introduction to Data Collection Techniques (I).
Exercise on Study Type and Data Collection Techniques.
LUNCH
Forum with Donors sponsoring HSR (optional,
if not done on Day 1)
TEA BREAK
Introduction to Data Collection Techniques (II)
Group work: Selection of Study Type and Data Collection
Techniques.
Plenary
Start with development of data collection tools.
Day 4
0800
0915
1000
1020
1050
1200
1300
1445 - 1500
1500 - 1530
1530 - 1630
1630 - 1715
Evening:
Day 5
0800
0900
1000
1015
1045
1115
1230
1330
1430
1500
1520
1700
0900
1000
1015
1045
1115
1230
1330
1430
1500
1520
1700
1800
Introduction to Sampling Procedures and Bias.
Group Work
TEA BREAK
Exercise on Sampling
Introduction to Sample Size
Group Work
LUNCH
Plenary on Sampling Procedures and Sample Size.
Introduction to Plan for Data Collection.
TEA BREAK
Group Work
Plenary
81
Time
Responsible
Person(s)
Activity
Day 6
0800 - 1230
Finalization of data collection tools;
Revision of first part of research
proposal including literature review.
Day 8
0800 - 0830
0830 - 0930
0930 - 1130
1130 - 1230
1230 - 1330
1330 - 1500
1500 - 1520
1520 - 1600
1600 - 1800
1800 - 1900
1900 - 2100
Introduction to Plan for Data Analysis and Interpretation.
Exercise
Group Work including
TEA BREAK
Plenary
LUNCH
Ethics - Introduction and Group Work.
TEA BREAK
Plenary Ethics
Discussion of each other’s Data Collection Tools in Groups
DINNER
Plenary on Data Collection Tools
Day 9
0800 - 1000
1000 - 1020
1020 - 1100
1100 - 1230
1230 - 1330
1330 - 1400
1445 - 1615
1615 - 1715
1715 - 1900
Revision of Data Collection Tools
TEA BREAK
Introduction to Pre-test (field exercise).
Group work to prepare field exercise.
LUNCH
Introduction to Utilization and Dissemination of Results.
Group Work
Plenary
Finalizing questionnaires
Day 10
0800 - 1200
1200 - 1300
1300 - 1700
Pre-test
LUNCH
Group Work: Discussion of field work and
revision of data collection tools.
82
Time
Responsible
Person(s)
Activity
Day 11
0800 0915
0915 1000
1000 1020
1020 1230
1230 1330
1330 1500
1500 1520
1520 1550
1550 1700
Presentation of pre-test results in plenary.
Introduction to Work Plan.
TEA BREAK
Group Work: Preparation of Work Plan.
LUNCH
Group Work: Finalization of Work Plan.
TEA BREAK
Introduction to Administration, Monitoring and Evaluation.
Group work
Day 12
0800 - 0845
0845 - 1115
1115 - 1130
1130 - 1230
1230 - 1330
1330 - 1345
1345 - 1500
1500 - 1530
1600 evening
Introduction to Budget
Group Work
TEA BREAK
Plenary
LUNCH
Introduction to Finalizing the Research Proposal.
Post-test
TEA BREAK
Finalizing the research proposal
(group work)
Day 13
0800 - 1600
1700 - 1800
1800
Finalizing the Research Proposal
Evaluation
Closing Ceremony
83
Annex 4
EVALUATION QUESTIONS HSR TRAINING
WORKSHOP IN
1.
2.
Have the objectives of the Workshop been achieved?
1.1
Have you gained the expertise to develop a research proposal?
1.2
Have you developed a proposal which you think can be carried out as a group
within the coming five months?
1.3
Do you feel (1) motivated and (2) confident to start some small research project in
the future in your own working situation?
Do you have any comments on the course content?
Would certain parts need extension?
Could some parts be slimmed down?
Were the presentations clear enough?
3.
What is your opinion on the training method?
type of teaching)
WHO/AFRO/SHDS
Joint HSR Project
84
(compared for example to the “lecture”
4.
Did you find the division of time between lectures, group work and plenary satisfactory
or would you propose more or less time for any of these three components of the course?
5.
Were you satisfied with the way of facilitation?
Would you have any suggestions for similar courses in the future?
6.
How did you function as a group?
Do you feel that every group member had a chance to equally gain from and contribute to
the course?
Would you have any suggestions for similar courses in the future?
7.
Acknowledging that you are all busy people, but that the course was quite compact:
Would you have liked the duration of the course longer, shorter or was it just OK?
8.
What is your opinion on the organization/accommodation/working conditions of the
course?
Do you have any suggestions for the next workshop in this respect?
9.
(Use back of page)
ANY OTHER COMMENTS
85
4’
Course on Health Systems Research
(October 16-26» 1989)
Group A (Operations Res earch)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10 .
11 .
12 .
Dr.A.S.Mohammed (Chairman)
Dr.Ajit Sahay
Dr.S.N.Deb Sarma
Dr.P•Bandyopadhyay
Dr.S.Na&esh
Dr.D.N.Mullick
Dr.Tonmoy Baruah
Rao (Rapporteur)
Dr . S . K . Bhal el.-,.- - M.D.(CHA) Student
Dr.Jaya
Dr . Marv/ah
Dr.(Mrs.) Malhotra
Dr.Batra
Resource Faculty
1.
2.
Dr.A.K.Agarwal
Dr.Y.P•Gupta
Group B (Behavioural Research)
(Chairman)
1 . Shri P.Sivagnanan
2 . Dr.R.P.Singh
3 . Dr.R.s■Phaneendra Rao
4 . Dr.A.Mani
5 . Shri A.D.Pansare
6 . Dr.A.G.Umakantha
7 . Dr.P.G.Pandya
8. Shri P.G.Thakar
9. Ms Ray Chaudhuri (Repporteur)
Resource Faculty
1 . Dr.(Mrs.)S.Bhatnagar
2 . Dr.S.D.Kapoor
Group C (Programme Or i ent ed)
Dr.R.K.Sachar (Chairman)
1.
Dr.A.K.Sood
2.
Ms
Samta Sen
3.
Dr.(Mrs.)
D.D.Pandit
4.
Dr.S.Haider
5.
Dr.S.Quadir
6.
Dr.D.L.Bhatkal
7.
Dr.G.S.Mohan Rao
8.
Dr.M.Raja Reddy
9.
10. Dr.S.G.Ambwani
11 . Dr.Mohammad Yunus Khan
12. Dr.Dinesh Agarwal
Dr '.E.E. Daniel (Rapporteur)
13. L---
Resource Faculty
1 . Dr.(Mrs.)Indira Murali
2 . Mrs.M.Kataria
I
HEALTH SYSTEMS
RESEARCH TRAINING
COURSE
How to develop research proposals
to solve priority health problems
Adaptation for district level of the Health Services Research Course,
developed by the WHO Regional Office for Africa and the
Project for Strengthening Health Delivery Systems in
Central and West Africa
(WHO/AFRO-SHDS)
Joint WHO/DGIS/RTI Project on Health Systems Research
WHO Subregional Health Development Office III
Harare, Zimbabwe
PREFACE AND ACKNOWLEDGEMENTS
In its original form, this course wasdeveloped by staff and consultants ofthe Project for Strengthening Health
Delivery Systems (SHDS) in Central and West Africa, at the request of the WHO Regional Office for Africa
(AFRO). SHDS, under the directorship of Dr David French and funded by USAID, had its technical basis at
the Boston University Health Policy Institute. Dr Ann Brownlee, Dr Thomas C. Nchinda and Dr Yolande
Mousseau-Gershman developed the course materials, in cooperation with local researchers and health
managers who participated in courses conducted in West and Central Africa and commented on earlier
versions. It was published in 1983 in two volumes:
1.
Health services research course:
Howto develop proposals and design research to solve priority problems (371 pp), and
2.
Course guide for administrators and trainers for the health services research course (139 pp)
by the Boston University Medical Center, sponsored by USAID.
The WHO/AFRO SHDS course has a practical orientation. With a number of presentations/discussions,
exercises and work group sessions it takes the participants, step by step, through the process of developing
their own proposal.
Besides in Western Africa, the course has also been conducted in Southern African countries, with support
from IDRC and AFRO: in Zimbabwe (1985), in Swaziland (1986) and in Lesotho (1987), where it was well
received.
Nevertheless, some modifications seemed necessary. It was felt that for use at district level, the course
material should be simplified and trimmed down to the most essential. Moreover, the content should be
more adapted to health systems research, and a larger variety of research methodologies than offered in
the original course presented. The modules on methodology, in particular modules 8-11 (Selection of
Variables; Selection of Study Type; Data Collection Techniques; Sampling) have been expanded and
completely rewritten.
Great care has been taken to maintain the participatory "learning by doing” method and the practical
orientation of the WHO/AFRO - SHDS course which is its major and highly appreciated characteristic.
Focusing on the district level, the present modified version carries this idea even further. When discussing
theoretical issues such as study types, orsampling techniques, for example, the modules start with concepts
with which middle level health staff are most familiar.
Though in a large number of countries in the Subregion the effect of the three AFRO/SHDS courses already
conducted can be felt, the actual implementation of the proposals developed during the courses leaves
much to be desired. Therefore it was decided to adapt the course to include a period of 4-5 months for the
research proposals to be implemented, and then organise a second workshop of two weeks to train the
participants in data analysis and reporting, using the data they themselves have collected in the field.
i
The joint HSR Project, based at the WHO Subregional Health Development Office (SRHDO) in Harare, took
the initiative of revising the WHO/AFRO/SHDS modules. The Joint HSR Project is a joint enterprise of WHO
(HQ and AFRO) and the Royal Tropical Institute in Amsterdam, with financial support from the Netherlands
Ministry for Development Cooperation (DGIS). It promotes health systems research in the 14 countries of
WHO Subregion III.
The rewriting of the modules took place at the SRHDO in Harare from 15-26 February 1988, with a multi
disciplinary group of 10 researchers from five countries whose names appear below. Four of them had
experience as facilitators with the original WHO/AFRO/SHDS modules.
Botswana:
Dr L.Omondi
Chairman, Health Research Unit, Ministry of Health
Lesotho:
Mr T.J.Makatjane
Lecturer, Demography Unit, National University
of Lesotho (NUL)
Dr M.E. Sebatane
Senior Lecturer, Research and Evaluation Division,
Institute of Education, NUL
Dr P. Chimimba
Chairman, Health Science Research Committee,
Ministry of Health
Mr L Msukwa
Director, Centre for Social Research, Zomba
Dr AKitua
Head , Epidemiology and Health Statistics Research
Unit, Ministry of Health
Ms E . Savy
Head, In-Service Education,
Ministry of Health
Dr R. Munochiveyi
Head, Department of Epidemiology and Disease
Control, Ministry of Health
Dr P. Taylor
Director, Blair Research Laboratories
Mr G. Woelk
Head, Research and Development Unit, Harare City
Health Department
Dr C.M. Varkevisser
Manager, Joint HSR Project
Dr M.W. Borgdorff
APO, Joint HSR Project
Mr B. Campbell
(Editor)
Staff Member
PHC Sub-Programme
Malawi:
Seychelles:
Zimbabwe:
WHO SRHDO:
Royal Tropical
Institute,
Amsterdam:
ii
During these two weeks of module writing and the following month in which the draft was finalized, the
administrative staff of the SRHDO provided much appreciated support.
The modules were tested during three 2-week workshops in Malawi (April 1988), in the Seychelles (May
1988) and in Botswana (June 1988). Then another revision took place taking into account the comments
of course participants and facilitators.
Again the office staff of WHO SRDHO was very helpful in finalising the present version. Sue Laver and
Godfrey Woelk (Community Health Department, Medical School, University of Zimbabwe) provided highly
valued assistance in editing the text. Systron (Harare) did the final layoutand prepared the text forthe printer.
20 October 1988
Joint WHO/DGIS/RTI Health Systems Research Project
WHO Subregional Health Development Office
HARARE
iii
TABLE OF CONTENTS
Preface and Acknowledgements
Module
i
1.
Orientation to the Course
1
2.
Introduction to Health Systems Research
7
3.
Selecting a Research Project
15
4.
Statement of the Research Problem
25
5.
Review of Available Literature and Information
31
6.
Formulation of Research Objectives
39
7.
Introduction to Health Systems Research Methodology
45
8.
Variables
49
9.
Study Type
57
10.
Data Collection Techniques
65
11.
Sampling
87
12.
Plan for Data Collection
101
13.
Plan for Data Analysis and Interpretation
109
14.
Ethical Considerations in Health Systems Research
125
15.
Pre-testing the Methodology
131
16.
Utilization and Active Dissemination of Results
139
17.
Work Plan
145
18.
Administration, Monitoring and Evaluation
155
19.
Budget
163
20.
Finalizing the Research Proposal
169
iv
Annex
5.1
Abstracting Journal: Current Health Information Zimbabwe . 174
5.2
Index Medicus
175
5.3
Computer Search
176
5.4
Index Card
177
9.1
Experimental and Quasi-Experimental Studies
178
10.1 Focused Group Discussion
180
10.2 Questionnaire using pictures
181
11.1 How to use random number tables
182
11.2 Formulas for calculating sample size
184
11.3 Explanation on sample size calculation
given in the text of module 11.
186
15.1 Summary of points to be checked during
a pre-test or pilot study
...
188
16.1 Steps in Project Planning and Implementation
Related to Utilization and Dissemination of Results
193
19.1 Sample Budget
194
4
V
COURSE ORIENTATION
Module 1 - COURSE ORIENTATION
COURSE OBJECTIVES
At the end of this course, you should be able to:
1.
Describe what health systems research is and understand the contribution it can
make towards solving priority problems in health care within the local
context
2.
Prepare a health systems research proposal by completing the following steps:
Problem identification
Review of literature and other available information
Formulation of research objectives
Development of an appropriate research methodology
Development of a strategy for distribution and utilization of results
Preparation of a work plan
Identification of resources required and preparation of a budget
3.
Implement this proposal in your own working situation over a period of 5-6 months
4.
Analyze and interpret the results
5.
Prepare a final report including recommendations for implementation of the
research findings
6.
Evaluate the quality of other health systems research proposals
WHO/AFRO/SHDS
Joint HSR Project
2
Who is the Health Systems Research course aimed at?
The HSR course has been developed for mid level managers and junior researchers working in health and
health-related services.
What training method is used in the HSR course?
The training method applied is based upon learning by doing. Participants will themselves develop research
proposals which they will actually carry out in the field.
How long is the course?
The course will take approximately seven months to complete, and will comprise three main parts:
Part 1. The first two-week workshop will provide an introduction to health systems research. Participants
will work in small groups and step by step design a research proposal on a priority problem in their
own working situation. As each new step is introduced, new concepts and research procedures
will be presented. The participantswill immediatelyapplytheseinthe proposal theyaredeveloping.
Part 2. During the following 5-6 months, the same groups of participants will implement their proposal. It
is therefore important that the groups are composed in such a way that they can easily cooperate
during the field work.
Part 3. A final two-week workshop will focus on analysis and interpretation of the results lead ing to a project
report with recommendations for implementation of the results.
In summary, therefore, the HSR course will proceed as follows:
1.
Health systems
research proposal
development
Implementation
of proposal
Analysis, inter
pretation and
reporting
2 weeks
5-6 months
2 weeks
HSR Proposal Development
A number of basic steps have to be taken when developing a research proposal. These steps are presented
in the diagram on page 3.
3
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
I
formulation of
objectives
V
research
methodology
Important elements
of each step
- literature and
other available
information
r- research
questions or
- hypotheses
i- variables
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
*— evaluation
p- material support,
and equipment
L money
summary of proposal
4
2.
Implementation of the proposal
Research proposals developed by groups of participants during the first two-week workshop will be imple
mented during the following five to six month period. Facilitators from the course will visit the research sites
to assist with problems which may arise and to review progress.
3.
Analysis of the data and reporting
After project implementation, participants will meet again for a further two-week workshop to review the
results of their research. During this workshop the data will be analyzed and interpreted. A final report with
recommendations for action will also be drawn up. Since many of the participants are in direct positions of
managerial responsibility, it is expected that recommendations can be implemented soon after completing
the study.
Notes:
By the end of the course, each participant should be thoroughly familiar with this training
document. It can be used both as a reference when preparing future research proposals
and as a training guide for training others in Health Systems Research.
Contribution of Participants
Each participant and trainer brings to this course his own experiences in applied research
and in the management of health or health related projects. Thus, the course should not
be perceived as having a student-teacher orientation. It should rather provide a forum for
sharing information where everyone can contribute the benefits of his or her own
experience and knowledge. This sharing will add greatly to the richness and relevance of
the course.
We hope that each of the participants and trainerswill depart with added knowledge and
skills, ready for further efforts in the area of health systems research.
5
INTRODUCTION TO HEALTH SYSTEMS
RESEARCH
Module 2 -
INTRODUCTION TO HEALTH SYSTEMS
RESEARCH
OBJECTIVES
At the end of this session you should be able to:
1.
Define health systems research and describe the difference between basic and
applied research
2.
List important problem areas and topics that could be investigated through health
systems research
3.
Give reasons for the present increased interest in health systems research
4.
Suggest guidelines on how health systems research can make a real contribution
to the improvement of health
WHO/AFRO/SHDS
Joint HSR Project
8
What is health systems research?
Health systems research aims at supporting the decision making process
at all levels of the health system with relevant information, in order to achieve a more
effective operation of the system leading to an improvement in the health of the
population.
_____
Health systems research concentrates on health and health care within its local context. It is acknowledged
that economic, socio-cultural and political factors may influence health. Health systems research will
therefore not merely concentrate on the contribution that health services can make to the improvement of
health, but will include the possible contribution of, for example, traditional health care, agricultural
extension and adult education services.
Health systems research can be organized at various levels, starting at the community level of health care
and proceeding through the district or regional to the central level. It may focus on issues of policy
development, management and decision making, administration, supply and utilization of services.
Who should be involved in health systems research?
Everyone directly concerned with a particular health or health care problem should be involved in all stages
of a HSR project. Depending on the problem, these may be health workers (including workers in healthrelated fields), managers of health care services, policy makers, community members or professional
researchers. There should be close cooperation with everyone mentioned above from the time that the
research is designed to the implementation of the research results.
Note:
In this and other modules that follow we use the term researcher to mean anyone
actively involved in planning and conducting the research, including health staff.
9
How do basic and applied research (such as health systems research)
differ?
Basic research is designed to extend the base of knowledge in a discipline,
primarily for the sake of understanding itself.
Ultimately part of this new knowledge may be applied, but the researchers are usually not
themselves involved in the application.
Applied research concentrates on finding solutionsto immediate problems of
a practical nature.
It is therefore action oriented. The researchers are usually closely involved in implementing
the research results.
For what important problem areas could health systems research be
useful?
1.
Political: ■ Unsuitable health policies or failure to translate policies into action; ■ inequitable
allocation of resources between rich and poor, urban and rural areas, between various sectors of
development or within the health sector.
Example of a project
An analysis of factors that contribute to an inequitable distribution of health manpower resources
over rural and urban areas, in order to change this distribution.
2.
Economic: ■ low budgets for health; ■ increasing costs; ■ wastage of resources.
Example of a project
A study into the possibilities of introducing a fee for
service, nevertheless ensuring that health
services remain accessible even for the economically least well-off.
10
3.
Technological: ■ difficulties of applying available technologies; ■ technologies inappropriate to
local conditions; ■ lack of local production of equipment and supplies.
Example of a project
An exploration of the possibilities to use simple diagnostic and treatment methods for bilharzia at
health centre level.
4.
Management: ■ Unsatisfactory management processes; ■ low effectiveness, ■ utilization and
quality of health services; ■ low coverage; ■ lack of integration of services.
Examples of projects
A study to ensure that MCH preventive and curative services are properly integrated.
A study to identify ways in which the control of village health activities can indeed to a large extent
be delegated to the village level.
5.
Socio-behavioral: ■ resistance to change; ■ attitudesand practices adversely affecting health;
■ lack of community involvement in decision-making; ■ high levels of social problems affecting
health.
Examples of projects
An investigation of cultural values attached to female circumcision, in order to find ways to stress
the health hazards of female circumcision without offending the community.
A study of factors leading to alcohol abuse.
6.
Ethical: ■ dehumanized medicine; ■ poor patient-physician relationships; ■ objectionable medical
practices; ■ issues of consent of population and patients.
Example of a project
A study on possible coercion exercised on women in the reproductive age groups to accept family
planning methods.
11
7.
Coordination: insufficient coordination between health and other sectors of development.
Example of a project
A multi-disciplinary study on factors leading to malnutrition, in order to address the problem with
all relevant services.
Health systems research can positively contribute to solving problems such as those listed above, by
providing health policy makers, managers, health staff and community members with suggestions for
alternative solutions and, if possible, with evidence for the best approach.
What are the reasons for increased interest in health systems research?
There is increased awareness that while enormous progress has been made in developing health
technologies, the ability to bring these technologies to the most needy is still very limited. As the World
Health Assembly observed, it is not so much a “technology gap” from which we suffer, as an “application
gap”.
This awareness has led to a growing emphasis on planning and evaluation of health interventions. It has
also led to the realization that better knowledge of management issues is essential for correct implemen
tation of these processes.
Basic Guidelines for Health Systems Research
1.
Research should focus squarely on priority problems in health. It should also consider health issues
in the broader context of each country’s overall socio-economic development.
2.
Emphasis should be placed on comparatively simple, short-term research designs which are likely
to yield practical results when needed for decision making.
3.
In order to promote cost effectiveness, the focus should be on low cost studies which can be
undertaken by managers and service personnel, but supported by experienced researchers.
4.
All parties concerned (health authorities, health staff, community members) should be involved in
designing and implementing the study, in the analysis of the results and the formulation of
recommendations.
12
n
5.
Results should be presented in formats most useful for administrators, decision-makers and the
community. The research report should include:
A clear presentation of findings
Honest discussion of practical or methodological problems that could have affected the
findings;
Alternative courses of action that could follow from the results and the advantages and
drawbacks of each.
6.
The effectiveness of the research undertaken should be judged not by the number of papers
published but by its ability to influence policy, improve services and ultimately lead to improved
health in the community.
Discussion points on country approaches to health systems
research.
The following questions could be discussed in plenary:
1.
Mechanisms and organizational structures
What mechanisms or organizational structures are now in place in your country to
promote, coordinate or regulate bio-medical and health systems research.
2.
Constraints
What local constraints at this time seem to hinder development and use of health
system research? Can you see any way to overcome some of the constraints.
3.
Recommendations for HSR
What do you think could be done to improve health systems research and its
contribution to the solution of health problems in your country?
13
SELECTING A RESEARCH PROJECT
Module 3 - SELECTING A RESEARCH PROJECT
OBJECTIVES
At the end of this session you should be able to:
1.
Identify criteria for selection of health related problems to be given priority in
research
2.
Recognize the importance of considering what the needs and possibilities are for
financial and technical support (particularly at the local level), before selecting a
topic
3.
Use the criteria identified and a group discussion .method to set priorities for
research
4.
Select in groups by consensus an appropriate subject for a research proposal that
will be developed during the course.
WHO/AFRO/SHDS
Joint HSR Project
16
A.
ESTABLISHING CRITERIA FOR SELECTING A RESEARCH
PROJECT
Each topic that is proposed for research has to be judged according to certain guidelines or criteria. There
may be several proposals to choose from. Before deciding on a research topic each proposed topic must
be compared with all other options. The guidelines or criteria discussed below can help in this process:
What criteria could be suggested?
1.
2.
3.
4.
5.
6.
7.
8.
Relevance
Avoidance of duplication
Feasibility
Political acceptability
Applicability
Cost-effectiveness
Timeliness
Ethical considerations
It is important that the criteria used should be related to the problem to be studied and the situation under
which your research will be carried out. The criteria discussed here can be used both to prioritize studies
on specific health problems such as mal- nutrition, diarrhoea; or to study management related issues such
as utilization of health services, distribution of manpower, appropriateness of training or effectiveness of
specific health interventions.
How could these criteria be defined?
1.
Relevance
Ideally, the topic you choose should be a priority problem. This may be identified by you and your
team working in an area, together with representatives from that community. Questions to be asked
include:
How big is the problem?
Who is affected?
How severe is the problem?
Try to think of serious health problems that affect a great number of people, or of the most serious
problems that are faced by managers of health systems in the area you work
17
2.
Avoidance of Duplication
Before you decide to carry out a study, it is important that you find out whether the suggested topic
has been researched within the proposed study area or another area with similar conditions.
If the topic has been researched, the results should be reviewed in order to explore whether major
questions that deserve further investigation have remained unanswered. Otherwise another topic
should be chosen.
3.
Feasibility
Look atthe project you proposed and consider the resources you will require to carry out the project.
Thought should first be given to manpower, time, equipment and money that are locally available.
In situations wherethe local resources necessary to carry out the project are not sufficient, you might
consider resources available at the national level; for example, in research units, research councils
or local universities. Finally, explore the possibility of obtaining technical and financial assistance
from outside sources.
4.
Political Acceptability
In general it is advisable to research a topic which has the interest and support of the authorities.
This will enhance the chance that the results of the study will be implemented. However, under
certain circumstances you may feel that a study is required to show that a government policy needs
adjustment. In order tolimitthe chance of confrontation, one should in that case make an extra effort
to involve the policy makers concerned in an early stage.
5.
Applicability
Is it likely that the recommendations from the study will be applied? This will depend not only on
the blessing of the authorities but also on the availability of resources required for implementing the
recommendations. The opinion of the potential clients and of staff directly concerned with the
problem will influence the implementation of recommendations as well.
6.
Cost-effectiveness
The basic question here is: Are the resources of time, money and manpower we are investing in the
study worthwhile given the result(s) that we expect? To answer this question we have to ask as well:
What difference or change will the findings of the study make to existing programmes?
18
Timeliness
Are the findings going to be available in time to enable us to make the necessary decisions? .You
need to be clear about how urgently the results are needed in order to make a decision about which
research should be done first and what can be done later.
Ethical Considerations
8.
How acceptable will the research be to those who will be studied? (Cultural sensitivity must
be given careful consideration.)
Can informed consent be ensured to carry out the study?
Has the condition of the subjects been taken into account? For example, if individuals are
identified during the study who require treatment, will thistreatment be given? What if such
treatment interferes with your study results?
B.
SELECTING A RESEARCH PROJECT
The process of actually setting priorities for research projects and selecting one problem to start with is
introduced with an exercise.
EXERCISE 1:
The Chobe District Health Team, selecting a research
project
(To be carried out in plenary, 1/2 hour)
Introduction to the exercise
The Chobe District Health team, responsible for the health of a population of 8000, has to choose between
two important study topics:
Possibility 1
The first possibility is to determine the nutritional status of all children under five in the district. This is
important because the entire country has been experiencing drought for seven years. However, compared
with other districts, Chobe has received a relatively higher rainfall. The government has to make a decision
whether Chobe District should keep participating in the National Drought Relief Programme. The politicians
and the community members feel strongly that Chobe should continue benefitting from this programme like
the other districts.
19
In order to convince the government, the District Health Team has to quantify the problem of malnutrition
and provide data on the nutritional status of children under three years of age in the district within two months
from now. The team intends to use the < 5 cards, but among children 1 -3 years of age only between 40-50%
attend the MCH clinic, and only few do so regularly.
Q:
How could the District Health Team make a rapid assessment of the nutritional
status of this target group in the district?
Possibility 2
The second urgency is to examine the reasons for the assumingly increasing perinatal mortality amongst
children delivered at the District Health Centre. Various community members have expressed their concern
over expectant mothers returning home from the District Health Centre “without babies”. They are
demanding an explanation from the health workers before they approach the government with the problem.
The District Health Team wants to prevent the community from approaching the politicians and first of all
wants to assess whether the perinatal mortality among children born at the District Health Centre has indeed
gone up over the past 5 years, and if so, what could be the possible explanation.
Q:
How could the District Health Team obtain this information?
Scaling Criteria
You will now use the criteria discussed in this session to select one of the problems to be studied as the first
priority.
Use the criterion scales belowto help you decide which of the problems should be studied as the first priority.
Enter these ratings in the table on page 17.
Relevance:
1.
2.
3.
Not relevant
Relevant
Very relevant
20
Avoidance of Duplication:
1.
2.
3.
Related studies available
Related studies available but major issues not covered
No related studies available
Feasibility:
1.
2.
3.
Not feasible considering available resources
Feasible considering available resources
Very feasible considering available resources
Political Acceptability:
1.
2.
Not acceptable
Acceptable
Applicability:
1.
2.
3.
No chance of findings being implemented
Some chance of findings being implemented.
Good chances of findings being implemented.
Cost-Effectiveness: (Judge whether the difference the results can make to existing programs are
substantial in view of the cost of the study)
1.
2.
3.
Cost too high in view of expected effect of study
Cost justified in view of expected effect of study
Cost low in view of expected effect of study
Timeliness:
1.
2.
3.
Not timely
Timely
Very timely
Ethical Considerations:
1.
2.
Not ethical
Ethical
21
Priority ratings for two research proposals in Chobe District.
Criteria for Selection
Proposed
6
1
2
3
4
5
Releva
nce
Dupli
cation
Feasi
bility
Polit
ical
Applica Cost
Effebility
Accept
ability
Topic
7
8
Timeliness
ctive
ness
Ethics TOTAL
SCORE
Possibility
A
Nutritional
Status
Possibility
B
Perinatal
Mortality
Group Discussion Methods
There are different methods that you may like to consider in order to arrive at a group agreement on a
research topic.
Option 1.
Working as a group propose ratings for each research possibility. Review the ratings and try
to reach a group consensus.
This method can be relatively quick but it has the disadvantage that persons with strong
opinions, whether right or wrong, can easily influence final decisions.
Option 2.
Working individually, propose ratings for each criterion for each research possibility (i.e.
each group member having his/her own “blank” table). Follow this by a group discussion
of the points allotted for each criterion.
Individuals then rate the proposals again. The points allotted per criterion for each pro
posed topic are then added and the totals finally compared.
22
Option 3.
The group first discusses the various criteria for each proposal. This is followed by an
individual rating. This helps to ensure that everyone understands the criteria before working
as an individual, whereas each group member has an opportunity to express his personal
opinion.
The totals are calculated and compared as outlined in Option 2.
Note:
Option 3 has usually proved to be the most efficient method as it promotes the attain
ment of an acceptable group decision within a reasonable time.
GROUP WORK
(Approximately 2-1/4 hours)
You are now requested to meet in your working groups to list a number of research topics
that you would consider for the development of a research proposal.
1.
It is suggested that each individual writes one or two topics on a piece of paper.
Then all topics can be listed and briefly discussed to delete duplications. Combine
similar topics or omit proposals that are obviously less relevant or too difficult to
carry out. Ideally you should select no more than six or seven topics for individual
rating.
2.
Give individual ratings, using the scoring sheet on the following page, add the
scores for each proposal per criterion and calculate the total scores.
3.
Now thoroughly review the (two or three) proposals that received the highest
scores. At this point it is important to take into account which proposal can most
realistically
be carried out by your group within the coming 4-5 months. Ideally, this implies
that all group members can participate actively and will directly benefit from the
results.
4.
Finally, selectthe topic of your upcoming researchand prepare a brief presentation
for the other members of your course. List the six or seven topics that have been
considered by the group and provide reasons for your final choice.
23
Module 3 - Rating sheet for group work
Criteria for Selection of Research Proposal
Propo sed
Topic
6
1
2
3
4
5
Releva
nee
Dupli
cation
Feasi
bility
Polit
ical
Accept
ability
Applica Cost
bility
Effe
ctive
ness
1.
2.
3.
4.
5.
6.
7.
8.
24
7
8
Time
liness
Ethics
TOTAL
SCORE
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
*- information
r— research
questions or
- hypotheses
-------- i—
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
L pilot study
or pre-test
r— manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
T
summary of proposal
- administration,
monitoring and
- evaluation
material support,
and equipment
L- money
Module 4 - STATEMENT OF THE RESEARCH PROBLEM
OBJECTIVES:
At the end of the session you should be able to:
1.
State the reasons for including a statement of the problem in your research
proposal
2.
Identify the range of persons who could be involved in problem identification
3.
List the main points to be included in the statement of the problem
4.
Describe the importance of defining terms used in a proposal in order to facilitate
communication and understanding
5.
Prepare the statement of the problem for your research proposal being developed
during the course
WHO/AFRO/SHDS
Joint HSR Project
26
Introduction
The first major section in a research proposal is the “Statement of the Problem” or description of the
background concerning the topic selected for study.
WHY is it important to accurately state and define the problem?
Because you will find that the statement of the problem:
is the foundation for the development of a research proposal (research objectives, methodology,
work plan, budget, etc.)
x
allows you to clarify your chosen topic
enables you to systematically describe the problem, to think about its importance, its priority in the
country and in the local area and to point out why the proposed research on the problem should be
undertaken
can be important as a working tool during the discussion with community members, health staff, the
relevant Ministry and others that will help you with your study
■
facilitates the presentation of your project to the authorities and to possible donor agencies
WHO should be involved in formulating the statement of the problem?
Community leaders, representatives of relevant groups in the community, including minority groups
Programme managers and policy makers
The research team itself (including health staff).
WHAT information should be included in the statement of the problem?
1.
Background information
A short description (one paragraph) of the important geographical and socio-economic
characteristics of the country.
27
Background information concerning health status and health care in the country (includ
ing relevant statistics). Ensure that this information is directly relevant to the problem to be
studied.
2.
Elaboration of the problem
A more detailed description of the problem could include:
■
what needs have been observed or expressed;
■
what is the size of the problem;
■
why is it important to try to solve it?
Main socio-economic, cultural or political factors that may affect the problem.
Services already offered to reduce the problem;
■
possible bottlenecks in the services that should be investigated;
■
position of the Ministry on the problem (if applicable)
A brief description of earlier local studies on the problem, justifying the need for research
(more detail about a literature review will be provided in Module 5).
A summary of how you think that the results of the study could be used to solve the problem.
HOW should the statement of the problem be written for the research
proposal?
The information concerning the problem should be summarized in a concise manner. This should
not take more than 2-3 pages, (i.e. 1 000 words)
An outline listing the major points to be covered should be prepared before any writing is done. This
will ensure that all the points are covered and that the statement is logical and well organized.
In considering possible causes of the problem, all relevant factors should betaken into account. It
is helpful to visualize these factors and their interrelationships in the form of a diagram.
Such a diagram forms a one page summary of the research content. It will be used to formulate research
objectives and to further develop the research methodology.
An example of a diagram is presented on page 23.
It illustrates how various factors may influence
malnutrition.
It is important to note that some factors may have a direct influence on the problem while others affect the
problem indirectly. For example, malnutrition in a child could be caused by illness (such as diarrhoea) and/
or a poor food intake. Mothers education is a factor which could influence both illness and food intake. A
low income or poor access to land may influence the mother’s educational level, but may also be directly
responsible for a lack of food in the household.
28
FLOW DIAGRAM
FACTORS RELATED TO MALNUTRITION
Poor hygiene/
sani tat ion
ILLNESS
Limi ted access
to health service
Low education
of mother
Limited access
to land, money
other resources
Prob 1em:
MALNUTRITION
‘ POOR
FOOD
INTAKE
8
Poor Ditribution
of food within the
families (eg taboos)
ILLNESS and POOR FOOD INTAKE are the
factors directly related to the
problem of malnutrition.
At the same
time, they influence each other
Limited availability
of food in the region
(drought)
All other factors indirectly influence
the problem, through illness and poor
food intake.
Lack of food in
the household
Note:
In order to make understanding of your proposal as easy as possible, prepare a list of
definitions of all special terms and abbreviations used in the study. Ensure that the
meanings are clearly explained. Be careful not to make the list too long. Think of exactly
who will be reading your proposal and define the terms you think they may not be familiar
with. The definition of terms may come right after the statement of the problem, or in an
appendix.
GROUPWORK
(21/2hours)
1.
Prepare a statement of the problem on the topic you have selected in your group.
2.
Make a diagram of the most important factors you think are influencing the
problem.
3.
Ensure to define terms where you think they need clarification.
4.
Critically examine your own statement of the problem, and if possible that of
another group. Ask the following questions:
Is there too little, sufficient or too much background to the statement of the
problem?
Is the problem clearly stated?
Why is the problem important?
Are the factors which are most relevant to the problem adequately
described? (For example : geography, population characteristics, cul
ture, religion, economics, social behaviour).
Have these factors been clearly depicted in a diagram? Does this show
their relationship with each other?
Have there been other studies on the problem? If so, has it been described
how these studies relate to the problem of concern?
What, if anything, has been done about the problem?
Is there a clear indication of how the research results can be used to solve
the problem?
5.
Prepare a flipchart with the title diagram for presentation in plenary session.
30
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
L- information
research
questions or
- hypotheses
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
P material support,
and equipment
money
summary of proposal
Module 5 :
REVIEW OF AVAILABLE LITERATURE
AND INFORMATION
OBJECTIVES
At the end of the session you should be able to:
1.
Describe the reasons for reviewing available information during the preparation
of a research proposal
2.
Describe the resources that are available when carrying out an information review
3.
Prepare an index card that summarizes important information obtained from
literature or interviews
4.
Prepare a review of literature and other information pertaining to the research
proposal which will state background knowledge and information supporting your
intended research
WHO/AFRO/SHDS
Joint HSR Project
32
WHY is it important to review information when preparing a proposal?
■
It prevents you from duplicating work that has been done before.
It helps you to find out what others have learnt and reported on the problem you want to study. This
may lead to refining of your statement of the problem.
It helps you to become more familiar with the various types of methodology that might be used in
your study.
It should provide good reasons for others to support your
proposed research.
WHERE can we find different sources of information?
Many different sources of information can be consulted and reviewed at various levels of health care delivery.
These may include the following:
Examples of Resources
Levels of
Health Care
Clinic and hospital based data from routine
statistics, registers.
Opinions, beliefs of key figures
Clinical observations, reports of
incidents etc.
Local surveys, annual reports
Statistics collected at national,
provincial and district levels
Books, articles, newspapers
mimeographed reports etc.
Community and
intermediate
levels
Documentation and reports/raw data of:
Ministry of Health
Central Statistical Offices
Non-Governmental Organizations
Multilateral organizations e.g. UNICEF, WHO
National Level
Literature reviews making use of
libraries at university,
polytechnics, colleges etc.
International
33
You need to develop a strategy to a get access to each source and obtain information in the most productive
manner. Your strategy may vary according to where you work and the topic under study. It may include
the following steps:
Identify a key person who is knowledgeable on the topic and ask if he can give you a few good
references or/and the names of other people whom you could contact for further information.
Look up the speakers at a conference on the topic to identify key persons.
Contact persons who are currently in key positions and those who occupied those key positions
during the recent past and ask about reports, minutes of meetings, administrative circulars, manuals,
conference reports etc.
Look up the bibliography or reference list in every key paper and book to identify references.
Look up references in Indexes (e.g. Index Medicus) and abstracting journals.
Contact librarians in universities, research institutions, Ministry of Health and newspapers and
request relevant references.
Request a computerised literature search.
Some agencies will assist with your literature review if requested by telephone or in writing. The request,
however, should be very specific and present the statement of the problem in a small paragraph. Examples
of such agencies are:
For Southern Africa:
Joint Project on Health Systems Research
WHO, SRHDO III
P.O. Box 5160, Harare, Zimbabwe
(Free service)
University of Zimbabwe, Medical Library
Box A178, Avondale
Harare, Zimbabwe
(Partly free)
In general:
(Limited free service
available)
IDRC, P.O. Box 8500, Ottawa, Canada
University of Liverpool
Community Health Support Service
Pembroke Place, Liverpool, L3 5OA,
(Charged)
England
Note:
Facilitators may also be able to provide specific information regarding other national facilities.
34
WHAT should be done with the references that are identified?
Firstly they should be skimmed or read.
Then a summary of the important information from the references should be recorded on index cards
(annex 5.4). These can then be classified so that the information can be easily retrieved.
Finally a literature review should be written.
How should information on an index card be organized?
For an article the information should be organized as follows:
Author(s) (last name first). Title of article. Name of Journal Year;Volume numbenpage numbers.
Example:
Gwebu ET, Mtero S, Dube N, Tagwireyi JT, Mugwagwa N. Assessment of nutritional status in preg
nancy: use of a reference table of weight-for-height. Central African Journal of Medicine
1985;31:193-6.
For a book the information should be organized as follows:
Author(s) (last name first). Title of book. Edition.
Place: Publisher, year: page numbers.
Example:
Taylor CE. The uses of health systems research. Public Health Papers 78. Geneva: World Health
Organization, 1984:1.
The index card should contain a few quotations and information such as:
Key words
A summary of the contents of books or articles, concentrating on information relevant to your study
A brief analysis of the content, with comments such as:
appropriateness of the methodology
strong points of the study
how information from the study can be used in your research
35
Note:
Index cards can also be used to summarize information obtained from other sources such
as informal discussions, local health statistics and internal reports.
HOW do you write a review of literature?
There are a number of steps you should take when preparing your review of available literature and
information. To do this:
First organize your index cards in groups of related statements according to which aspect of the
problem they discuss.
Then decide in which order you want to discuss the various aspects. If you discover you have not
yet found literature or information on some aspects that you suspect are important, make a special
effort to find this literature.
Finally write a coherent discussion of one or two pages, using all relevant references.
3
Note:
It is recommended that you use consecutive numbers to refer to your references in the text.
Then list your references in that order, using the format described in the section on index
cards (see page28). Add this list as an annex to your research proposal.
What about bias?
,•
?.■
■
■
;
■
■
■
•.
.•
■'
'
-
Bias in the literature or in a review of the literature can be seen as a distortion of the
available information so that it reflects opinions or conclusions which do not represent the
real situation.
___________________________________________________________________________
It is useful to be aware of various types of bias. This will help you to be critical of the existing literature. If
you have reservations about certain references or if you find conflicting opinions in the literature, discuss
these openly and critically. Such a critical attitude may also help you to avoid biases in your own study.
36
Common types of bias include:
using rhetoric rather than reason to convince the reader, thus putting him on the wrong trail
playing down controversies and results that are different
restricting references only to those that support the point of view of the author
reporting insignificant differences between control and experimental groups as if they were
significant
drawing far reaching conclusions from preliminary or shaky results or making sweeping generali
zations from just one case
Note:
It is important to recognize the relationship between the statement of the problem, the
review of the literature, and the methodology in a research proposal.
GROUPWORK
(2hours)
1.
Outline the topics that need to be researched for the “Review of the Literature” for
your proposal.
2.
List the sources of information you could use for your review (here or later when
you return home).
3.
Search throughthe documents (books, articles and bibliographies) available in the
course library. List the most useful references you can find on your topic.
4.
Summarize the most important information from the references. Place this infor
mation onto index cards.
5.
Prepare a critical literature review for your proposal. A list of the references used
should be presented as an annex.
37
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
information
I— research
questions or
- hypotheses
------------ —
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
L pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
T
1
summary of proposal
- administration,
monitoring and
evaluation
[_ material support,
and equipment
money
Module 6 -
FORMULATION OF RESEARCH
OBJECTIVES
OBJECTIVES
At the end of this session you should be able to:
1.
State the reasons for writing objectives for your research project
2.
Define and describe the difference between general and specific objectives
3.
Discuss the characteristics of research objectives
4.
Prepare research objectives in an appropriate format for the project being
developed.
WHO/AFRO/SHDS
Joint HSR Project
40
WHAT are research objectives and WHY should they be developed?
Objectives o.. research proled summarize „hal is .0 he achioeed 0, ,he mud,.
-
■■
-
-
.
■
Objectives should be closely related to the statement of the problem. For example, if the problem identified
is low utilization of Child Welfare Clinics, the objective of the study will be to identify reasons for this low
utilization, in order to improve it.
If an objective states what is to be accomplished by the study in such general terms,
it is called a general objective.
It is possible (and advisable) to break down general objectives into smaller, logically
connected parts. These are normally referred to as Specific Objectives.
______________
The general objective to determine factors which influence low utilization of Child Welfare Clinics could for
example be broken down into the following specific objectives:
Identify the pattern of utilization of Child Welfare Clinics in district X, in relation to the target set
Determine possible variations in utilization of child welfare clinics, per season and per child
Identify factors offered in the child welfare services which make them attractive or not attractive to
mothers. (This objective may again be broken down in smaller parts, referring to home-clinic
distance, acceptability of the services to mothers, quality of the services, etc)
Identify socio-economic and cultural factors that may influence mothers in their utilization of the
services (again, these may be further broken down)
Make recommendations, with all parties concerned (managers, health staff, mothers) on which
factors might be most easy to change and how
Note:
An objective on implementation of the results should be included in every applied study.
41
The formulation of specific objectives allows us to:
focus the study (narrow it down to essentials)
prevent us from collecting data which are not strictly necessary to better understand and solve the
problem we have identified
organize what we hope to accomplish in the study in clearly defined parts or phases
Properly formulated, specific objectives, also help to guide us in the development of our research
methodology and to orient the collection, analysis, interpretation and utilization of data.
Note:
Objectives have to cover the different parts of the problem as defined under “Statement
of the Problem’’ (Module 4). If the problem has been defined with precision, theformulation
of objectives will become easier.
HOW should you state your objectives?
Take care that the objectives of your study are:
phrased in such a way that they focus on what the study is attempting to solve. They should cover
thedifferent parts of the problem in a logical way;
■
realistic within the local conditions; i.e. it should be feasible to achieve your objectives;
clearly phrased in operational terms specifying exactly what one is going to do. You should keep
in mind that when the project is evaluated, the results will be compared to the objectives. If these
have not been spelled out clearly, the project cannot be evaluated;
stated using action verbs that are specific enough to be measured. Examples of action verbs
include:
To determine, to compare, to verify, to calculate, to describe, to establish.
Avoid the use of vague non-action verbs such as:
To appreciate, to understand, to study.
42
Examples of how to state an objective
Using the previous example on the utilization of Child Welfare Clinics we may want to:
compare
the level of utilization of the Child Welfare Clinic services amongst various socio
economic groups.
verify
the extent to which increasing distance from the health facility reduces the level of
utilization of the Child Welfare Clinic services.
establish
the pattern of utilization of Child Welfare Clinic services by season
describe
mothers perception of the quality of services provided during the Child Welfare
Clinics.
Open Research Questions versus Hypotheses
Based on your experience with the study problem, it might be possible to have an explanation for the problem
or predict causes which can then be tested. One can in this case formulate a research question in the form
of a hypothesis.
• is an explanation or a prediction of why one or more factors
A u
hypothesis
influence other factors.
___________
In our example of low utilization of Child Welfare Clinics it would be possible to formulate the following
hypotheses:
i.
Utilization of child welfare clinics is lowest in the rainy season due to the high workload of mothers
in that season.
2.
Utilization of child welfare clinics is lowest in those clinics in which staff is poorly motivated to provide
preventive' services.
Note:
Policy makers and field staff usually feel the need for research because they do not have
enough insight in the causes of a certain problem. In that case, it is better to formulate open
research questions, and avoid cutting down the study to a limited number of hypotheses.
43
In most HSR proposals, therefore, one would rather formulate the specific objectives in the form of open
statements as given in the examples on page 32.
GROUPWORK
1.
(1-1/2 hours)
Prepare general and specific research objectives for the research proposal you
are developing. After formulating your objectives ask yourself the following
questions:
Do the objectives deal with all parts of the research problem in a logical
and coherent way?
Are the objectives clearly phrased?
Are the objectives
defined in operational terms that can be measured?
realistic.
Do the objectives indicate that recommendationswill be made for pos
sible implementation of the research results?
2.
Prepare a flipchart with your objectives for use in exercise 2 and in the plenary
discussion.
EXERCISE 2:
Formulation of research objectives
(1/2 hour)
Assess the research objectives formulated by another team using the criteria mentioned
above.
44
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
literature and
other available
information
research
questions or
hypotheses
variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
type of study
data collection
techniques
sampling
plan for data
collection
plan for data
analysis and •
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
ethical
considerations
pilot study
or pre-test
manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
P- material support,
and equipment
L money
summary of proposal
Module 7 -
INTRODUCTION TO HEALTH SYSTEMS
RESEARCH METHODOLOGY
OBJECTIVES
At the end of the session you should be able to:
1.
Identify the pertinent questions to consider when developing a research design
for your project
2.
Describe the components of a research design that should be dealt with in the
methodology section of your research proposal.
In the previous modules you have selected a research topic; made a brief description of the problem and
its importance; carried out a literature and information review; set out objectives precisely stating the
purpose of the study; and you have explained how the results will be used.
Now you must decide exactly how you are going to achieve your stated objectives. The following questions
will be asked to introduce each stage of research design in the modules to come.
WHO/AFRO/SHDS
Joint HSR Project
46
HEALTH SYSTEMS RESEARCH METHODOLOGY
Components of Research
Design:
Questions you should
ask:
1.
What do we want to
measure?
Selection of variables
2.
How can we measure it?
Selection of type of
study
3.
What tools do we need
to measure it?
Selection of data
collection techniques
4.
Where should we measure it?
Sampling
How many subjects do we
include in the measurement
and how do we select them?
V
5.
Plan for data
collection.
How do we collect the data ?
I
6.
Plan for data analysis
and interpretation
What will we do with the
collected data?
t
7.
Are there any ethical
considerations related to
the implementation of the
study?
8.
How can we check whether
our methods for data
collection are correct
before beginning a study?
Ethical considerations
in health systems
research
Pre-testing
Note that steps to be taken according to this schedule are inter-related. After completing a step it is
useful to review previous steps to ensure consistency in your proposal.
47
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
formulation of
objectives
V
research
methodology
Important elements
of each step
- literature and
other available
L information
r— research
questions or
- hypotheses
J variables
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
r— manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
p. material support,
and equipment
*- money
summary of proposal
Module 8-VARIABLES
OBJECTIVES
At the end of the session you should be able to:
1.
Define what variables are and describe why their selection is important in research
2.
Distinguish the differences between dependent and independent variables and
indicate how they are used in research designs
3.
Identify the variables that will be measured in the research project you are
designing
WHO/AFRO/SHDS
Joint HSR Project
50
WHY is selection of variables important?
Look again at the diagram in module 4 describing the factors that might influence malnutrition.
In our research project we may want to determine to what extent these factors actually influence or cause
the problem.
Therefore we may try to express both the problem and the factors influencing it in so called variables or
characteristics which are measurable. For example, we may express malnutrition in the
variables “wight for height” and “weight gain”. Illness can be expressed in the variable “number of illness
episodes during one year, to be assessed at two-week intervals”. If children who are ill most frequently show
less weight gain and have a lower weight for height after one year than children who have not been ill, we
might assume that illness contributes to malnutrition.
Determining causes of a problem is important, because if we can remove or alter a cause we may be able
to reduce the problem.
What is a variable?
ESH
A variable is a characteristic of a person, object or phenomenon, which is
measurable and thus can take on different values.
__________________________
_________________________________________
Variables can be:
Quantitative and measured numerically (for example, age, height, weight, blood pressure, etc.)
or
Qualitative, and measured in terms of categories.
Example:
Variable
Categories
Sex
male, female
Outcome of
disease
recovery, chronic
illness or death.
Mode of transport
to come to clinic
foot, bicycle,
motor cycle, bus,
car, other
51
Think of a few examples of quantitative and qualitative variables that you will measure in your own research
project. For each variable that you select, the method of measurement should be clearly indicated. For
quantitative variables the unit of measurement (for example centimetres or kilograms) should be stated.
What types of variables are important in research and why?
Where a researcher looks for a causal explanation, the variables that are used to
measure the problem under study may be called the dependent variables.
The variables that are used to measure the factors that are assumed to cause or to
influence the problem may be called indepondont variablOS.
Often, these independent variables are mutually related. Therefore, some researchers
prefer to call them differently: contributing factors to the problem, or risk factors, or
determinants.
__________________
______________________ ——-----------—
A variable may be dependent or independent according to the objectives of the study. It is therefore
important in the context of a particular study to be clear which are the dependent and independent variables.
The decision on which variables are dependent and which are independent follows from the statement of
the problem.
example:
If a researcher investigates whether smoking causes lung cancer, lung cancer is the dependent
variable and smoking the independent variable. If a researcher investigates why people smoke,
smoking would be the dependent variable, and pressure of peers to start smoking could be an
independent variable.
Whereas it may be easy to show an association between two variables, demonstration of a causal
relationship is much more difficult.
A variable that is associated with the problem AND with a possible cause of the
problem is a potential confounding variable.
A confounding variable may either strengthen or weaken the apparent relationship
between the problem and a possiblecause.
___________________
52
For instance, an association is shown between breastfeeding of the child and improved nutritional status
(measured by weight/height). However, age of the child is strongly associated with the way of feeding:
children of 18 months old are less likely to be breastfed than those of less than 6 months old. But age is also
strongly associated with nutritional status: children of 18 months old are more likely to be malnourished than
those of less than 6 months old for a variety of reasons, including the more common occurrence of infectious
diseases in that age group. These associations are presented in the following figure:
Age
Nutritional Status
(weight for height)
Breastfeeding
Because age is associated with both breastfeeding and nutritional status it is a potential confounding
variable. A study aiming to show that breastfeeding is associated with good nutritional status, should
measure age as well. Age should be taken into account in the analysis, for instance by subdividing the study
population into eight age groups of three months.
Certain variables such as age, sex, socio-economic status, marital status and religion, are related to many
other variables and thus often need to be measured as potential confounding variables. Because they
appear in almost every study they are called BACKGROUND VARIABLES. However, you have to make
sure for each particular study, whether you really need the information and why, before you include them.
HOW should variables for a particular study be selected?
To avoid unnecessary data collection and to ensure that all data relevant to the objectives are collected,
variables must be carefully identified to be specific to the research problem. A common problem in research
projects is the failure to collect information about key variables and the collection of too much information
about others.
Note:
The number of variables should be as many as are necessary to cover all objectives. It may
be easiest to derive your variables from the diagram of factors influencing your problem,
as presented in the statement of the problem.
53
An example of how these factors can be measured by using variables is presented in the table below.
PROBLEM
DEPENDENT VARIABLES
Malnutrition
Weight/Age, Weight/Height
Upper Arm Circumference
FACTORS
INDEPENDENT VARIABLES
Illness
Diarrhoea during past two weeks
Other illnesses during past two weeks
Number of meals per day
Type of food in each meal
Actual amount of food consumed
past 24 hours.
Poor Food Intake
Module 8
EXERCISES:
Identification of Variables in Research
(To be carried out in plenary, 1/2 hour)
Look at the following descriptions of research problems and then answer the questions that follow.
Problem 1.
In developing countries, the highest rate of mortality occurs in infants and children, mostly from
preventable diseases. Many health care programmes and research activities have therefore been
directed towards children and pregnant women. A health researcher, however, believes that
anaemia, malaria and malnutrition are serious problems among adult males, particularly farmers.
He therefore wishes to study the prevalence of these diseases among adult males of various ages,
occupations and educational backgrounds in order to determine how serious a problem these
diseases are for this population.
Q:
What are the dependent and independent variables in the study? Which of these
are qualitative and which are quantitative variables?
54
Problem 2
A District Medical Officer receives a complaint from the community that Village Health Workers are
often out of drugs. In preliminary investigations this shortage of drugs is confirmed. Village Health
Workers get their drugs at monthly meetings at the Health Centre. The DMO decides to investigate
why the supply of drugs to VHWs is unsatisfactory.
Q:
What are the dependent and independent variables for the study?
Problem 3
Occasionally, a research project is carried out without considering several of the important variables.
This may result in deceptive findings or an unclear relationship.
Ina study concerning the pattern of distribution of schistosomiasis in the adult population of a village
community, a researcher found that the adults were predominantly farmers and overall 20% of them
had schistosomiasis. The researcher believed that the prevalence of the disease was moderately
low in the adult population.
Q:
Are there any variables whose inclusion in the study might have shown that the
prevalence of the disease varied greatly among different categories of adults in the
village?
GROUPWORK
1.
(1-3/4 hours)
Identify variables that will be measured in your research proposal. Include the
following:
What are your dependent variables?
What are your independent variables?
Are all your variables measurable?
Show the relationship between variables in a diagram. This diagram will
be an adaptation of the one prepared under the statement of the problem.
2.
Prepare a flipchart of the diagram of variables for presentation in plenary.
55
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
*- information
P research
questions or
- hypotheses
------------ —
i— variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
-| type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
’’
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
7
1
summary of proposal
- administration,
monitoring and
- evaluation
P- material support,
and equipment
*- money
1=
L
Module 9-STUDY TYPE
OBJECTIVES
At the end of the session, you should be able to:
1.
Describe the types of study most often used in health systems research as well
as how and when to use them
2.
Describe common problemsthat arise in usingthe varioustypes of study in typical
field situations and how they might be solved
3.
Describe what types of study are most appropriate for various research problems
commonly dealt with through health systems research
4.
Identify the most appropriate types of study to use in the research design for the
proposal you are developing during the course.
WHO/AFRO/SHDS
Joint HSR Project
58
Overview of Study Types
There are essentially three types of study:
Descriptive study
Analytical study
Experimental study
1.
Descriptive Study:
A descriptive Study involves the systematic collection and presentation of
data to give a clear picture of a particular situation.
Examples:
Community diagnosis (baseline survey)
Nutritional status survey
Determination of the proportion of vaccinated children in a population
3
In descriptive studies two main types of research can be distinguished:
Descriptive
Study
Quantitative
Research
Qualitative
Research
59
Qualitative Research
Qualitative research involves exploration of a number of often mutually related
variables that give insight to the nature of a certain problem or situation.
Such studies may be concerned with opinions, perceptions and attitudes to a topic (for example, family
planning). They may also describe changes over time (for example, in nutritional habits). Or they may
attempt to identify various possible factors associated with a particular problem, about which little is known.
For example, a study may try to identify the major factors contributing to the increasing gap between rich
and poor in a certain society, and describe how these factors reinforce each other.
Note:
A qualitative study will usually concentrate on a small study population (see sampling),
in order to remain flexible in the number of variables one would wish to include, and to
permit in-depth description of these variables.
Quantitative Research
Quantitative research aims at quantifying the distribution of certain vari
ables among a study population.
A quantitative study usually includes (quantitative as well as qualitative) variables that are easily measurable.
Ina community diagnosis, this may includethe counting of water sourcesand latrines, the measuring of the
size of plots, or establishing the prevalence of certain diseases. It may, however, also include qualitative
variables which are less easily measurable, provided they have been carefully categorized in a preceding
small scale study.
Note:
Quantitative studies usually cover a relatively large study population (see module 11,
sampling). It may be advisable to restrict the number of variables, in order to avoid
problems with analysis.
60
Qualitative and quantitative research can fruitfully be combined in almost any descriptive study.
Often this occurs in one single questionnaire. However, if hundreds of respondents are required to quantify
certain variables it is not wise to include qualitative research in the same questionnaire. This would make
the researchers “drown” in data which are more difficult to analyze and which are not needed in these
quantities. In that case it is more appropriate to take a subsample and conduct the qualitative research
separately.
Example:
One may want to conduct a survey on the use of various family planning methods in a district. To
describe what percentage of the population uses what method one may interview say 1000
respondents.
To find out why some people prefer certain methods or do not use any method at all we could take
small samples of say 20 non-users and 20 users of various methods. Each group would be inter
viewed to find out reasons for using or not using particular methods.
2.
Analytical Study
An analytical study triesto establish causes or riskfactors for certain problems
in a more rigorous way than in a descriptive study. This is done by comparing two
or more groups on the variables one is interested in.
Two commonly used types of analytical studies will be discussed here:
Analytical
Study
Cohort Study
Case-Control
Study
61
Case-Control Study
In a case-control study the investigator compares one group of informants
among which the problem that he wishes to investigate is present (e.g. malnutrition)
and another group, called a control or comparison group, where the problem is
absent, in order to find out what factors have contributed to the problem.
•
•
.«
.•
••
•'
•
••
•
*
I -— . . A >» * A ? *
\
Cases and controls may be matched on certain confounding variables such as age and sex, but not on the
variables for which one wants to determine to which extent they contribute to the problem.
Ina study of malnutrition, for example, you may select children as controls (those without malnutrition) who
are of the same age as the cases (with malnutrition), if it is already known that malnutrition tends to be
concentrated in a certain age group. However, you will not match the children on socio-economic and
cultural variables, or on the presence or absence of communicable disease (see diagram in Module 4), if
you are looking for risk factors among these variables.
If in the study of malnutrition it is found that the father is a migrant labourer who does not contribute much
to the household for 65% of the malnourished children (cases) and for 15% of well-nourished children
(controls), this is a risk factor.
a
Cohort Study
<
In a cohort Study, a group of individuals that are exposed to a risk factor (study
group) are compared to a group of individuals not exposed to the risk factor (control
group). The researcher follows overtime in both groups whether and when diseases
or problems appear that she or he expects to be related to the risk factor, and
compares the occurrence of these problems.
A well known example of a cohort study is that carried out among smokers and non-smokers to determine
the importance of smoking as a risk factor for developing lung-cancer.
The control group should be selected at the same time as the study group, and both should be followed up
with the same intensity.
You may use either of the types (case-control or cohort) to study possible causes of a problem.
For example, if you assume a causal relationship between the use of a certain water source and the inci
dence of diarrhoea among children under five in a village with different water sources:
62
You can select a group of children under five years and check at regular intervals (two weeks) how
often the children develop diarrhoea, and how serious this diarrhoea is. Children using the sus
pected source will be compared with those using other sources of water supply with regard to the
incidence of diarrhoea (cohort study);
You can also conduct a case-control study. In that case you may compare the water source used
by children who present themselves at a health centre with diarrhoea (cases) with the water supply
used by children presenting themselves with other complaints of roughly the same severity, for
example acute respiratory infections (controls).
However, for financial and practical reasons, a case-control study would usually be preferred to a cohort
study.
Case-control studies are relatively quick, and cheap to undertake. The major problem with a case-control
study is the selection of an appropriate control group.
Cohort studies take a long time and are labour intensive, and therefore expensive. The major problems
are usually related to the identification of all cases in a study population, and to the inability to follow up all
personsincluded in the study because of population movement.
3.
Experimental Study (optional)
An experimental study is the strongest possible type of study to prove causation.
■
In an experimental study, individuals are randomly allocated to at least two
groups. One group is subject to an intervention, or experiment, while the other group
is not.
■
■■
The outcome of the intervention (effect of the intervention on the dependent variable/
problem) is obtained by comparing the two groups.
__________________________
An example of an experimental study would be a drug trial or vaccine trial (See Annex 9.1). In practice it is
often difficult to fulfil all conditions of an experimental study. If individuals are not randomly allocated to an
intervention or control group, the study may be called “quasi-experimental”. Such a study will, however,
always experiment with an intervention (for an example, see Annex 9.1).
63
Summary of Types of Study
Descriptive
Studies
Analytical
Studies
Experimental
Studies
■
■
■ case-control
■ cohort
■ experimental
■ quasiexperimental
qualitative
quantitative
Note:
For health systems research, we will most often conduct descriptive studies (small
scale, qualitative studies as well as larger scale quantitative surveys) and case-control
studies, or a combination of these study types.
Most research requires a combination of study types. From descriptive studies we may extract certain
factors that we would like to test for their causal relationships using an analytical study or even an
experimental study.
Research using a limited number of variables, however, tends to lead to further questions about the context
of those variables, so that the researcher is obliged to go back to the field and undertake another descriptive
study.
The following diagram offers an example of how different study types might be combined in the field of EPI.
64
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
formulation of
objectives
Important elements
of each step
- literature and
other available
L information
r— research
questions or
- hypotheses
r
’’
research
methodology
r- variables
type of study
data collection
techniques
- sampling
plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
L. pilot study
or pre-test
•— manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
*- evaluation
j- material support,
and equipment
*- money
r
summary of proposal
Descriptive Study (Qualitative)
During a community diagnosis the District HealthTeam finds that
people complain that children get measles, even though they are
vaccinated. This finding is confirmed by interviewing health staff
at the MCH clinic.
Analytical Study (Case-control)
The team compares the vaccination status of children admitted
to hospital with measles (cases) with the vaccination status of
those admitted for other diagnoses (controls). From this com
parison it is estimated that vaccine efficacy is only 50%.
Descriptive Study
Further investigation shows that the cold chain is poorly main
tained.
Corrective action is taken
Module 10 - DATA COLLECTION TECHNIQUES
OBJECTIVES
At the end of this session you should be able to:
1.
Distinguish different kinds of data collection techniques, and state their advan
tages and disadvantages.
2.
State how and when to use a certain technique to deal with a given research
problem.
3.
State the benefits of using a combination of different data collection techniques.
4.
Distinguish various stages in questionnaire design.
5.
Demonstrate appropriate techniques of wording and lay-out of a questionnaire
to ensure maximal quality of responses.
6.
State various sources of bias in data collection, and ways to prevent bias.
7.
Identify appropriate data collection techniques for your study.
8.
Prepare your data collection tools, taking care that you cover all important
variables.
WHO/AFRO/SHDS
Joint HSR Project
68
A.
OVERVIEW OF DATA COLLECTION TECHNIQUES
Data collection techniques enable us to systematically collect information about our objects of study
(people, objects, phenomena) and about the setting in which they occur.
In the collection of data we have to be systematic. If data are collected haphazardly, it will be difficult to
answer our research questions in a conclusive way.
Example:
During a nutrition survey three different weighing scales were used in three villages. The researchers
did not record which scales were used in which village. After completion of the survey it was
discovered that the scales were not standardized and indicated different weights when weighing
the same child. It was therefore impossible to conclude in which village malnutrition was most
prevalent.
Various data collection techniques can be used:
Observation
Interview (face to face)
Administering of written questionnaires
1.
Observation
•-
Observation is a technique which involves systematically selecting,watching
and recording behaviour and characteristics of living beings, objects or phenomena.
Here the eye is the most important research tool.
The observation of human behaviour is a much used data collection technique. It can be done in different
ways:
Participant observation (the observer takes part in the situation he observes)
Non-participant observation (the observer watches the situation, openly or concealed, but does
not participate)
69
Observations may serve different purposes. They may be the principle source of information (for example:
systematic observations on the motoric development of infants: when do they start sitting, creeping,
walking). Or they may complete and correct information obtained from interviews about behaviour of
people. For example, they will give valuable additional information with respect to sensitive issues such as
alcohol or drug use, or community interaction with patients who suffer from feared diseases (leprosy, TB,
AIDS).
Observations of human behaviour can form part of any type of study, but as it is difficult to standardizethem,
they are most often used in smaller scale studies.
Observations can also be made on objects. For example presence or absence in the household of a latrine
and its state of cleanliness may be observed.
If observations are made against a defined scale they may be called measurements.
Measurements usually require additional tools. For example, in nutritional surveillance we measure weight
and height by using weighing scales and measuring board. We use thermometers for measuring the body
temperature.
2.
Interview
An interview is a data collection technique that involves oral questioning of re
spondents, either individually or as a group.
■
Answers to the questions posed during an interview can be recorded by writing, or by tape recording
The answers can be recorded either during an interview itself, or
after the interview (soon after the interview, if possible, so that responses are not forgotten).
Interviews can be conducted in different ways: i.e. with flexibility or with a fixed list of questions as
described below.
Interviewing with flexibility is done:
From an interview schedule or checklist, to ensure that all issues are discussed, but with
flexibility concerning timing and the order in which the questionsare asked. The interviewer
may ask additional questions on the spot, in order to gain as much useful information as
possible.
From a prepared questionnaire,but mainly with open questions. (For explanation of open
questions, see Section B, Questionnaire Design).
70
The unstructured or loosely structured way of asking questions can be used for interviewing individuals as
well as groups of key informants (see annex 10.1 on focussed group discussions).
Note:
This method of interviewing is useful in small scale qualitative studies in which the
researcher still has to explore various aspects of the problem.
Interviewing with a fixed list of questions is done:
In a standard sequence, where questions usually have fixed or precoded answers (see
Section B, Questionnaire Design)
Note:
This technique is useful when the researcher is relatively knowledgeable about expected
answers and when the population being interviewed is relatively large.
3.
Administering a Written Questionnaire
A written questionnaire (also referred to as self-administered questionnaire)
is a data collection tool in which written questions are presented that are to be
answered in written form by the respondents themselves.
A written questionnaire can be administered in different ways, for example by:
Sending questionnaires with clear instructions on how to answer the questions by mail, and asking
for mailed responses
Gathering all or part of the respondents in one place at one time, giving oral or written instructions,
and letting the respondents fill out the questionnaires
Hand-delivering questionnaires to respondents and collecting them later.
The questions can either be phrased in a flexible and open way or precoded, as in the interview.
71
4.
Systematic Analysis of Available Data
All data collection techniques discussed above refer to the collection of new data. However, there usually
is a mass of information that has already been collected by others but has not been analyzed. This
information can be fruitfully used for identification or analysis of certain research problems.
Analysis of the information routinely collected by health facilities can be very useful,for example, to identify
problems in certain interventions or in flows of drug supply, or to identify increases in incidence of certain
diseases. Sometimes the factors contributing to the problem may also be identified from the same source;
sometimes additional research will be necessary to solve the problem.
Analysis of census data, unpublished reports, publications in archives and libraries or offices at the various
levels of health and health-related services, may also be a study in itself.
The advantage of using existing data is that it is inexpensive. However, it is sometimes difficult to get access
to the records or reports required, and the data may not always be reliable.
Differentiation between Data Collection Techniques and Data Collection
Tools
In order to avoid confusion in the use of terms, the following table points out the distinction between
techniques and tools applied in data collection.
Data Collection Techniques
Data Collection Tools
Observing
eyes; pen/paper; watch;
scales; microscope,etc.
Interviewing
interview schedule;
questionnaire; tape recorder
Administering a written
questionnaire
questionnaire
72
Advantages and Disadvantages of Various Data Collection Techniques
The following table summarizes the advantages and disadvantages of various data collection techniques.
TECHNIQUE
ADVANTAGES
DISADVANTAGES
Observation
(in particular
of behaviour)
gives more detailed
information
observer bias
(researcher may only
notice what interests him)
permits collection
of information on
facts not mentioned
in the questionnaire
Interview
presence of researcher
can influence the situation
he observes
permits tests on
reliability of
responses to
questionnaires
future plans cannot be
observed
suitable for
illiterates
presence of researcher
can influence responses
of respondents
clarification of
questions is possible
higher response rate
than written
questionnaires
Small scale
flexible
interview
records of events may be
less complete than with
observations
observer bias, if
information not
scheduled
extra information
can be gathered
through observation
more difficult to analyze
in a standardized way
Larger scale
fixed interview
easy to analyze
easy to miss important
information
Written
Questionnaire
(self
administered)
no need for research
assistants to
administer it
not possible to use with
illiterate respondents
low rate of response
no bias in asking
the questions
differently
questions may be
misunderstood
anonymity: more
honest responses
73
Importance of combining Data Collection Techniques
When discussing different data collection techniques, and their advantages and disadvantages, it becomes
clear they complement each other. Combining different techniques maximizes the quality of the data
collected and reduces the chance for bias (see below).
Wherever possible, observations should be added to information obtained by interviews and question
naires, preferably collected in a systematic way (spread out in place and time, and well recorded).
Note:
Large scale surveys are better suited for measuring facts, small-scale interviews and ob
servations for obtaining the information that places these facts in their context.
During supervisory visits, we already combine different methods: we observe; we make calculations by
using the information system; we pose a number of routine questions, and we have informal discussions
with staff and patients. Also in daily life we inform ourselves in various ways before taking important
decisions. Research is a more sophisticated form of data collection, but the principle remains the same.
74
Modules 9, 10
EXERCISE 4:
Selection of Study Type and Development of Data
Collection Techniques
In all the health management problems below, you are asked to state what type(s) of study and data
collection techniques would be most appropriate and why. Read each problem thoroughly and
answer the questions asked.
1.
You noticed a number of women with goitre in your district and you are concerned that this may
be a problem. What type of study would you do to find out the size of the problem?
How would you find out whether the population perceives goitre as a problem? How would you
identify the most important risk factors, and possible interventions which would be acceptable for
the population?
2.
A District Health Team has evaluated its malaria spraying programme and did not find significant
flaws in the functioning of the services in different divisions and villages. Nevertheless, the incidence
of malaria and mosquito counts show peaks in certain villages which are most likely related to
differences in quality of the services. How would you try to find out if there is something wrong with
the services?
3.
You are a mid-wife in charge of a maternity unit in a district hospital. You suspect that the number
of low-birthweight babies is increasing, and you would like to know more about the condition of the
mothers to see if remedial actions could be taken. The clinic records are not complete enough to
draw conclusions and you have neither the time nor the money to do a large community survey.
What type of study and data collection techniques would you use?
4.
You have recently been appointed to be the District Nursing Officer in a remote, previously unders
erved district. One of your tasks is to develop a district health plan. What information would you
like to collect, and how?
5.
There are long queues at the out-patient department of your District Hospital. You are concerned
about this and you think that the problem may be related to the organization and management of
the department.
Q.
What type of study would you do to identify the causes of the problem?
What technique would you use to collect your data?
6.
You are in charge of a large Health Centre in a rural area. You feel that there is a problem in the
TB services: few patients seem to report themselves, and the regularity in attendance of registered
cases is poor.
Q.
What type(s) of study would you do to find out why TB patients do not attend the
TB clinic?
What data collection technique(s) would you use?
75
B.
QUESTIONNAIRE DESIGN
The questionnaire in its different forms is probably the most commonly used instrument for collecting and
recording information. The specific form and content of questionnaires will vary from study to study.
However, each researcher designing a questionnaire will have to keep in mind:
What exactly do I want to know?
■
Who are the people that I am going to ask the questions?
What is their level of understanding of the topic?
Are they mainly literate or illiterate?
■
How am I going to process the data, by hand or by computer?
Steps in Designing a Questionnaire
1.
Take the variables as your starting point, first your dependent variables that define the problem,
then the independent variables. Phrase questions where required (for some variables you may use
observations, measurements, or existing data).
2.
Phrase clear, short questions, using every day words (never scientific language).
Be aware that your research questions usually can not be answered directly by informants, and
often have to be broken down in several questions.
Example:
In an investigation for reasons why many nurses quit the profession, it is assumed that for
unmarried nurses it will be easier to search for other employment than for married nurses.
The question
“Did marital status influence your decision to quit nursing’’, however, may not
make sense to the informant. Instead you may ask:
“Were you married (or living together) when you quit the profession?’’
A number of questions could then follow to clarify under what conditions married women
managed to stop working or change jobs.
76
3.
Check whether each question measures one thing at a time.
Example:
The question “Have you used ORS when your child had diarrhoea the last time and were
you satisfied with the result” should be split up in a number of questions:
When did your child have diarrhoea for the last time?
What did you do?
(If use of ORS is mentioned)
How did you prepare it?
How often did you give it to the child
Were you satisfied with the result?
(Explain why, or why not)
4.
Check that there are no leading questions
Example:
The question "Did you use ORS” makes it easy for the mother to say yes even if she did
not use it.
If the open question “What did you do” reveals she used ORS, the answer has more value.
5.
Check whether the questionnaire has been designed “consumer friendly”. It should prefera
bly start with an open question that raises the interest of the informant. Link background variables
such as age, marital status, religion, and economic status to other questions so that the information
makes sense for the informant, or ask them towards the end of the questionnaire. Put more delicate
questions towards the end.
6.
Formulate “control” questions to cross check responses on earlier sensitive questions, for
example about income, or to further specify questions that are likely to yield inexact answers (e.g.
on age).
7.
Make the questionnaire as short as possible. If the nature of the topic requires a longer ques
tionnaire (more than one hour): conduct the interview in two parts.
8.
Make sure that every question in the questionnaire will ultimately be used. A plan of how to
analyze the data should be prepared in advance, including the preparation of dummy tables.
Dummy tables are used to make clear what variables you want to relate to each other to check for
77
possible causal relationships after data collection has been completed. This is another way of
making sure that every question in the questionnaire serves a purpose, and that no important data
are missing.
9.
Make sure that all the interviewers or respondents have the same understanding of questions.
To ensure standardization it is useful to prepare special instructionsforthe interviewers or, in case
of written questionnaires, for the respondents. Instruct the interviewers when to “probe” for answers
for an open question, and how to do this in a systematic way.
10.
Pre-test the questionnaire to ensure that the format, the wording and sequencing is correct. If it
is impossible to obtain reliable answers on a certain question, even after rephrasing, throw it out.
11.
Use the pre-test to establish whether it is necessary to translate the questionnaire in one or
more local languages. In case it is necessary to translate the questionnaire, make the translation
and have it retranslated to the original language. You can then compare the two versions and make
a decision on the wording of the final version.
Types of Questions Used in the Questionnaire
Basically, two types of questions may be used in questionnaires:
Open questions
Closed questions
1.
Open Questions
H
efnnc
• ♦
list of answers from which to choose.
S^ponses:
resp
.•
■
pr
d h
th
________________________________________________________________________________
Such questions are useful to test knowledge (Q 1) or ask for opinions (Q 2):
01:
What are the activities of the Village Health Committee?
02:
What do you think are the reasons for a high drop-out rate of Village Health Committee
members?
78
Advantages of open questions
Open questions allow the respondent to “tell it like it is”in his/her own words. Such information can
be very useful to make the final report more lively.
Open questions make it possible for the respondent to bring up issues/ideas not previously thought
of in planning the study, which may provide valuable new insight in the problem.
As the information comes spontaneously, the answer has more value than when it would have been
suggested as one of the options from which the informant could choose (see point 4 in question
naire design on page 61).
Disadvantages of open questions
Makes heavy demands on the interviewer. Interviewers have to be skilled in note taking, in distin
guishing what information is relevant for the topic under study, and in steering the discussion back
to the remaining questions. Otherwise the process could be very long and tiring to both the
respondent and interviewer, and much valuable information will be lost.
The respondent does not benefit from a list of possible responses which could serve as a reminder,
hence a tendency for the respondents to say they “Don’t Know” or to forget important aspects.
Analysis of an open question is time consuming. Responses have to be categorized after the field
work is completed.
Possible Solutions to these Problems:
Thoroughly train the interviewers, or select skilled people for this type of research.
Prepare a list of further questions to keep at hand to “probe” for an answer in a systematic way.
Pre-test the open questions and categorize the most common responses, leaving enough space for
other answers. Do not suggest any answer!
Closed Questions
2.
Closed questions offer a list of possible options (answers) from which the re
spondents must select at least one answer.
W.
____
79
Closed questions are useful if there are only a limited number of options:
Q3:
Have you ever gone to the local
VHW for treatment?
1.
2.
Yes
No
Q4:
What is your marital status?
1.
2.
Single
Married
/living together
separated /divorced
widowed
3.
Closed questions may also be used if one is only interested in some aspects and does not want to waste
time of the respondent and interviewer by obtaining more information than one needs
For example, a researcher who is only interested in the protein content of a family diet may ask:
Q5:
Did you yesterday eat any:
peas, bean, lentils
Yes
No
fish or meat
Yes O
No
eggs
Yes
No
milk or cheese
Yes
No
Closed questions may be used as well to make respondents express their opinion by choosing points on
a scale:
Q6:
How useful would you say the activities of the Village Health Committee have been in the
development of this village?
1.
Extremely useful
2.
Very useful
3.
Useful
4.
Not very useful
5.
Useless
80
Closed questions should aim to:
Offer a list of options (answers) that are exhaustive, i.e. cover all possibilities
If possible, offer a list of options that are mutually exclusive, i.e. any one option chosen excludes
all others.
Keep the number of options as small as possible
Advantages of closed questions
Answers can be recorded quickly
It makes it easy to carry out analysis because the responses are already set out in a way suited for
analysis
Disadvantages of closed questions
As a list of choices is provided to the respondent, this may influence or bias his/her answer.
Respondents may select an answer which they would not have thought of otherwise, just to please
the interviewer.
Fixed responses leave little room for the respondent to think of other possible answers.
Both respondent and interviewer might soon get bored with the interview.
Possible solution to these problems
Only use closed questions on issues that are simple andXincontroversial
Test questions out as open questions before and after categorizing the answers. If you find that too
much valuable information gets lost, you may still decide to keep the question open.
Leave opportunity for additional answers by adding a category “other (specify)’’:
Q7.
How did you become a member of the Village Health Committee?
1.
Volunteered
2.
Selected at a community meeting
3.
Nominated by the health staff
4.
Other (specify)
81
Use opinion scales (such as from extremely useful to useless in Q6) not in isolation but rather as
a provocation to a next, open question: “Why?”
Notes:
A questionnaire usually hasa combination of open and closed questions arranged
in such a wdy that the discussion flows as naturally as possible.
Questions can be asked as open questions, though possible answers may to a
large extent be pre-categorized.
In case all possible answers are to be presented to the respondents to make them
choose one answer, this has to be made clear in the question itself, or in the in
struction for the interviewers. Otherwise the question may be asked in different
ways, and the answers will be incomparable.
Sometimes it can be useful to use pictures or drawings when asking questions
(see Annex 10.2, questionnaire to be used by illiterate Traditional Birth Atten
dants).
C.
BIAS IN INFORMATION COLLECTED
representative of the true situation.
..;
Possible Sources of Bias:
1.
Defective Instruments, such as:
pre-coded questionnaires utilized when not enough is known about the topic
loosely structured questionnaires without guidelines on how to ask the questions (see
questionnaire design)
weighing scales which are not standardized
This can be prevented by a careful study preparation and by pre-testing the data collection tools.
82
2.
Observer Bias:
This is a distortion which may in particular appear in observations and when utilizing loosely
structured group- or individual interviews. There is a risk that the researcher will only see or hear
things in which he is interested. These techniques should therefore preferably be applied with at
least two persons, and the information should be recorded and discussed immediately after its
collection.
3.
Effect of Observer on Informant:
This is a possible factor in all interview situations. The informant may mistrust the intention of the
researcher and dodge certain questions or give misleading answers. In small scale flexible
interviews the chance for honest answers and spontaneous illuminating remarks will be higher than
in one-time interviews with fixed questionnaires.
4.
Seasonality
For the collection of some data it is insufficient to visit an informant only once. Wet seasons in rural
areas are characterized by a high workload of the farming population, a relatively high prevalence
of certain diseases (malaria, diarrhoea), and a relative shortage of food and money, compared to
the post-harvest period.
As also the roads are better in the dry season, researchers are often concentrating their activities
in that period. As a consequence they will underestimate certain problems.
Note:
By being watchful for these biases, one can to a certain extent prevent them. If one does
not fully succeed, one should always mention honestly what the (possible) distortion of
the data collected is.
83
GROUPWORK
1.
(4hours)
Decide what type(s) of study you will apply in your own research proposal, and
what data collection techniques you will use. Make your choice on the basis of
the objectives and the variables you have decided to include in the study.
Types of study:
Descriptive (may be qualitative research, quantitative research, ora com
bination of both)
case-control
Analytic
cohort
(Quasi) Experimental
Data collection techniques:
Observations (including observing behaviour; observing/counting ob
jects; measurements such as weight, height).
Interviews (with interview schedules or
questionnaires)
individual interviews
group interviews
Analysis of records, or other existing information
When you use questionnaires, remember that you can have
Open questions (if you do not know enough about certain variables and
want the respondents to talk freely)
Closed questions (if you have a pretty good idea of what answers you
might get)
Questions with drawings or photographs
2.
Prepare your data collection tools, taking care that you cover all important
variables.
3.
Discuss the possibility of bias occurring when using the data collection tools you
have developed. Try to limit this as much as possible.
84
EXERCISE 5:
Commenting on Data Collection Tools of
Other Groups (2 hours)
Review and critically comment on the data collection tools of the other groups. Your
working group will give particlar attention to the tools of one other group to be identi
fied by your facilitators, on which you will report in plenary.
85
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
I
formulation of
objectives
’’
research
methodology
Important elements
of each step
- literature and
other available
L information
r— research
questions or
- hypotheses
i- variables
type of study
data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
I
1
summary of proposal
- administration,
monitoring and
- evaluation
p- material support,
and equipment
J- money
Module 11 - SAMPLING
OBJECTIVES
At the end of the session you should be able to:
1.
Identify and define the population to be studied
2.
Identify and describe common methods of sampling
3.
Critically discuss problems of bias that should be avoided when selecting a
sample
4.
List the factors to consider when deciding on sample size
5.
Decide on the sampling method and sample size most appropriate for the
research design you are developing
WHO/AFRO/SHDS
Joint HSR Project
88
A.
INTRODUCTION
WHY sampling?
When conducting research it is usually impossible to study the entire population. Constraints in both time
and money force us to study only a fraction, or a sample of the population we want to study.
As we would like to draw conclusions about the population, it is necessary to take care that the sample is
fully representative of that population.
sjlfililiii• ■ feswt ilt|ilttii
<
?■ - ■ .• <
A representative sample should possess all the important characteristics
of the population from which it is drawn.
The first step in sampling is to clearly define the study population and its characteristics.
Example:
We want to investigate malnutrition in children. The study population
could be defined in this case as children under five. The child aged 0-4 years is
called the study unit.
A study population need not necessarily be people, but may consist of for instance clinics or water supply
points.
Besides identifying characteristics, we have to make a decision about the scope of the study: will it be
national, regional or local? Often, financial and logistic constraints force us to restrict the study to a limited
number of areas (study areas).
B.
SAMPLING PROCEDURES
There are various procedures that can be used to obtain a sample. The most common ones are discussed
below.
Non-Probability Sampling
Two sampling procedures are particularly useful for pilot studies on attitudes or opinions where the aim is
to get only a rough impression of how certain variables are distributed in the population: convenience
sampling and quota sampling. They are called non-probability sampling.
89
1.
Convenience Sampling
Convenience sampling is a sampling method in which the “sample” that
happens to be available at the time or period of the research is selected, for
convenience’s sake.
Many patient or clinic studies are conducted using convenience samples. For instance, all the patients
attending a particular clinic on a certain day or week might be asked to participate in the study.
A drawback of convenience sampling is that the sample may not be representative for the population you
want to study. Quota sampling is an attempt to ensure more representativeness.
2.
Quota Sampling
Quota Sampling is a method of sampling which ensures that all the known
elements in the population occur in the sample.
One may try to ensure that these occur in the same proportions as in the population. The investigator
interviews as many people in each category as he can find until he has filled his “quota”.
Quota sampling is useful when it is felt that a convenience sample would not provide the desired balance
of elements in the population.
Example:
A researcher wishes to study the attitudes of villagers towards a clinic in a village. He decides to ask
patients questions following their visit while they are still at the clinic, rather than take a random
sample from the village. This is done simply because it is more convenient. However, he suspects
that the religion of the patients interviewed will have a strong effect on their attitudes toward the clinic.
Hetherefore decides to select a quota of patients of different religious groups, in the same proportion
as they are found in the village.
Probability Sampling
If the aim is to quantify variables and to generalize the findings to the population (such as in a study that aims
to determine the malnutrition prevalence rate in a population), this can not be achieved with non-probability
sampling. In that case, probability sampling methods should be used.
90
Probability sampling employs random procedures to ensure that the sam
pling unit (i.e. the individuals, groups of people, objects, villages, etc. which are the basic
unit of the sample) is selected on the basis of chance. Every member of the population
should have a known chance of being included in the sample.
■
This method requires that a complete listing of all sampling units is available. This listing is called the
sampling frame.
1.
Simple Random Sampling
The simplest form of robability sampling is simple random sampling.
InSimple random sampling each unit in the population list (sampling
frame) has an equal chance of being selected for the sample.
________________________
To select a simple random sample you need to:
Make a numbered list of the units in the population that you want to sample;
Decide on the size of the sample (this will be discussed below);
Select the required number of sampling units, using a “lottery” method or a table of random
numbers (In Annex 11.1 it is explained how to use a table of random numbers).
Example:
A simple random sample of 100 students is to be selected from an urban school having 1200
students. Each student is given a number, and from a table of random numbers, 100 numbers are
selected. All pupils with these numbers constitute members of the sample.
Note that:
the list of 1200 students in the school is called the sampling frame
the group of 100 students selected is a simple random sample
91
2.
Systematic Sampling
In systematic sampling individuals are taken from the list (sampling frame) by
KI
taking every, say, third or tenth (or any other number) person on the list. Ideally we use
a random number to tell us where to start.
■
______
Example:
A systematic sample is to be selected from 1000 records of antenatal attendances. The total sample
is chosen to be 100. The sampling fraction is
100 (= sample size)
1
1000 (= study population)
10
The sampling interval is therefore 10.
The number of the first clinic record (i.e. sampling unit) is chosen in a random fashion, say by blindly
picking numbered pieces of paper (1-10). If 6 is picked, then every following tenth antenatal record
is in our sampling frame until we have 100 records. (The numbers on the records correspond to
6, 16, 26, 36, 46 etc).
Systematic sampling may lead to bias if the sampling interval coincides with systematic variation within the
study population. For instance if we want to select a random sample of days on which to count clinic
attendance, systematic sampling with a sampling interval of 7 days would be inappropriate as all study days
would fall on the same day of the week.
3.
Stratified Sampling
The simple random sampling described above does not ensure that the proportions of individuals with
certain characteristics in the sample will be the same as those in the target population.
If it is important that the sample includes representative sub-groups of individuals
(forexample^rbanandruralresidents;oragegroups),thenthesamplingframemust
be divided into sub-groups, or strata, for these characteristics. Random or system
atic samples of a pre-determined size will then have tobe obtained from each stratum.
This is called stratified sampling.
________________________________
Stratified sampling is only possible when it is known what proportion of the study population belongs to each
stratum.
92
An advantage of stratified sampling is that we can take a relatively larger sample from smaller sub-groups.
This allows us to get a sample which is big enough to enable us to draw valid conclusions about relatively
small sub-groups without having to collect an unnecessarily big (and hence expensive) sample of the other,
larger groups.
However, in so doing, we are using unequal sampling fractions and it is important to correct for this when
generalizing our findings to the whole population.
Example:
A survey is conducted on household water supply in a district comprising 20,000 households of
which 20% are urban and 80% rural. It is suspected that in urban areas the access to safe water
sources is much more satisfactory. In such a situation it may be decided to include 100 urban
households (1 in 40 sample) and 200 rural households (1 in 80 sample). Because we know the
sampling fraction for both strata, the access to safe water for all the district households can be
calculated.
4.
Cluster Sampling
It may be difficult or impossible to take a simple random sample of the individuals in the target population,
either because a sampling frame of individuals does not exist, or because visiting scattered individuals is
logistically difficult. However, when a list of groupings of individuals is available (e.g. villages or schools)
or can be easily compiled, a number of these groupings can be randomly selected.
The selection of groups of individuals (clusters) instead of the selection of individuals
separately is called cluster sampling.
Example:
In a study of the knowledge, attitudes and practices (KAP) of family planning in rural communities
of a region, a list was made of all the villages. Using this list, a random sample of villages was chosen,
and all the adults in the selected villages were interviewed.
5.
Multi-stage Sampling
After selection of a sample of clusters, further sampling of individuals may be carried out within each cluster.
In that case, the method may be called two-stage sampling. If within the clusters we select in the first stage
(e.g. districts) we select other clusters (e.g. villages) and within each selected village we select individuals
this would be three-stage sampling.
Sampling
procedures involving more than one stage in which units are selected are
called multi-stage sampling.
__
_
93
The main advantages of cluster and multi-stage sampling are that:
A sampling frame of individual units is not required for the whole population. Initially a sampling
frame of clusters is sufficient. Only within the clusters finally selected the individuals may need to
be enumerated.
Logistically the study will be easier to carryout than a simple random sample of similar size, because
the individuals in the sample are together in groups, instead of scattered all over.
■
Their main disadvantage is that - compared to simple random sampling - there is a larger probability that
the final sample is not representative for the total study population. The likelihood of being unrepresentative
mainly depends on the number of clusters that is selected in the first stage. Ideally this number would be
20 or more.
C.
BIAS IN SAMPLING
Bias is a systematic error leading to a distortion in the results.
_
I n Module 10 it was discussed how the use of faulty data collection tools can lead to biased results. However,
bias can also result from improper sampling procedures whereby the sample is not representative for the
study population. There may be several sources of bias.
Non respondents
It may happen that a sample is correctly identified, but that a large proportion of individuals in the selected
sample is absent or refuses to cooperate. Non-respondents from the sample may exhibit characteristics
which differ systematically from those of respondents. How do we deal with this problem? This depends
on various factors and we need to ask the following questions:
1.
How many non-respondents are there?
2.
Are non-responses due to absence or refusal of the individuals concerned?
3.
Is the problem identified during pre-testing of the methodology or only during study implementa
tion?
Depending on the answers to the above questions a number of measures to reduce bias may be considered:
94
Follow-up of non-respondents;
Change of data collection techniques to ensure better co-operation;
An extra, separate study on non-respondents, to identify to what extent respondents differ from non
respondents;
Replacement of absent non-responders. This can be justified only if their absence is very unlikely
to be related to the topic being studied.
The following are some other sources of bias:
Use of volunteers. The fact that volunteers are motivated to participate in the study may mean that
they are also different from the study population on the factors being studied. Avoid using non
random procedures which bring in human choice (consciously or unconsciously).
In studies on prevalence of disease, cases of short duration are more likely to be missed. This
may often mean missing fatal cases, cases with short episodes, and/or mild cases.
Tarmac bias. Study areas are usually selected in such a way that they are easily accessible.
However they are likely to be systematically different from inaccessible areas?
Sampling of only registered patients. Patients reporting toaclinicare likely todiffer systematically
from people treating the condition at home.
GROUP WORK (1 hour)
1.
Develop in your working group:
a definition of your study population;
appropriate sampling procedures for your study.
2.
Prepare a summary on a flipchart for use in exercise 5 and plenary discussion.
EXERCISE 6: Sampling Procedures
(1 /4 hours)
Critically examine the definition of study population and sampling procedures developed
by another group. Identify sources of bias and suggest improvements of the proposal.
95
D.
SAMPLE SIZE
Having decided on how we select our sample, we now have to determine our sample size.
It may be helpful to distinguish sampling procedures in qualitative studies from those in quantitative studies
as discussed in Module 9. Determining attitudes towards family planning would be an example of a
qualitative study. The sample size in this type of research needs to be large enough to reflect important
variations in the population, but small enough to allow for intensive study methods. Once you have defined
the categories of informants (e.g. non-users, users of pills, of injections, of condoms), you might start with
20 to 30 interviews per category. This number could be increased if the information obtained per category
is not providing a certain trend or provides conflicting information.
Quantitative studies are ideally used to quantify well defined variables and to test hypotheses. An example
might be a study to determine what proportion of the population uses various family planning methods. The
level of knowledge before embarking on this type of study is probably sufficient to make a rough estimate
(or range of estimates) of the expected outcome, and thus, to do sample size calculations using the statistical
. formulae given annex 11.2.
Notes:
The eventual sample size is usually a compromise between what is desirable and what is
feasible. However, in quantitative research it is very important to do sample size
calculations before embarking on a study, because it may not be worthwhile to do a study
at all if the feasible sample size is much smaller than the desirable sample size.
The feasible sample size is determined by the availability of resources:
time
manpower
transport
equipment
Remember that resources are not only needed to collect the data, but also to analyze them!
The desirable sample size for quantitative studies can usually be worked out, with some
assistance, if you are able to make a rough estimate of the outcome of the study, and are
clear about the main objectives of the study.
96
A descriptive study may aim to calculate a proportion (e.g. number of children 12-23 months that
are vaccinated against measles) with a given precision.
Example;
The precision required may be for instance, that, if we find 80% of children in our sample are
vaccinated, we are 95% certain that the proportion of children vaccinated in the study population
is between 70% and 90%. (It is also said that the 95% confidence interval is between 70% and 90%.)
To achieve this precision a simple random sample of approximately 100 children is needed.
If we want more precision we need a larger sample. Often, if the estimated proportion (e.g. of women having
goitre) or rate (e.g. maternal mortality rate) is small we want more precision.
Example:
Estimated
Proportion of
women with
goitre
District A
District B
1/100
1/1000
1%
0.1%
Desired 95%
confidence
interval
(precision)
Required sample
size
0.5 -1.5%
0.05-0.15%
1600
16000
This table shows that in district B more precision is desired. This increased precision can be achieved by
taking a larger sample.
(In annex 11.3 it is explained how you arrive at these sample sizes.
Ananalytic study may aim to show a significant difference between two groups. For instance, the
aim of an analytic study may be to show that malnutrition is more prevalent in children of mothers with a low
level of education than in children of mothers with a high level of education.
The sample size depends primarily on the size of the difference betweenthe two groups that are compared.
Thelargerthedifference.thesmallerthesamplethatisneededtoshow this difference. Secondly, the sample
size depends on how large we want the probability to be that we will find a significant result. The larger
the sample size is, the larger the probability of finding a significant difference.
97
Example:
Required Sample
size in each group
Mother’s Education
Country A
Prevalence rate
malnutrition in
under fives
Country B
Prevalence rate
of malnutrition
in under fives
High Level
Low Level
10%
30%
60
10%
15%
600
The table shows that in country A the difference in malnutrition prevalence rates between the two groups
is much larger than in country B. Consequently, the required sample size to show that there is a difference
is much smaller.
The sample sizes in the table ensure a 75% probability of finding a significant difference in malnutrition
prevalence rates between the two groups (at the 5% level). If, however, in a study in country B we want to
give ourselves a 90% probability of finding a significant result, the sample size has to be 900 instead of 600.
(In Annex 11.3 it is explained how you arrive at these sample sizes.)
Conclusion
When conducting a study on quantitative variables it is advisable to consult a person, who is able to calculate
the required sample size.
For sample size calculations to be made, estimates may be required of the occurrence of important
variables. In addition the main aims of the study have to be made explicit. Is the aim to measure a variable
with a certain precision? Is it to show a significant difference between two sub-groups in our study
population? It may be useful to conduct sample size calculations for each of the objectives of the study.
These calculations may reveal, for instance, that some but not all objectives are feasible. Or they may show
that some variables only need to be measured on a sub-sample.
98
GROUPWORK (1/2hour)
1.
Determine in your working group the sample size requirements to achieve each
of your objectives. The following considerations are crucial for establishing the
sample size of your study.
Is your study mainly qualitative?
■
Do you aim to determine a proportion or a quantity with a certain
precision?
Do you aim to show a significant difference between two groups?
2.
After determining sample size decide whether it is feasible to carry out the study,
and adjust your study type or objectives if necessary.
99
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
I
formulation of
objectives
Important elements
of each step
- literature and
other available
*- information
P research
questions or
- hypotheses
------------ ------
''
research
methodology
I— variables
type of study
- data collection
techniques
sampling
plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
[— manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
I
±
summary of proposal
- administration,
monitoring and
- evaluation
material support,
and equipment
*- money
Module 12 - PLAN FOR DATA COLLECTION
OBJECTIVES
At the end of this session, you should be able to:
1.
Identify and discuss the most important points to be considered when preparing
a plan for data collection
2.
Determine what resources are available and needed to carry out your study
3.
Describe typical problems that may arise during data collectionand how they may
be solved
4.
Prepare a plan for data collection for the research proposal you are developing.
WHO/AFRO/SHDS
Joint HSR Project
102
WHY should you develop a plan for data collection?
A plan is developed so that you can:
Have a clear programme of your work
Minimize errors which may result from the lack of planning (for example the population not being
fully available or data forms being misplaced)
Avoid unnecessary delays by lack of preparation
Organize both human and material resources effectively
It is likely that during the formulation of your plan for data collection, some constraints may be recognized
(such as lack of manpower) which will then require modification of the proposal. Such modification might
require a reconsideration of the sample size or extension of the period during which data will be collected.
WHAT are the main stages of the data collection process?
The data collection process can be looked at in three main parts as presented in the figure below:
Part I
Part II
Part III
Permission to
Proceed
Data
Collection
Data
Handling
Part I:
Permission to Proceed
Consent of the relevant authorities, individuals and the community in which the project is to be conducted
must be obtained. This could involve organizing meetings at district level, village level or even with a group
of households. For clinical studies this may also involve obtaining written informed consent.
Liaison must be maintained with leaders throughout the study to ensure their continued support.
103
Part II:
1.
Data Collection
WHO is to be involved in the data collection process?
Consideration should be given to those who are actually going to collect the data and their
supervisors including:
An examination of the project’s research design, to determine what types of personnel are
likely to be required at every step of the research process. Make sure to consider each task
that has to be accomplished.
A survey of what staff are available and what additional staff would be needed.
An assessment of what additional training would be required.
2.
WHEN should the research project be carried out?
The actual time that the project is carried out will be determined by the type of data to be collected
and the demands of the project. Consideration should be given to:
Availability of staff
The appropriate season or seasons to conduct the project
Accessibility and availability of the sampled population
Public holidays and vacation periods.
3.
HOW LONG should the research project take?
Consideration should be given to:
Time required to reach the study area
Time required to locate the units in the sample
Time required for the investigation once the units have been located
The frequency of visits to each unit. For some studies it may be necessary to visit units a
number of times. In certain qualitative studies additional questions may be asked of the
same informant. Allowing time for follow-up of non-respondents should also be consid
ered.
104
Note:
It is always advisable to over-estimate the duration of the data collection to allow for
unforeseen delays.
4.
Quality Control to prevent Bias
Quality of data from questionnaires can be improved through:
checksbytheinterviewertoseewhetherthequestionnaireisfilled incompletely i.e. before
the interview is terminated.
checks by the supervisor on the day of the interview to see that the questionnaire is filled
in completely and that the recorded information makes sense. Problems should be
discussed with the interviewer.
checks by researchers during data analysis to ensure that data are complete and
consistent
Insufficient quality control during the data collection process may result in biases in the data, leading to false .
or misleading conclusions. Most of these sources of bias have been discussed in previous modules.
Supervision in the field should try in particular to prevent the following biases:
Deviations from the sampling procedures set out in the proposal
Variations in interviewer style and research procedures due to deficiencies in training,
poor selection procedures, stress or overwork
Consistency amongst various members of the research team should be ensured. This
includes the adherence to standard procedures (e.g. all children may have to be undressed
before weighing them)
Biases in assessment by observers because they make a mistake in favour of expected
results when recording observations or answers to open questions
Differences in measurement due to different persons making measurements
It may be useful to have measurements repeated on some individuals by different members
of the research team to check for variation between researchers.
The use of different, unstandardized, items of equipment
If measuring equipment is used (e.g. weighing scales), a rule must be made to regularly
check that they perform to standard. For example, make it a rule that before each weighing
session the scales are set to zero and that once a week each scale will be tested using a
standard weight of, say, 10 kg.
105
Changing criteria for measurement or for categorizing answers part way through the
study;
Measurements or responses which change when the subject is under investigation
(for example, blood pressure and pulse may increase when the subject is apprehensive;
a respondent may dodge certain sensitive questions).
5.
Pretest
In order to estimate the time required for the data collection, the number of staff required and the
appropriateness of data collection techniques, it may be necessary to carry out a pre-test of the
methodology (See Module 17).
Part III:
Data Handling
As the data are collected, a clear procedure should be identified to process them.
1.
First, it is necessary to check that the data gathered are complete and accurate (see control above).
2.
At some stage questionnaires will have to be numbered. Decide if this should be done at the time
of interview or at the time they are put in storage.
3.
Identify the person responsible for the storing of data and where it will be stored.
4.
Decide data should be stored. Storage should be organized so that record forms are logically
placed in a numbered or data sequence.
5.
Decide whether the data will be processed or tabulated in the field or whether they are to be
transported back to the main base for this exercise.
f
106
GROUPWORK
(1-1/2hours)
Divide into your working groups, and prepare the plan for data collection for your research
proposal, keeping in mind the following points:
Permission to proceed
Data collection
Who will do what?
When will it be done?
Time required?
Quality control: How and by whom?
Where can biases be expected?
Pretest on what?
Data handling
How will accuracy be checked?
When and how will questionnaires be numbered?
How will data be stored?
Who is responsible for data storage?
Where will data processing be carried out?
I ' ‘
107
Module 13 - PLAN FOR DATA ANALYSIS AND
INTERPRETATION
EXERCISE 7 : Smoking (Questionnaire)
Questionnaire Number
Please answer all questions below one by one after reading them carefully.
If boxes are provided please put a tick
1.
Age:
2.
Sex:
in the applicable box.
(number of years)
male
female
3.
Have you had an episode of cough in the last two days?
Yes
D
No
4.
Do you smoke cigarettes?
Yes
□
No
5.
If “yes” to question 4, how many cigarettes do you smoke per day?
6.
If “yes” to question 4 why are you still smoking cigarettes?.
cigarettes
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
L- information
r research
questions or
- hypotheses
------- i
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
type of study
- data collection
techniques
sampling
- plan for data
collection
plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
P- material support,
and equipment
L- money
summary of proposal
Module 13 - PLAN FOR DATA ANALYSIS AND
INTERPRETATION
OBJECTIVES
At the end of this session you should be able to:
1.
Identify and discuss issues to be considered when preparing a plan for data
analysis
2.
Describe the important points to cover when interpreting results
3.
Prepare a plan in consultation with relevant experts or statisticians for the analy
sis of data and the interpretation of results for the research proposal being devel
oped
WHO/AFRO/SHDS
Joint HSR Project
112
WHY is it necessary to prepare a plan for analysis of data and interpre
tation of results?
Such a plan helps the researcher to make sure that, at the end of the study:
All the information he needs has indeed been collected, and in an appropriate way (standardized);
He has not collected unnecessary data which will never be analyzed.
Note:
This plan for analysis of data must be prepared before the data is collected in the field.
WHAT should the plan include?
It should include provisions for:
■
Systematic quality control of the data;
Processing data;
Data analysis;
These three points are discussed in more detail below.
1.
Systematic Quality Control of the Data
In the field, the data have already been checked to ensurethat all the information has been properly collected
and recorded.
However, before and during processing, the information should be checked again for completeness and
internal consistency.
If an inconsistency is clearly due to a mistake made by the researcher/research assistant (for example if a
person in an earlier question is referred to as a non-smoker, whereas all other questions reveal that he is
smoking), it may still be possible to check with the person who conducted the interview and to correct the
answer.
If the inconsistency is less clearly a mistake in recording, it may (in a small scale study) be possible to return
to the respondent and ask for clarification.
113
Ina large scale study, information that does not make sense may be excluded in the processing and analysis.
However, a decision to exclude data should be thoroughly considered.
2.
Processing Data
Decisions have to be taken on what type of data processing to use:
Doing the sorting, tallying and analysis by hand, with assistance of a calculator, or
By computer, i.e., using a micro computer with diskettes, and making use of packaged (or self
written) programmes for analysis.
Further, decisions have to be taken on whether and how to code the responses to questions.
The development of a well-planned coding system is important. Coding may be useful for sorting data when
doing the analysis by hand, and is essential for data entry when using a computer.
Codes may need to be developed for:
Quantitative variables: Decisions on howto group quantitative data, for example age, home clinic
distance, and therefore how to code them are usually taken after the data have been collected.
Qualitative variables: The coding of closed questions and of certain types of observations (for
example presence or absence of latrine in household) may be done before starting the study (pre
coding). This will save much time in the processing and analysis of data.
A particular problem may be coding answers from open-ended questions. Their coding can only
be decided upon after the data have been collected.
Note:
Here the sorting of data is not in the first place done to code and count, but to get an insig ht
into what action could be taken.
An open question into reasons for smoking, for example, will provide many answers (see Exercise
7, Part 4). We could sort them in categories of related answers, and rank them according to the
chance we attribute to each category to find a suitable entry for health education on smoking.
114
Data missing (for example unanswered questions) or data thrown out, need special codes.
Finally, you have to consider how you will summarize the raw research data in an efficient way, to facili
tate data analysis.
If the study is small, you may tally by hand all answers of individual respondents on a data mastersheet.
Example of a data mastersheet:
M
1
1
2
1
3
etc.
Cough
Sex
Respondent
F
Yes
No
No response
1
1
1
1
(See Exercise 7, Part 1 for more details).
From such a “mastersheet” you can more easily make tables which describe the study population than from
the original questionnaires.
3.
Data Analysis
In order to be able to further process and analyze the data meaningfully, we have to sort them in relation
to the objectives of the study and the variables we have selected. Remember that our study has to provide
answers to the problem we are investigating.
Imagine that we have conducted a study among a certain population, in order to find out who are smokers,
how smoking affects their health, and how health information could be provided to reduce smoking.
First of all we use the data collected to describe the population more in detail. This can be done in simple
tables, which may record, for example:
sex of respondents,
age,
residence,
115
smoking habits and/or
health status of informants.
More interesting than simple counts of the total study population would be the comparison of smokers and
non-smokers for the same variables.
This can be done by cross-tabulations. Cross-tabulations may show an association, and thus a possible
causal relationship between smoking and another variable, e.g. coughing.
I n our example cough, in the last to days would be the dependant variable; smoking of cigarettes would be
the independent variable or risk factor.
If we process the data by hand, we would have to make several dummy tables relating the dependant
variables to the independent variables, for example:
No cough in last two days
Cough in last two days
Smoker
Non smoker
m mi
III
3
One could tally in two ways, LH1 or
Will
W1___ 52
W1W1W1W1W1W1W1I
W1W1W1W1W1W1W1 71
LRl LW1
11
. The latter type of tallying has been used by village health
workers who were hardly literate.
Note:
It is extremely important to realize before you start collecting the data, what tables you will
need in order to look for possible explanations of the problem you have defined. This will
prevent you from collecting too little or too much data in the field. It will also save much
of your time in the data processing stage.
A schedule or work plan for the analysis of data should include:
The preparation of dummy tables for the description of the study population and for the establish
ment of relationships between variables;
The sequence of tables to be analyzed, or some priority of what should be analyzed first;
A specification of the staff required for the analysis;
A decision on whether all or some parts are to be analyzed by hand tallying or computer;
116
An estimate of the time the analysis is likely to take, and which aspects might take longer
than others;
An estimate of the total costs of the analysis.
Note:
Data analysis usually involves an amount of calculation and or statistical procedures.
Statistical procedures allowthe data, in particular of quantitative studies, to besummarized
and interpreted. In quantitative studies it is advisable to consult a person with statistical
knowledgefrom the start, in order that:
a correct sampling method is selected;
decisions on coding are made that will facilitate data processing and analysis;
a clear understanding and agreement is reached on the planned data analysis and
interpretation, including an agreement on the tables to be prepared and the testing
of relationship between certain variables. For this purpose, “dummy” tables have
to be prepared.
In general, a researcher carrying out a quantitative study should have or obtain some
statistical knowledge in order to appreciate the whole process of data analysis and inter
pretation.
The following diagram shows the steps in analysis data and interpretation of results:
Sort the data
I
Perform quality control checks on the data
i
Process the data
I
Analyze the data
I
Interpret the results
117
EXERCISE? :
Smoking (Data Analysis and Interpretation)
(In plenary, 3/4 hour)
Introduction
The questionnaire you have completed at the beginning of this session will be used in our group of work
shop participants:
To determine the age and sex distribution of the workshop participants;
To confirm the relationship between cigarette smoking and episodes of coughing during the past
two days;
To identify reasons why cigarette smokers do not stop smoking.
EXERCISE 7, Part 1 : Data Mastersheet
1.
Number the questionnaires for easy referral.
2.
Summarize the data on a data mastersheet. For fixed-response questions (Q 2, 3, 4,) this can be
done straight from the forms. For open quantitative questions (Q 1,5) just fill in the numbers.
The open-ended qualitative question (Q 6) will be analyzed in part 4 of this exercise.
118
EXERCISE 7, Part 1 : DATA MASTERSHEET
Question
naire
number
Q1
Q2
Q3
Q4
Q5
Age
(years)
Sex
Cough
Smoking
Number of
Yes No NR
Yes No NR
cigarettes
M
F
NR
2__
2__
A__
2. _ ._
6_
.8—
_9 —
10„.
11_„
12 —
13—
14_
13__
13
17___
IB___
19__
23___
21__._
22 __
23 __
2.4__
25__
26___
27 __
28 __
29 __
33___
31
32 _ . .
34
35
119
EXERCISE 7, Part 2 : Frequency Distribution
Quantitative variable: Age
The data on age can be interpreted more easily if they are summarized by grouping them. It would be useful
therefore to go through the following steps.
I nspect the column of numbers . What is their range? Could we divide the range in, say, three groups and
have a reasonable number in each group? In defining groups it is common practice to start with round
numbers (e.g. 15-29;
30-44; 45-59 etc.) Set up a table indicating how data are grouped, and tally all
observations:
For instance:
Age Group
No
%
15-29
18
40
30-44
14
31
45-59
12
27
No response
1
2
TOTAL
45
100
Qualitative variable: Sex
Calculate also the sex distribution in percentages. The total numbers can be taken from the mastersheet.
Sex
%
No
Male
Female
No response
TOTAL
120
EXERCISE 7, Part 3: Cross-tabulation
After having made a simple overview of the distribution of variables, we would now like to see how variables
are related by making cross-tabulations. Cross tabulations can be made by hand tallying from the master
sheet. For instance, in table 1 each individual can be tallied in one of the four cells. Check that the totals
fit with the totals on your mastersheet!
Another convenient way of preparing cross-tabulations by hand is hand-sorting of the individual forms. For
instance to make a cross-tabulation of coughing in the past two days of smokers and non-smokers (table
1) we go through the following steps:
1.
Divide the forms into two piles, one for smokers and one for non-smokers.
2.
Divide each pile into one for those without cough and one for cough (we now have four piles).
3.
Count the number in each pile and fill in the table.
You may notice that when sorting you have to manipulate the forms a lot. This is a reason for choosing fairly
strong material when developing forms for sorting.
Table 1. Episodes of coughing during past two days in smokers and non-smokers
Cough
No cough
TOTAL (No)
Smokers
Non-smokers
TOTAL (No)
TOTAL (%)
After completion of the tables, answer the following questions:
1.
What percentage of smokers had cough within the past two days?
2.
Does this study confirm that smoking causes coughing?
121
TOTAL (%)
EXERCISE 7, Part 4 : Analysis of qualitative data from open ended
questions
You will now analyze and interpret answers to question 6 of the questionnaire: i.e. Why are you still
smoking?
Imagine that the following answers were given to this question:
I have tried to give up so many times but I have been unable to
I like the feel of the cigarette in my hand
Because it gives me pleasure
I do not see why I should give up smoking
Because I like to blow the smoke through my mouth and
nose
Because I feel confident and in charge when I am smoking
It helps me to think better
I like the image that comes with smoking
I feel that people respect me more as a smoker
All my friends are smokers
It helps to make people more friendly and comfortable, especially when offering a cigarette
Why not?!!
Smoking makes me feel like a man
I like to blow smoke rings
I like the taste
It is too difficult to give up
It helps me to relax
It helps me to reduce the work pressure and tension
My wife likes a man who smokes
122
These answers may be analyzed and interpreted through the following steps:
1.
Develop a post-coding system by grouping the responses provided. Try to group them in such a
way that it is easy to find possible suggestions or recommendations for action that can be taken
(for example, entry points for health education in order to reduce smoking).
2.
After grouping and coding give your interpretation of these responses.
123
GROUPWORK (2hours)
Using the time guide provided each working group will now make a plan for data analysis
and interpretation considering the following points:
10 min
1.
Systematic control of data: Who? How often?
consistency checks can be suggested?
50 min
2.
Processing the data:
What
How will you do it (by hand, by computer?)
How many open ended questions do you have which
require categorizing or coding? How many closed
ended questions? How big is your sample? How much
time would it require?
Who will do the categorizing/coding? Who will super
vise?
Prepare mastersheets for your proposal
Ihr
3.
Analyzing and interpreting the data
Prepare dummy tables in which you relate variables to
each other to analyze possible causal relationships
Select the most important ones that you would like to fill
in before we have our next workshop on data analysis
and interpretation
Make estimates on staff, time and costs involved in the
analysis.
5 min
Plenary
Present your mastersheet, three dummy tables and
rough estimates of manpower time and costs required to
the plenary.
Note:
In our special case you only have to plan the time and staff it would require to prepare and
fill in the master-sheets and the selected dummy tables, as we will continue to work on the
analysis during the second workshop after your research project has been implemented.
124
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
■ other available
L- information
P research
questions or
- hypotheses
----- r~
r— variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
type of study
- data collection
techniques
sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
ethical
considerations
1— pilot study
or pre-test
r- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
material support,
and equipment
*- money
summary of proposal
Module 14 - ETHICAL CONSIDERATIONS IN HEALTH
SYSTEMS RESEARCH
OBJECTIVES
At the end of this session you should be able to:
1.
Identify ethical issues and problems that should be considered when designing
studies in health systems research
2.
Describe the responsibilities of the researcher concerning ethical issues
3.
Locate sources of information and guidelines available concerning ethics in health
systems research
4.
Identify and take action on ethical issues which should be considered in the
research project you are developing during this course
WHO/AFRO/SHDS
Joint HSR Project
126
Introduction
Ethical considerations in the health profession and in health systems research are extremely important. It
is crucial that the effects that your research might have on individuals, groups, institutions and communi
ties be thoroughly considered before final plans for research are made. Ethical considerations also apply
to the scientific integrity of the researcher during all phases of the research process.
■
.
■
Ethics are rules of conduct concerned with preventing the possibility of inflicting
harm upon others.
■
•
-
Ethical issues can apply to every step in the research process
1.
In the choice of research questions, statement of the problem and literature review, ethical
issues to be considered could include:
Are the research topics acceptable to the population studied? For example, is there
enough consideration for religious and cultural beliefs and practices?
Have we left out relevant information in the literature review or in the justification of the
research proposal because it does not support our views on the problem?
2.
In choosing study types, data collection techniques, and when planning for data collection,
questions concerning the following ethical issues should be raised:
Are we withholding beneficial procedures to certain informants,
for example by:
witholding services from those who need it;
witholding treatment of the control group, or any other persons in need of
treatment whom we might encounter during our research;
Might research procedures cause any physical or emotional harm, for example:
through researchers failing to observe/respect certain cultural values or tradi
tions;
by intruding into informants’ privacy with sensitive questions;
127
Have we planned to obtain informed consent of the participants? Will the goals and pro
cedures of the research be explained beforehand, and will consent be obtained? This is
particularly obligatory with respect to medical procedures such as:
taking samples for blood tests;
drug or vaccine trials;
Note:
Advantages should not be taken of minors, mentally disabled adults, etc.
Should we assure confidentiality of data to protect the respondents?
Should names, addresses or even names of villages be recorded?
Are informants observed without their knowledge (hidden cameras)? How do we
record this information?
Is scientific integrity assured?
Are data collection tools developed in such a way that distortion of information to reflect
what the researcher wants is difficult?
3.
In the analysis and dissemination of results, you need to ask the following questions:
Are the conclusions drawn justified, in view of the data collected?
Has feed back been given to the community and health staff involved in the research?
Are we revealing data to the authorities in such a way that the privacy of respondents is as
sured?
Has action been taken to ensure utilization of the research results?
Has all information that will be useful to others been published, and in a form that is under
standable and accessible to local staff and local researchers?
Note:
It is important to realize that most proposals will need ethical approval by national
authorities before the research can begin.
128
GROUP WORK
(45 minutes)
Working as a group, spend 45 minutes looking at your own proposal to see whether ethical
issues and considerations have been included:
1.
Consider whether there are any ethical issues or problems that may arise in the
research project being developed as it is now designed. To do this, go through
each phase in your proposal and ask the question: “Will any harm come to
someone if we follow our plan as specified?”
Be aware that harm might also come to people by not carrying out the research.
Make any changes necessary in the proposed methodology.
2.
Provide suggestions on possible ethical issues during data analysis and/or the
dissemination of data.
3.
Determine and outline local procedures necessary to have consent for your
research proposal. Make sure that these steps are incorporated as you develop
your overall project workplan.
4.
Develop a small section for the proposal which discusses ethical issues in relation
to the project being proposed.
5.
Summarize these issues on a flipchart for presentation in plenary.
129
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
I
formulation of
objectives
Important elements
of each step
- literature and
other available
information
r- research
questions or
- hypotheses
j- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
type of study
- data collection
techniques
sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
[- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
P material support,
and equipment
money
summary of proposal
Module 15 - PRE-TESTING THE METHODOLOGY
OBJECTIVES
At the end of this session you should be able to:
1.
Describe the necessary components of a pilot study that will allow you to test, and
if necessary, revise your proposed research methodologybefore starting the study
2.
Plan and carry out pre-tests of research components for the proposal being de
veloped.
WHO/AFRO/SHDS
Joint HSR Project
132
WHAT is a pre-test or pilot study of the methodology?
A pre-test refers to a small scale trial (field-testing) of a particular research
component.
Ap
t
rying out a preliminary study going through the
entu
ntire research procedure with asmall samplee'
:"
aBte
___
WHY do we carry out a pre-test or pilot study?
A pre-test or pilot study serves as a trial run which allows us to identify potential problems with the proposed
study and make revisions if necessary. Although this means extra effort at the beginning of a research
project, the pre-test and/or pilot study allows us to encounter problems on a small scale before starting a
major study. Adaptations can then be made, saving a good deal of time, effort and money in the long run.
WHAT aspects of your research methodology can be evaluated during
pre-testing?
i.
Reactions of the respondents to the research procedures can be observed in the pre-test to
determine:
Availability of the study population and how respondents’ daily work schemes can best be
respected
Acceptability of the methods used to establish contact with the study population
Acceptability of the questions asked
Willingness of the respondents to answer the questions and collaborate with the study
2.
The data collection tools can be pre-tested, to determine:
Whether the tools you use allow you to collect the information you need and if those tools
are reliable. You may find that some of the data collected are not relevant to the problem
or are not in a form suitable for analysis. This is the time to stop collecting these data and
to revise, or to consider using additional techniques
The time needed to administer the questionnaire, to conduct observations or group
interviews, or to make measurements
133
Whether there is any need to revise the format or presentation of questionnaires/interview
schedules:
Sequence of questions
Wording of the questions (clear?)
Space for answers
Precoding of some answers, or changing closed questions into open questions
Accuracy of translations
Need for additional instructions for interviewers (guidelines for "probing” certain
open questions?)
3.
Sampling procedures can be checked for the following points:
Whether the instructions to obtain the sample are used in the same way by all staff involved
How much time is needed to locate individuals to be included in the study
4.
Staffing and activities of the research team can be checked, while all are participating in
administering the pilot study, to determine:
How successful the training of the research team has been
What the work output is of each member of the staff
What the group dynamics are within the team
Whether logistical support has been adequately arranged
Reliability of the results when instruments or tests are administered by different members
of the research team
Accuracy of interpretation
How well the proposed plans for staff supervision function
The pre-test can be used as a period of extra training for the research team in which sensitivity to
the needs and wishes of the population studied is developed.
134
5.
Procedures for data processing and analysis can be evaluated during the pre-test, including:
The appropriateness of data master sheets and d ummy tables and the ease with which they
can be used
The effectiveness of the system for quality control of data collection
The appropriateness of the statistical procedures (if used)
The clarity and ease with which data can be interpreted
6.
The proposed workplan for research activities can be assessed during the pre-test including:
The appropriateness of the amount of time allowed for the different activities of planning,
implementation, supervision, coordination and administration
The accuracy of the scheduling of the various activities
Note:
If all the above are undertaken in a field test the process would be called a pilot study.
Which components should be assessed during the pre-test?
The pre-test may resolve two different issues:
1.
Firstly, if you suspect certain limitations or weaknesses in your methodology, you may try to find
out how serious they are.
When planning the pre-test, the first activity you might undertake is preparation of an outline of
weaknesses and limitations in the present proposal. It may be useful to go through the possible
sources of bias in data collection techniques and sampling and through the section on ethical con
siderations to identify major issues.
Can some of these limitations be overcome by adapting the research design? Is the extent of some
of these limitations unclear and should we try to measure them in the pre-test?
135
2.
Secondly, the pre-test may help you to plan in more detail the time and resources needed for your
study.
Therefore, during the pre-test, you should focus in particular on those components that are
likely to have the largest impact on thetotal time and budget. You must pay special attention
to areas where the greatest uncertainty exists. (For example the testing of questionnaires
or other data collection techniques).
What issues should be considered when selecting the sample popula
tion with which the pre-test or pilot study will be conducted?
The sample population studied should be similar in important characteristics to the population that
will be used in the study itself.
The sample for the pre-test should include neither the same respondents as for the full study nor
respondents who are likely to influence the participants in the full study in any way.
WHO should be involved in the pre-test or pilot study?
The entire research team, headed by the principal investigator.
HOW LONG should the pre-test or pilot study last?
The time required for a pre-test or pilot study will be determined by a number of factors which may include:
What and how many components of the research methodology you want to pre-test
The size and duration of the research project (the longer the study will take, the more time you might
reserve for the test run)
The complexity of the methodology used in the research project
The main objective: to test the methodology to ensure that the research can be carried out according
to plan, must however be achieved.
136
GROUPWORK
(1-1/2hours)
A field exercise is being planned during which each working group will have a chance to
pre-test certain research components. Data collection tools will probably be one of the
components tested.
1.
Develop a plan for a pre-test of your research proposal, including the
following information:
What sample population will be used for the pre-test, and how will they be
selected?
Who will carry out the pre-test?
What data collection tools, research procedures and other aspects of the
project will be tested during the pre-test and how? (see Annex 15.1 for
ideas) How will this information assist in the final design of the study itself?
Be aware that only half a day is available for this purpose
2.
Draft a summary description of the pre test for inclusion in the appropriate
methodology sections of the proposal.
3.
Select data collection tools that you will try out during the pre-testing and
prepare for the exercise:
It should be easy to pre-test the data collection tools selected with respon
dents during the exercise
Ensure that your tools are ready for use. If they have not been pre-coded
and this would be useful, you might pre-code them now
Prepare a short list of questions you wish to answer during the pre-test of
the data collection tools
Ensure that sufficient copies will be available for use during the field
exercise.
4.
During the field exercise you are advised to work in pairs. Each pair will have
an opportunity to test the data collection techniques.
5.
At the end of the field exercise, you will analyze with the working group as a whole
what you have learned from the pre-test and (if necessary), make changes in your
research design and data collection techniques or tools.
6.
Prepare a summary of the changes you have made to your research proposal
as a ressult of the pre-test for presentation in plenary.
137
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
L- information
research
questions or
- hypotheses
t- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
T
summary of proposal
- administration,
monitoring and
- evaluation
material support,
and equipment
L- money
Module 16- UTILIZATION AND ACTIVE
DISSEMINATION OF RESULTS
OBJECTIVES
At the end of this session you should be able to:
1.
Defend the fact that the most important reason for undertaking health systems
research should be to obtain results that can be disseminated and used to improve
health care
2.
Suggest what the researcher can do during the design and preparation of the
proposal, during the implementation of the study and at the end of the project in
order to facilitate utilization and active dissemination of the research results at all
levels
3.
List the essential components of a good research report
4.
Prepare a plan for actively disseminating results and fostering their utilization for
the project proposal being developed.
WHO/AFRO/SHDS
Joint HSR Project
140
WHY should the researcher be concerned about utilization and dissemi
nation of research results?
The fundamental reason for undertaking health systems research is to obtain results that can be used to
improve health and health care. Results may be used to:
Improve or expand the delivery of health services
Improve the training of health workers
Facilitate community involvement in health activities
Solve health problems by means other than health services delivery
Provide useful information to other groups and service organizations with similar problems
WHO will be interested in the results?
The results may be useful to the community, staff and managers of health and health-related services, donor
agencies and even people in other countries.
This may require different summary reports to be prepared, suitable to each audience. Often researchers
try to make a single report for all of those who would be interested. This reduces its relevance and actual
usefulness.
WHEN should planning for the utilization and dissemination of results
take place?
Before the project begins, while designing the research and drafting the proposal
During your implementation of the project
Towards the end and when you have finished your project
We will discuss these phases in more detail below.
1.
Design Phase
Several activities, if carried out during thedesign phase, will make it more likely that your results will be utilized
and disseminated.
141
Before settling on a research topic, involve government officials, health workers and managers,
as well as community members, to make sure that you select a priority problem that has their
interest.
If the project is designed to produce results that should be fed back directly to improve a particular
programme, you might consider organizing a Project Advisory Committee that meets regularly,
including members such as:
The principal investigator and key staff
Major potential users of results, including key decision-makers
Representatives of the target community
The Project Committee may in fact be small but it would it would serve a vital function in keeping
key people informed.
Identify additional individuals, groups, and institutions that you feel would be interested in your
results and be able to use them, for example NGO’s, school-teachers, agricultural extension
workers or managers of other public sectors and international organizations.
Determine how the results might best be used for decision-making and how the project should
be scheduled to ensure the results will be available when they are needed for this purpose.
Devise one or several strategies for disseminating the results to the target audiences you have
identified and for promoting utilization of these findings. This may involve dissemination through
discussions/workshops, news media, television, posters, and various reports.
Outline a plan for disseminating results and promoting utilization of findings and include it in the
research proposal. Allow adequate funds and time in the project for the activities you propose.
2.
Implementation Phase
Once the research project has been approved and in the process of implementation, some activities may
help to keep everyone concerned interested in utilizing the final results. These include:
Continuing to involve everyone concerned in implementation of the project
Preparation, distribution and discussion of regular progress reports on the research with project
supervisors, the study population, funding agencies, and others interested in the project.
3.
Final Phase
Planning dissemination of results for the final phase includes preparation for reporting and several other
methods of dissemination.
142
■
Reports
A preliminary report is most useful if it is produced as quickly as possible and directed to the
decision makers and managers whose cooperation is needed for implementation of the research
results. If there is an Advisory Committee, this would meet to discuss findings and preliminary
recommendations of a draft report.
Don’t neglect to report the findings to the subjects and/or community or organization studied
before the report is finalized.
This is done to fulfil an obligation to those studied, to obtain
information on possible errors in the report, and to discuss your proposed recommendations in
more detail.
Prepare a final report that is clear, interesting, practical and complete. Include recommendations
for action that give clear instructions on how the results might be used. These recommendations
may differ according to your various target groups (health staff, community).
Other methods of dissemination
Decide on other methods for disseminating your results that are appropriate to the project and the
variety of audiences you want to reach. These might include:
Discussions;
Workshops or seminars where results are presented and possible use by interested groups
is explored
Short presentations of results through slide shows, film, radio, television, posters, or
newsletters.
Submission of articles for professional journals published both within the region and
internationally
Development of technical or training manuals
Note:
Remember not to wait too long before disseminating your results. If you delay too long you
may find that interest and enthusiasm have dropped and that the results are no longer
relevent.
143
Preparation of a Research Report
In the final two weeks of your course (following the 5-6 month period of project implementation), you will
be covering the topic of report writing in great detail. However, in order to plan realistically for this phase,
and allocate appropriate time and resources, the following guidelines for preparation of your final report can
be helpful.
WHO should write the final report(s)?
The final reports are written by the research team, for example split up in groups of two persons per
chapter or section. The principal investigator has the overall responsibility for delegating tasks and
putting the different parts together.
WHAT essential sections should be included in a final report?
The content would vary according to the type of report and the target audience. Annex 16.1 gives
an outline of a comprehensive research report. For most target audiences only small parts of this
comprehensive report would be useful.
What guidelines might be given concerning preparation of a good report?
Simplicity and clarity are important. Avoid technical jargon or a stuffy style.
Provide a short summary of the problem, research objectives, methodology used, and
major findings and recommendations at the beginning of the report.
Be brief, concise and to the point. If lengthy detail is necessary, try to include as much as
possible in your appendices.
Avoid emotional and subjective statements and back up all conclusions with data.
Present the report in an attractive format, using appropriate charts, graphs and pictures,
if relevant.
Provide a report that can be read by itself, rather than one that refers the reader to other
sources of information not readily available. Make the report practical in orientation, so that
readers can easily utilize the research results in their own working situation.
GROUPWORK
(1-1/2 hours)
Divide into your working groups and prepare a plan for disseminating and promoting
utilization of the research results, using the checklist provided in annex 16.2.
144
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
I
formulation of
objectives
Important elements
of each step
- literature and
other available
information
r- research
questions or
- hypotheses
--------- i—
1
research
methodology
i— variables
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
J
I
summary of proposal
- administration,
monitoring and
L- evaluation
material support,
and equipment
*- money
Module 17 - WORK PLAN
OBJECTIVES
At the end of this session you should be able to:
1.
Describe the characteristics and purposes of various project planning and
scheduling techniques such as “work scheduling” and “GANTT charting”
2.
Determinethe staffyou need for the various tasks inyour projectand describe how
you will select and train appropriate individuals
3.
Prepare a work schedule and a GANTT chart for the project proposal you are
developing
WHO/AFRO/SHDS
Joint HSR Project
146
A.
WORK PLANNING AND SCHEDULING
WHAT is a work plan?
A Work plan is a schedule, chart or graph that summarizes in
• a cle.r,ashlw
various components of a research project and how they fit together.
_____________________________
■
It may include:
The tasks to be performed;
When the tasks will be performed;
Who will perform the tasks and the time each person will spend on them.
Various Work Planning and Work Scheduling Techniques
1.
Work schedule
A work Schedule is a table that summarizes the tasks to be performed in a
research project and the duration of each activity.
•
;
Va.
.
.
.
, . .
-
■
.
-
;.
j
The version of a work schedule given on page 119 includes:
The tasks to be performed;
The dates each task should begin and be completed;
Professional person-days required;
Professional personnel assigned to the task;
Non-professional person-days required;
Non-professional personnel assigned to the task.
The categories used in a work schedule can be adjusted, depending on the needs of the particular project.
147
EXAMPLE OF A WORK SCHEDULE
Calendar Week
Date
Begin End
Professional
Person-days
required
Prof. Personnel
Assigned to
task
Non-Prof.
Person-days
required
Non-Prof.
Personnel
Assigned
task
1. Prepare proposal
and submit to govt.
Week 1 through
Week 6
March 15-Apr. 25
30 days
Principal Investi
gator (P.I.),
Research
Assistant (R.A.)
30 days
Secretary
2. Discuss study with
government
Week 2
March 22-26
1 day
P.I.
3- Receive reply from
government
Week 12
April 21-25
4. Order water testing
equipment and supplies
and receive
Week 12 (6 weeks
delivery time)
April 21-May 6
2 days
Secretary
5. Advertise, interview
and select data
gatherers
P.I. (10 days)
R.A. (10 days)
3 data gatherers
(30 days total)
10 days
Secretary
P.I., R.A., and
3 data gatherers
full time
30 days
Secretary
P.I. (10 days)
R.A. (20 days)
30 days
Secretary
Tasks to be performed
P.I.
2 days
R.A.
Week 12 through
Week 14
May 2-20
20 days
P.I. (15 days)
R.A. (5 days)
6. Conduct training
of data gatherers
Week 16 through
Week 17
May 30-June 10
50 days
7. Conduct testing
of water
Week 18 through
Week 23
June 13-July 22
150 days
8. Analyze data
Week 24 through
Week 30
July 25-Aug. 19
30 days
9. Prepare report
Week 26 through
Week 30
Aug 8-Sept. 9
40 days
P.I. (25 days)
R.A. (15 days)
10 days
Secretary
10. Prepare workshop
on findings
Week 31 through
Week 32
Sept. 12-23
20 days
P.I. & R.A.
full time
10 days
Secretary
11. Hold workshop
Week 33
Sept. 26-30
10 days
P.I. L R.A.
full time
5 days
Secretary
WHO/AFRO/SHDS
Joint HSR Project
148
Comments
2.
GANTT Chart
The GANTT chart is used for planning activities by depicting graphically the order
in which various tasks must be completed and the duration of each activity.
The GANTT chart shown on page 121 lists:
■ f
The tasks to be performed;
The responsibility for each task;
The time each task is expected to take.
The length of each task is shown by a bar that extends over the number of days, weeks or months the task
is expected to take.
How can work plans be used?
Work plans can serve as:
Tools in planning the details of the project activities and later in budgeting funds
Visual outlines or illustrations of the sequence of project operations. They can facilitate presenta
tions and negotiations concerning the project with government and institutional authorities and
other funding agencies
Guides or management tools for the principal investigator and members of his team, showing what
tasks and activities are planned, their timing, and possibly when various staff members will be
involved in various tasks
Tools for monitoring and evaluation, when the current status of the project is compared to what had
been foreseen in the work plan
149
§
I
O
>
"n
X
O
----------------- r
l
Tasks
1. Prepare proposal and
submit to govt.
c/>
X
□
cn
Example of
A GANTT Chart
Title of Project:
Phase of Project:
Responsibility
Principle Investi
gator (P.I.) and
Res. Asst. (R.A.)
2. Discuss study with
govt.
P.I.
3. Receive reply from
govt.
P.I.
4. Order water testing
R.A.
equipment and supplies
cn
o
5. Advertise, interview
and select data gaterers
P.I. and R.A.
6. Conduct training of
data gatherers
P.I. and R.A.
7. Conduct testing
P.I., R.A. and 3
lata gatherers
8. Analyze data
P.I. and R.A.
9. Preoare report
P.I. and R.A.
10. Prepare workshop on
findings
P.I. and R.A.
11. Hold workshop
P.T. and R.A.
!
i
i
i
l
i
•
4onth 1 ^onth 2 1ponth 3 [ ' Month 1 Month -SMonth
6
i
____________ i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
■!
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
—i--- r~
i
th 7 Month
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
rh 9
I
l
I
i
I
I
l
I
I
I
I
I
i
i
i
i
i
I
i
i
i
i
i
I
I
I
I
i
i
I
I
l
i
I
i
I
I
i
I
I
l
l
I
I
i
I
i
i
I
i
i
l
i
I
i
I
I
I
I
J
I
I
I
I
I
I
I
I
I
I
When should the work plan be prepared and when should it be revised?
Note:
The first draft of the work plan should be prepared when the project proposal is being
prepared, so the schedule can be discussed easily with the government, institutional
authorities and other possible funding agencies.
A more detailed work plan may be prepared for each stage of the project as it begins.
There should be no hesitation in revising work plans or preparing new ones as the project gets
underway, based on an assessment of what can be realistically accomplished in the coming
months, considering the current and future situation and their constraints.
Note:
Some crucial factors should be kept in mind when preparing
a work plan:
It should be simple, realistic, and easily understood by those directly involved.
It should cover the preparatory and the implementation phases of the project, as well
as data analysis, reporting and dissemination/utilization of results.
The activities covered should include technical or research; administrative; secre
tarial and other support level; and training tasks.
The realities of local customs (local holidays, festivals) and working hours should be
considered, when preparing the work plan.
Also seasonal changes and their effect on travel, work habits, and on the topic you
are studying (such as incidence of disease, nutritional status), are factors to keep in
mind.
151
B.
SELECTION OF PROJECT STAFF
The manpower needed to carry out a project forms a crucial component of the work plan and budget of any
project.
Staffing needs can be determined by:
Looking at your project research methodology and work plan, and considering the tasks to be
accomplished, deciding upon the types of personnel likely to be necessary at each stage of the re
search process
Making a list of the types and numbers of staff needed and the time period and number of hours
for which each person would work on the project
Considering whether the use of consultants is necessary. This should include whether or not the
project staff have the necessary skills for certain technical tasks, and whether short term assistance
is helpful. Always consider using local consultants. If consultants are used, involve them already
in the planning stage of the project.
In reviewing your tentative staffing plan you should ask:
Is the plan realistic, taking into account the project budget likely to be available?
Are specific requirements in terms of types of personnel and levels of expertise likely to be available
for the project?
For example: Isthere a sufficient range of disciplines represented among the project staff, including,
where appropriate, personnel from outside the health field?
Towhat extent can community members, traditional healers, students or other non-professionals
be involved in the study?
■
If special staff has to be recruited or reassigned from other ministries or agencies; what regulations
will have to be observed, and what will the conditions of service be?
Would it be useful to provide pre-project or on-the-job training for certain staff members participat
ing in the project? (This may increase motivation to participate.)
152
Presentation of the project staff in the work plan
Specify, in a clear, brief style, the type of staff needed and how they will be used in the project. After
an introductory statement, it may be useful to list each position and give a short description of the
job qualifications, responsibilities and work schedule involved
If particular individuals have been selected for project positions, their names and a brief statement
of their qualifications can be listed.
GROUPWORK
1.
(3-1/4hours)
Prepare in your working group for the proposal you develop:
An outline for selection and training of project staff, which will form part of
your work plan.
List the types and numbers of staff you need. Aim for training of existing
staff, rather than selection of new staff. Include in your outline when and
for how long you will need local consultants.
A work schedule and a GANTT chart.
Ensure that you cover all phases of the project, and include activities such
as technical/research tasks, administration, and training. Use the infor
mation on time required that you collected during the pre-test. Take into
account the realities of local customs and work scheduling. Use a
calender to allow for public holidays etc.
Ensure that your work plan and outline for selection and training of project
staff are realistic, as you will have to use them as a guideline in your practical
work the coming 3-5 months.
2.
Copy your GANTT chart onto a flipchart for use in exercise 7 and/or in plenary
discussion.
EXERCISE 7 (Optional) : Project Work Plan
Discuss the work plan developed by another group fortheir research proposal and provide
constructive criticism.
153
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
L- information
j- research
questions or
- hypotheses
------ r—
r- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
L pilot study
or pre-test
p- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
administration,
monitoring and
evaluation
j— material support,
and equipment
L money
summary of proposal
Module 18 - ADMINISTRATION, MONITORING AND
EVALUATION
OBJECTIVES
At the end of this session you should be able to:
1.
Identify arrangements that must be made for administration of a research project
when the project is being designed
2.
List the responsibilities of the principal investigator related to the administration
and monitoring of an on-going project
3.
List the administrative and final evaluation activities that should be completed at
the end of a project
4.
Prepare a brief plan for administration, monitoring, and final evaluation of the
research projects being developed
WHO/AFRO/SHDS
Joint HSR Project
156
ADMINISTRATION IN RESEARCH PROJECTS
A.
WHAT is project administration?
Project preparation is the term for all the activities involved in managing the
human, material, financial and logistical resources of a project.
'
WHY is good administration important in a research project?
It assists in the management of selection, contract preparation, hiring and payment of project
personnel and local consultants
It allows for orderly and accurate purchase and procurement of equipment, payment of bills, and
preparation of financial reports
It allows researchers to foresee needs for funds and to make timely requests in order to avoid gaps
in the implementation of the project
Hallows researchers to devote most oftheirtimetothe technical and scientific aspects of the project.
WHAT administrative issues should be considered
When developing a research proposal, the principal investigator should consider the following:
Which organizational unit or official has the power to receive research funds?
Which service or organization unit has the “know- how” and resources to ensure good administra
tion of a project? In other words, what service or unit is best able to:
work in collaboration with the principal investigator as well as with the funding authorities
to ensure flow of necessary funds, including “petty cash” for minor operating expenses
avoid creating unnecessary bureaucratic or administrative difficulties for the research
Would the creation of a project advisory committee be useful, comprising of managers with interest
in the research project and funding authorities? (if not already considered when planning utilization
of results).
157
Related to administration, the role of the principal investigator during the project implementation
would be:
To make certain that the managerial or administrative team understandsthe work of the researchers
To alert administrative officials in timely fashion concerning staffing, materials and equipment
needs, supplying necessary information concerning costs and reasons for requests
To supervise the flow of funds, project accounting and preparation and submission of financial
reports
To discuss with the research project advisory committee, if any, the difficulties encountered in the
project and attempt to identity solutions adapted to the situation
To arrange for continuity in administration during periods when the principal investigator will be
absent
At the end of the project, the principal investigator would need to supervise the following administra
tive operations:
Writing and presentation of the final report and dissemination of results
Making sure that all financial obligations are met.
B.
PROJECT MONITORING
WHAT is project monitoring?
.
Monitoring is the on going process by which information is gathered concerning
activities designed to keep track of resourcesavailable and used, and the quantity and
quality of the operations carried out in each phase of the project.
____________________
Monitoring should take place throughout the project and should be helpful in alerting staff to problems that
develop, and changes that are needed. It is a valuable management and learning tool for everyone
concerned.
158
WHAT should be reviewed during monitoring sessions?
The resources needed for the project, including staff, equipment, supplies, logistical support and
funds, to assess if they are available, and appropriately used
The activities of each team member and their relations to the project as a whole, to assess if they
are being carried out as planned, and what delays or difficulties, if any, have developed that need
addressing
The flow and quality of the data that are being collected
The communication and coordination of the research team with the study population, other
collaborating groups and with funding authorities.
WHO should be involved in the monitoring process?
Those involved should include:
the principal investigator;
the principal administrator;
members of the research team;
members of the project advisory committee if any.
WHEN should monitoring take place?
The principal investigator of the research team should monitor the project on a continuous basis.
At project advisory committee may meet on a regular basis to discuss project development with the principal
investigator and/or research team.
159
C.
PROJECT EVALUATION
WHAT does final evaluation of a research project involve?
I/;'.’-#’ .V,
‘
Evaluation involves a systematic collection and analysis of data to assess
whether the objectives of the research project have been met.
■/
-
•
~
________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ _________________________________________________________________________________
What is the difference between project monitoring and final evaluation?
Monitoring is an on-going process carried out on a periodic basis throughout the project to assure that the
project remains “on target”. The final evaluation is undertaken at the end of the project to see if the project
has obtained its goals. Data available through the monitoring process can be used during the final
evaluation.
HOW should a research project be evaluated?
The final evaluation of a research project should not be lengthy, and it should be carefully organised and
carried out in a systematic way. The following factors should be considered:
Was the statement of the problem clear?
Were the objectives or hypotheses clear, measurable, and realistic? Were the objectives achieved
or the hypotheses tested?
Were the methodology and instruments used appropriate, valid, and reliable? Was the research
design implemented as planned?
Was the analysis of data and interpretation of results accurate and complete?
Have the results been adequately disseminated and the findings used?
Has the project been successful, as perceived by:
policy makers, local population, health workers, health managers and the research team itself?
Have any administrative problems arisen which affect the running of the project relating to staff,
resources, transport, etc.
Was thefinancial investment worth it, considering the project’s accomplishments and the relevance
of findings to priority health needs?
160
Are there any follow-up activities or further research efforts recommended? What has been learned
on this project that can be applied in future research?
GROUPWORK (1 hour)
1.
Develop inyour working groupaplanforadministration, monitoring and evaluation
of your project proposal. Take the following points into consideration.
Administration
Which organizational unit or official would be best suited to administer the
project? (Be aware that the principal investigator cannot at the same time
be the principal administrator.)
Who will be the principal investigator?
Which authorities are likely to fund the project?
How could a smooth flow of funds be assured?
Would the creation of a steering committee be useful? Who should be
members? At what phases of the project (or at what intervals) should they
meet?
Monitoring
Who will be monitoring the project?
How and how often would you plan to do this?
What will be monitored?
Evaluation
Who will participate in the final evaluation?
When will it take place?
On what major points should the evaluation concentrate, in your opinion?
How would you evaluate whether the policy makers, the local population,
the health workers/health managers and the research team itself perceive
the project as successful.
2.
Prepare a summary of your plan for administration, monitoring and evaluation on
a flipchart for presentation in plenary.
161
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL;
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
Important elements
of each step
- literature and
other available
L- information
research
questions or
- hypotheses
f— variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
- pilot study
or pre-test
j- manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
1
♦
summary of proposal
- administration,
monitoring and
- evaluation
p- material support,
and equipment
L money
Module 19-BUDGET
OBJECTIVES
At the end of the session you should be able to:
1.
Select or develop appropriate major categories for a
2.
Make reasonable estimates of the expenses in various budget categories
3.
Describe common difficulties encountered during budget construction and use,
and how to avoid potential problems
4.
List various ways a budget can be reduced, if necessary, without substantially
damaging a project
5.
Prepare a realistic and appropriate budget for the project proposal being devel
oped during the course
WHO/AFRO/SHDS
Joint HSR Project
164
budget
WHY do we need a budget?
Adetailed, carefully thought out budget will be useful in planning, implementing and monitoring the
project.
The process of budget design may help you think through aspects of the work plan you have not
thought of before and serve as a useful reminder of activities planned, as your research gets
underway.
A detailed budget helps you to identify which resources are already locally available and which
additional resources may be required.
WHEN should budget preparation begin?
A complete budget is not normally prepared until the end of project planning. However, cost is usually a
major limiting factor and therefore must always be kept in mind during the planning stages to avoid
unrealistic proposals being developed. (See Module 4: Statement of the Problem)
The use of locally available resources increases the feasibility of the project from a financial point of view.
HOW should a budget be prepared?
It may be convenient to use the work plan as a starting point. Specify for each activity in the work plan what
resources are required. Determine for each resource the unit cost and the total cost.
Example:
In the work plan of a study to determine the utilization of family planning methods in a district, it is
foreseen that 5 interviewers will each visit 20 households in clusters of 4 over a time period of 5
working days. A supervisor will accompany one of the interviewers each day using a car. The other
4 interviewers will use a motor cycle. The clusters of households are scattered over the district but
are on average 50 kilometres from the district hospital from where the study is conducted.
The budget for this component of the work plan could be made as follows:
Resources required:
1.
2.
3.
Personnel
Transport
Stationery
165
1.
Personnel
Number of staffdays (no. staff
x no. working days)
Daily wage
(ini. subsistence)
Total
wage
Interviewers
$10
25
$250
Supervisor
$20
5
$100
$350
1. Personnel TOTAL
2.
Transport
Number km
(no. vehicles x cost
no. days x km/day)
Cost per
km
Motorcycles
Car
2000
500
$0.10
$0.40
Total
$200
$200
$400
2. Transport TOTAL
3.
Stationery
Pens
Questionnaires
Unit Cost
Number
Total
$1.00
$0.20
12
120
$12
$24
3. Stationery TOTAL
$ 36
GRAND TOTAL
$786
However, if more than one budget source will be used.e.g. Ministry of Health and a donor, it would be useful
to split up costs from the start according to who will bear the costs. The total budget for the proposal in this
example might look like Annex 19.1.
166
Advice on Budget Format
A sample budget format is given in Annex 19.1. This is a budget pointing to the major categories which are
usually included in all budget preparations.
The type of budget format to be used may vary depending upon whether it is to be approved by your own
organization or submitted to a donor organization for funding. Most donor organizations have their own
special project forms including a budget format. If you intend to seek donor support it is advisable to write
to the potential funding organization as early as possible in the project formulation.
The budget summary should give a clear overview of the distribution of the costs of the project.
Advice on Budget Preparation
1.
Keep in mind the tendency to underestimate time needed for completing project tasks in “the real
world”. Compensate by somewhat overestimating the time and costs for various activities, while
not inflating the budget too much.
2.
Do not box yourself in tootightly with very detailed categories and amounts, especially if regulations
do not allow adjustments later. Press the supervising agency to agree that there may be some
transfer between “line items” in the budget, if needed.
3.
If your government or department has agreed, as part of the project plan, to contribute a certain
amount, try to arrange that the contribution can be accounted for separately, so that the
administrators remain aware of the committment. This may also ensure easier access to the funds.
4.
If the budget is for a period longer than a year, build in allowances for inflation before the project
begins and in subsequent years by increasing costs by a set percentage.
Budget Justification
It is not sufficient to present a budget without explanation.
The budget justification follows the budget as an explanatory note justifying briefly, and in the context of the
proposal, why the money is required. This is always necessary for additional finance requested or for items
costing a large amount of money, and for travel allowances.
The budget justification can make it easier for a proposal to be approved without cuts in important items.
167
How can budgets be reduced?
Explore whether other health-related institutions are willing to second personnel to the project.
When possible, use local rather than outside personnel. If consultants are needed at the beginning,
train local personnel as soon as possible to take over their work.
Explore the possible use of students or community volunteers where appropriate for minimal or no
pay.
Plan for strict control of project expenditures, such as those for vehicle use, supplies, etc.
GROUPWORK
(2 1/2 hours)
1.
Preparea budget for your project. Keep in mind the importance of having a realistic
budget, for which resources can actually be found. (See Annex 19.1 for an
example)
2.
Examine the work plan in your project proposal and consider costs involved in
each component.
3.
Indicate for each item, unit cost as well as number of units. Justify large budget
items, travel and allowances.
4.
Consider the “cost-effectiveness” of various budget levels.
obtained worth the expense?
5.
Consider the budget level that possible funding authorities would consider
appropriate:
Are the results
Examine their guidelines;
If appropriate, talk with donor representatives about their policies, if
appropriate.
6.
If an outside donor is requested for additional funding, make clear what contribu
tion the Ministry of Health/your own institution is making.
168
STEPS IN THE DEVELOPMENT OF A HEALTH SYSTEMS RESEARCH
PROPOSAL:
Questions you
must ask
Steps you will
take
What is the problem
or what is to be
studied?
statement of the
problem
What information is
already available?
literature review
Important elements
of each step
i—
Why do we want to
carry out the re
search? What do we
hope to achieve?
formulation of
objectives
- literature and
other available
information
I— research
questions or
- hypotheses
----- rr- variables
What additional
data do we need to
answer our research
questions? How are
we going to collect
this information?
research
methodology
- type of study
- data collection
techniques
- sampling
- plan for data
collection
- plan for data
analysis and
interpretation
How will the
results be used?
plan for
utilization and
dissemination of
results
- ethical
considerations
L- pilot study
or pre-test
r— manpower
timetable
Who will do what,
and when?
What resources do
we need to carry
out the study, and
what resources do
we have?
How will we present
our proposals to
authorities or
potential funding
sources?
work plan
resources required
and budget
- administration,
monitoring and
- evaluation
P material support,
and equipment
l-
summary of proposal
money
Module 20 - FINALIZING THE RESEARCH PROPOSAL
OBJECTIVES
At the end of the session you should be able to:
1.
Write a brief summary of the completed research proposal
2.
Finalize the research proposal for presentation to the relevant authorities
WHO/AFRO/SHDS
Joint HSR Project
170
Writing a Summary of the Research Proposal
When you have completed writing your research proposal, there is usually a need for the protocol to be
reviewed by senior authorities and policy makers orfunding agencies. Forthe purpose of obtaining approval
from policy makers or very busy administrators it is advisable to add a summary (of no more than two pages)
to the proposal.
A summary usually includes:
Title of the research proposal
Duration: (dates of onset and completion of the project)
Total budget: (in local currency and US$)
Contribution of Ministry of Health
Contribution of donor
Research team: (names and positions)
Principal Investigator
Co-investigators
Principal administrator
Content of study
A brief narrative summary of one page would contain the following elements;
One paragraph on the statement of the problem
General objective
Study population and data collection techniques
Indications of how the findings will be used
171
Finalizing the Research Proposal
At this stage you need to draw all your information together and prepare the final proposal
This task takes time, but it is very rewarding. You have to look at the text with the eyes of an outsider :
Ensure that each paragraph and each sentence is clear
Read each section carefully in sequence, check for “overlaps” and ensure one section logically
leads to the next
In particular the background information and statement of the problem may need revision. These are the
eye catchers of your proposal, and though the statement of the problem should be brief and to the point,
it should also read nicely. Read Module 4 once again and critically judge whether what you have written
meets the guidelines.
Note:
It is advisable to work in small groups of one or two persons, each with responsibility for
one or more sections. Two persons may be made responsible for final editing. They should
ensure a smooth flow going from one section to the next. The whole group, including the
facilitator, should have read all parts, before the final script is presented in for typing.
Finally a table of content should be added to the proposal. It is usually presented after the summary. We
suggest the following as a guideline:
172
Table of Content
1.
Background information
2.
Statement of the problem
3.
Literature review
4.
Objectives
5.
Methodology
5.1
5.2
5.3
5.4
5.5
5.6
Variables, study type, data collection techniques
Sampling
Plan for data collection
Plan for data processing and analysis
Ethical considerations
Pretest
6.
Plan for dissemination and utilization of results
7.
Plan for administration and monitoring
8.
Work plan (including project staff)
9.
Budget
Annex 1.
References
2.
List of abbreviations (if applicable)
3.
Questionnaires (if applicable)
GROUPWORK
(8 hours)
1.
Prepare a summary of your research proposal.
2.
Finalize your research proposal.
173
Annex 5.1
Module 5
CURRENT HEALTH
INFORMATION
ZIMBABWE
Volume 2 Number 2
April - June 1988
Jelliffe-D-B, Jelliffe-E-F.
Nutritional assessment in early childhood (newer
perspectives)
Papua-New-Guinea-Med-J 1987, V: 30 (2) P: 91-6
•n-R-C, Reddy-V, Naidu-A-N
Milton-R-C
Mild vitamin A deficiency and childhood morbidity-an Indian experience.
Am-J-Clin-Nutr 1987, V: 46 (5), p: 827-9
ABSTRACT: Over 1500 preschool urban Indian children
were followed weekly for morbidity from 12 to 18mo.
Examination for mild xerophthalmia (Bitot’s spots
and night blindness) was done initially and at 6 and
12 mo.
Children with mild xerophthalmia at the
start of a 6-mo interval developed respiratory
disease in the interval twice as often as children
with normal eyes at the start of the interval.
No
association was found between mild xerophthalmia and
incidence of diarrhea.
Author.
Narang-A.
Planning follow-up of high risk neonates
Indian-Pediatr 1987, V: 24 (1), P: 9-13
Seaman -J, Poore-P
Good intentions, unfortunate consequences (letter)
Lancet 1987, V: 2 (8571), P: 1334
Sen-B, Wilkinson-G, Mari-J-J.
A
Psychiatric morbidity in primary health care,
two-stage screening procedure in developing
countries:
choice of instruments and cost
effectiveness .
Br-J-Psychiatry 1987, V: 151, P: 33-8
ABSTRACT:
The use in two developing countries of a
two stage-stage psychiatric screening procedure with
an emphasis on minor morbidity is described.
This
was the first use of the Self-Reporting
Questionnaire and Clinical Interview Schedule in
conjunction. Health auxiliaries with limited
training administered the first-stage screening
instrument.
In comparison with other psychiatric
screening techniques used in developing countries,
the procedure described may prove to be more costeffective in primary care and community surveys, and
more appropriate for assessing the wide range of
minor psychiatric morbidity commonly encountered in
such settings.
Author.
174
Module 5
MAR 1988
UTILIZATION
Swing-beds: current experience and future directions.
Richardson H, et al. Health Aff (Millwood) 1987 Fall;
6(3):61-74
NURSING, ONCOLOGIC see ONCOLOGIC NURSING
Annex 5•2
INDEX MEDICUS
NURSING, SUPERVISORY
STANDARDS
Head nurses: their practice and education. Hodges LC, et
al. J Nurs Adm 1987 Dec; 17(12):39-44
NURSING PROCESS
NURSING THEORY
A nursing information system. The outcome of implementing
nursing diagnoses. Mehmert PA. Nun Clin North Am 1987
Dec;22(4):943-53
The impact of nursing diagnoses in a long-term care setting.
Miller J, et al. Nurs Clin North Am 1987 Dec;22(4):9O5-15
The implementation of nursing diagnoses. The Iowa Veterans
Home experience. Specht J, et al. Nun Clin North Am
1987 Dec;22(4):917-33
NUTRITION
METHODS
[Clinical effect of primary nursing and functional nursing]
Mei YC. Chung Hua Hu Li Tsa Chih 1987 Aug;22(8):337-8
(Chi)
STANDARDS
Intershift report: oral communication using the nursing
process. Kilpack V, et al. J Neuroad Nurs 1987 Oct;
19(5):266-70
NURSING RECORDS
[Opinions on writing a nursing history of the patient treated
by Chinese traditional medicine] Liu HH.
Chang Hua Hu Li Tsa Chib 1987 Aug;22(8):355-6 (Chi)
STANDARDS
A new dimension in documentation: the PIE method.
Buckley-Womack C, et al. J Neurosci Nun 1987 Oct;
19(5):256-60
Intershift report: oral communication using the nursing
process. Kilpack V, et al. J Neuroad Nun 1987 Oct;
19(5):266-70
NURSING SERVICE, HOSPITAL
ORGANIZATION A ADMINISTRATION
Gambling away absenteeism. Curran MA, et al.
J Nurs Adm 1987 Dec;17(12):28-31
Outcomes of shared governance on staff nurses. Howard DC.
J Nurs Adm 1987 Dec;17(12):9
Implementing nursing diagnoses. Maibusch RM.
Nurs Clin North Am 1987 Dec;22(4):955-69
The impact of nursing diagnoses in a long-term care setting.
Miller J, et al. Nurs Clin North Am 1987 Dec;22(4):905-15
[Experience in improving the management of the work of
nurses] Tsvang RL. Sot Zdravookhr 1987;(8):35-4O (Eng.
Abstr.)
(Rus)
STANDARDS
Unit-based approach for nursing quality assurance in The
Netherlands: one year experience. Giebing HA.
Aust Clin Rer 1987 Mar,7(24):28-32
NURSING SERVICES
ORGANIZATION A ADMINISTRATION
Modeling: a method for program development Budgen CM.
J Nurs Adm 1987 Dec; 17(12): 19-25
Organizational characteristics that facilitate the use of nursing
diagnoses. Maas ML. Nurs Clin North Am 1987 Dec;
22(4):881-6
NURSING STAFF, HOSPITAL
LEGISLATION A JURISPRUDENCE
When hospitals don \ listen to nurses' complaints. Cushing
M. Am J Nurs 1987 Dec;87( 12): 1547-8, 1550
ORGANIZATION A ADMINISTRATION
Gambling away absenteeism. Curran MA, et al.
J Non Adm 1987* Dec; 17(12):28-31
PSYCHOLOGY
Determining social and therapeutic rules for psychiatric
inpatients. Morrison EF. Hosp Community Psychiatry 1987
Sep;38(9):994-5
Prognostic conflict in life-and-death decisions: the
organization as an ecology of knowledge. Anspach RR.
J Health Soc Behav 1987 Sep;28(3):215-31
Outcomes of shared governance on staff nurses. Howard DC.
J Nurs Adm 1987 Dec;17(12):9
SUPPLY A DISTRIBUTION
Will the nursing shortage lead to liability suits? Boston CM,
et al. Hospitals 1987 Nov 2O;61(22):64-8
Statewide plan for nurses. Triolo PK. Iowa Med 1987 Oct;
77(IO):5l7-9
[Exploration of the thinking mode of nursing] Dang JP.
Chung Hua Hu LI Tsa Chili 1987 Aug;22(8):369-7O
(Chi)
see related
DIET
FOOD
The nutrient intakes of pregnant and lactating mothers of
good socio-economic status in Cambridge, UK: some
implications for recommended daily allowances of minor
nutrients. Black AE, et al. Br J Nutr 1986 Jul;56(1):59-72
Nutrition and cellular immunity in hospital patients. Dowd
PS, et al. Br J Nutr 1986 May;55(3):515-27
The distribution of nutrient intake within families. Nelson
M. Br J Nutr 1986 Mar;55(2):267-77
Nutrition knowledge of female varsity athletes and university
students. Barr SI. J Am Diet Asaoc 1987 Dec;87( 12): 1660-4
Nutritional adequacy of diet and supplements in patients with
rheumatoid arthritis who take medications. Bigaouette J,
et al. J Am Diet Assoc 1987 Dec;87( 12): 1687-8
Nutritional intake from food sources of high school football
athletes. Hickson JF Jr, et al. J Am Diet Assoc 1987 Dec;
87(12):1656-9
USDA software must be used onsite [letter] Kerbs-Smith
J. J Am Diet Assoc 1987 Dec;87(12):1709
The 1990 National Nutrition Objectives: lessons for the
future. Miller SA, et al. J Am Diet Assoc 1987 Dec;
87(12):1665-7
Use and misuse of balance studies. Mertz W. J Nutr 1987
Nov;117(11):1811-3 (12 ref.)
Analysis of nutrient-response rehtionships. Schulz AR.
J Nutr 1987 Nov; 117(11): 1950-8 (33 ref.)
Nutrition and health—an individual responsibility. Dodds
JM. Public Health Rep 1987 Jul-Aug;Suppl:29-33
Cultural issues in maternal and child health and nutrition.
Pelto GH. Soc Sd Med 1987;25(6):553-9 (50 ref.)
[Incidence of lactase deficiency in patients with involution
osteoporosis and in normal subjects. Its effect on the
nutritional intake of calcium and phosphorus] Jodry H, et
al. Schweiz Med Wochenschr 1987 Oct 31; 117(44): 1736-41
(Eng. Abstr.)
(Fre)
[Tell me what you eat... (interview by Robert Henry)]
Mirouze J. Union Med Can 1987 Oct;l 16(5):27O-5 (Fre)
EDUCATION
Nutrition and the podiatric physician. Results of a survey.
Connell CR 4th, et al. J Am Podiatr Med Assoc 1987 Oct;
77(10):570-5
Nutritional information and misinformation [letter] Walker
AR, et al. Med J Aust 1987 Nov 16;147(10):521
NUTRITION DISORDERS
BLOOD
Malnutrition during the reproductive cycle: effects on
galactopoietic hormones and lactational performance in the
rat. Kliewer RL, et al. Am J Clin Nutr 1987 Dec;
46(6):926-35
COMPUCATIONS
Malnutrition as a risk factor for severe visceral leishmaniasis.
Cerf BJ, et al. J Infect Dis 1987 Dec; 156(6): 1030-3
[Comparison of prognostic nutrition indices in preoperative
detection of risk patients. A prospective trial] Brenner U,
et al. InfusioDsther Klin Emahr 1987 Oct;14(5):215-21 (Eng.
Abstr.)
(Ger)
DIAGNOSIS
Detection of malnutrition [letter] Kapil U, et al.
Indiaa Pediatr 1987 Apr;24<4):352-4
DIET THERAPY
[Feeding soy meal and rice cereal to malnourished infants
during convalescence] Visquez-Garibay EM, et al.
Bol Med Hosp Infant Mex 1987 Oct;44(10):603-9 (Eng.
Abstr.)
(Spa)
METABOLISM
Cardiac sensitivity and responsiveness to beta-adrenergic
stimulation in experimental cancer and undemutrition.
Drott C, et al. J Mol Cell Cardiol 1987 Jul;19(7):67S-83
OCCURRENCE
The Fifth World Food Survey: an assessment of food supplies
and malnutrition. Dichter CR. J Am Diet Aaeoc 1987 Dec;
87(12): 1668-72
175
Module 5
DATASTAR
QUERY 0569
Annex 5.3
MEDX: MEDLINE FEBRUARY 1988
08.01.88
PAGE
2
1
AN 88048427.
AU Ormerod-L-P.
TI Reduced incidence of tuberculosis by prophylactic chemotherapy in
subjects showing strong reactions to tuberculin testing.
$0 Arch-Dis-Child 1987 Oct, VOL: 62 (10), P: 1005-8, ISSN: 0003-9888.
AB The introduction of chemoprophylaxis with rifampicin and isoniazid in
1981 significantly reduced the incidence of tuberculosis. Between
1978 and 1981 children accounted for 136 of>642 notified cases, and
this was reduced to 55 of 418 between 1982 and 1986. Jhis effect was
obvious among children from the Indian subcontinent* who
nost
comprised 80.2% of the children treated. The incidence among white
children was not affected. Few side effects occurred and only two of
339 (0.6%) later developed clinical tuberculosis. Chemoprophylaxis
plays an important part in the management of tuberculosis in a
district with a high incidence of the disease. Author.
2
AN 88048110.
AU Mohandas-V, Unni-J t Mathew-M, Steinhoff-M-C, Rajan-D-P, Mathan-M,
Pereira-S-M.
TI Aetiology and clinical features of acute childhood diarrhoea in an
outpatient clinic in Vellore, India.
SO Ann-Trop-Paediatr 1987 Sep, VOL: 7 (3), P: 16?-72, ISSN: 0272-4939.
AB. The aetiology of acute diarrhoea was investigated in 245 children
less than 3 years old in a paediatric outpatient clinic in southern
India, In 55% of the children organisms were found in the stools, and
quarter
were
infected with multiple organisms. Viruses,
one
E.
coli,
shigella and Campylobacter species
enteropathogenic
accounted for 75% of all isolates. Cholera and helminthiasis were
rare, and no child had amoebiasis. Clinical findings were useful in
the diagnosis of shigella and rotavirus infection only. Children with
shigellosis had classical dysentery, and a greater number of stools!
they were unlikely to be breast-fed, to be less than 6 months old, or
to have watery stools. Rotavirus infections were characterized by
vomiting. Only 20% of the 245 children had an infection which could
be treated effectively with antimicrobials. Author.
3
AH 88046629.
AU Milton-R-C, Reddy-Y, Naidu-A-N.
TI Mild
vitamin
A deficiency and childhood morbidity—an Indian
experience.
SO Aa-J-Clin-Nutr 1987 Nov, VOL: 46 (5), P: 827-9, ISSN: 0002-9165.
AB Over 1500 preschool urban Indian children were followed weekly for
morbidity from 12 to 18 mo. Examination for mild xerophthalmia
(Bitot’s spots and night blindness) was done initially and at 6 and
12 mo. Children with mild xerophthalmia at the start of a 6-mo
interval developed respiratory disease in the interval twice as often
as children with normal eyes at the start of the interval. No
association was found between mild xerophthalmia and incidence of
diarrhea. Author.
A
176
Module 5
Annex 5.4
Example of an Index Card
Hassouna, W.A., "Solving Peoples' Problems", World Health,
pp. 26-29.
WHO, Geneva, April, 1980,
— This article discusses health services research (HSR) as a relatively new area of
investigation (1960).
-- This method of research permits the health team and the community to study critical
problems, while econanizing on time and money. Jrportant to try to’collaborate
with service administrators.
— If HSR to be effective, must be done so results available in time
to solve problems it addresses—change in health status, not publication, most
important result of research.
— Example of HSR study in Maruit (Egypt) :
- In 2 days a multidisciplinary team (25 members) v^re able to identify the
critical problems affecting health and health care in area.
- Various aspects of the study are discussed.
- The study results are stated clearly and the role of the traditional healer
identified.
- Among major findings was that "the formal providers of health services were not
giving the people the service they required at the time they needed it, at a cost
they could pay, and in a manner acceptable to the people." (p. 27)
The reverse side of the index card appears belcw:
— Points that are emphasized in the article:
- There's little correlation between size and quality of health services available
to population and health status of population (p. 28) Problem is present nature
of medical technology.
- Use of med. technology to inprove health status would be more successful if
became integral part of socio-cultural and ec. behavioural change process, (p. 28).
- Article lists characteristics and advantages of PHC and role of camtunity in it.
- Discusses importance of HRS related to PHC - conviction HRS should form core
of WHO "Health for All by the Year 2000" strategy.
- Important to involve WIO staff in field activities so acquire practical
understanding of health service realities.
— Observations:
Good reference article on applied research, PHC, and research training.
WHO/AFRO-SHDS
177
Annex 9.1
Module 9
SELECTION OF STUDY TYPE
Experimental Study
The classical experimental study has three characteristics:
MANIPULATION - The experimenter does something to one group of subjects in the study.
CONTROL - the researcher has introduced one or more control group(s) to compare with the ex
perimental group.
RANDOMIZATION - the researcher has taken care to assign subjects to the control and experimen
tal groups on a random basis.
Study population
(Sampling)
Sample population
(Randomization)
Control (comparison) group
Experimental (study) group
I
I
First data collection (Same
time as in study group)
First data collection
(Before intervention)
I
I
No manipulation or
intervention
Period of intervention/
manipulation
I
I
Last data collection
(Same time as in study group)
Last data collection
(After intervention)
In a QUASI-EXPERIMENTAL STUDY one characteristic is missing, either randomization or use of
a control group. The quasi- experimental study, however, always includes the manipulation of an
independent variable which serves as the intervention.
Advantages and disadvantages of an experimental study
Advantage:
It is the most powerful method of testing the cause-and-effect relationship between
variables.
178
Disadvantages:
Certain variables can not be manipulated in this way.
In certain situations ethical issues are raised by proposed interventions or treatments.
Example of an experimental study
A researcher plans to study the effects of a new drug. (The drug has already been tested
extensively and approved fortrial use). He plans to use 300 patients as subjects who are
currently receiving the standard treatment fot the same condition for which the new drug
base been designed. He describes the study and new drug to the patients and their
spouses, asking their consent to be in a “double-blind” study in which the patients will be
divided into 2 groups. On a random basis, one group will receive the experimental drug,
while the other group receives the standard treatment. He plans that the medications will
be disguised and labelled so that neither the research assistant administering the drugs nor
the patients receiving them will know which drug is used.
Example of a quasi-experimental study
A researcher plans to study the effects of health education on the level of participation of
a village population in an immunization campaign. She decides to select a village in which
to give health education sessions concerning immunization and another village she
believes to be similar which will serve as a control. The immunization campaign will then
be carried out in the same manner in both villages. A coverage survey will then be
undertaken to determine if the level of coverage in the village where health education was
introduced before the campaign, is significantly different from the level of coverage in the
“control” village which received no health education. (Note: The study is quasiexperimental because the subjects were not assigned to the control or experimental
groups on a random basis).
179
Annex 10.1
Module 10
DEVELOPMENT OF DATA COLLECTION TECHNIQUES
Focused Group Discussion
A focused group discussion consists of a group discussion in which a small number of participants
(usually six to ten) of similar status under the guidance of a moderator, and with at least one
additional observer, talk about topics of relevance to the particular study. The informal group
situation and open-ended nature of the questions are intended to encourage participants to
comment on behaviour and elaborate opinions to an extent that is more difficult in more formal
individual interview situations. Interactions among participants stimulate discussion.
This group dynamic distinguishes focused discussions from more conventional individual in-depth
interviews.
The purpose of focused group discussions is not to provide statistically generalizable quantitative
data but ratherto obtain qualitative information exposing underlying attitudes, opinions and behavi
our patterns. It is complementary rather than an alternative to a sample survey approach.
Participants in a session are usually selected to be roughly of equal socio-economic status. Par
ticipants should perceive each other as sharing the same perspective in order to facilitate discus
sions. If one wants to discuss changes for example with regard to nutrition of young children, one
may also invite women of different age groups, as long as one takes care that each group with whom
one discusses is roughly of the same composition.
Separate sessions on the same topic with a number of groups selected according to different sets
of criteria would permit an indepth exploration of complementary or conflicting views, for example,
on the widening gap between rich and poor in a certain area.
The limitation of the focused group discussion is that in a group one participant may dominate the
group discussion.
WHO/AFRO/SHDS
Joint HSR Project
180
Annex 10.2
Module 10
MATERNAL RECORD
1
Name of mother.
Name of TBA—
Age of mother—
Date of delivery.
Parity
Number
1
a,
o
Presentation
of baby
s
o
o
o
pY-—|1
Abnormal loss
Normal loss
Blood loss
of mother
during and
after delivery
Normal
baby
Low weight
baby
j
Good
Sick
Dead at
delivery
Condition of
baby
at delivery
Dead
Condition of
mother after
delivery
No referral
Referral to
hospital
Good
Sick
Dead
1. This questionnaire which was made for use by non-literate
health workers was shared by Dr. Peter Lamptey.
WHO/AFRO-SHDS
181
Annex 11.1
Module 11
How to Use Random Number Tables*
1.
First, decide how big a number you need. Next, count if it is a one, two or more digit number. For
example if your sampling frame consists of 10 units, you must choose from numbers 1-10,
(inclusive). You must use two digits to ensure that 10 has an equal chance of being included.
You also use two digits for a sampling frame consisting of 0-99 units.
If however, your sampling frame has 0-999 units, then you obviously need to choose from three
digits. In this case, you take an extra digit from the table to make up the required three digits. For
example, the number in columns 10,11, row 27: 43, would become 431; going down, the next
numbers would be 107, 365 etc.
You would do the same if you needed a four digit number, for a sampling frame 0-9999 units. In
our example of the number on columnsW, 11,12, row 27 of the table: 431, this would now become
4316, the next down 1075, and so on.
>,down the page .across
2.
Decide before-hand whether you are going to go across the page R
the page L <-—, or up the page ~
3.
Without looking at the table, and using a pencil, pen, stick, or even your finger, pin-point a number.
4.
If this number is within the range you need, you take it. If not, you continue to the next number along
the direction you chose before-hand, (across or up ordown the page), until you get a number that
is within the range you need.
For example if you need a number between 0-50 and you began at column 21,22, row 21 you get
74 which is obviously too big. So you could go down (having decided whether to go up/down or
across, before-hand) to 97, also too big, to 42, which is acceptable, and which you take.
The random number table on the following page has been taken from Hill AB. A short
textbook of medical statistics. London: Hodder and Stoughton, 1977 : 306-7
182
RANDOM SAMPLING NUMBERS
1
2
3
4
5
6
7
8
9
10
11
12
co
GO
12 3 4
5 6 7 8
9 10 11 12
13 14 15 16
8 0 9 4
3 5 6 3
2 5 2 5
8
8
3
2
5
2
2
7
2
5
9
9
7
3
2
5
4
13 3 0
3 5 6 6
2 10 8
6 3 3 1
0 0 4 6
7 8 5.0
5 9 2 6
4 4 9 0
6 5 4 5
3 8 2 6
4 6 6 5
6 4 9 8
5 4 7 7
5 9 9 5
4 8 2 0
7 5 19
7
0
6 7 2 2
9 7 4 8
9 8 6 9
5 9 8 2
9
5
1
8
9 10 4
13
14
15
5 6 4 1
14 3 7
7 4 4 4
8 2 7 9
9 2 0 0
3 0 7 9
16
0 16 1
17
7 3 8 8
7 8 3 0
9 8 8 7
12 6 1
7 6 17
9 7 5 9
4 7 14
4 2 4 6
2 5 9 2
21
3 9 4 7
4 9 3 7
22
23
24
25
4 5 5 0
114 4
1
4
1
7
3
3
1
4
8
6
0
5
6
6
5
8
5
7
1
6
8 10 3
7 7 10
9 6 9 7
5 8 2 3
9 9 0 L
26
6 9 5 9
27
7 9 4 1
28
2 2 8 4
9 5 ^4
18
19
20
29
30
8 9 7 6
4 6 13
4
7
1
1
3
3
1
9
9
6
8
7
8
5
4
4
5
4
8
2
1
1
5
7
6
1
1
4
7
2
5
9
2
6
4
4
9
3
5
5
3
2
7
3
0
6
4
3
1
1
6
1
5
8
9
0
7
8
7
3
9
5
8
1
1
1
5
9
0
2
7
2
7
5
3
8
7
4
8
7
7
2
3
8
4
4
5
5
2
3
6
9
8
2
6
9
2
8
6
2
3
4
9
3 . 4
5 0
8 3
8 7
7 8
6 0 0 8
8
4
6
2
2 3 2 2
0 8 9 6
7 4 3 6
8 5 0 9
2
4
1
3
0
1
7
3
6
5
3
6
9
9
9
6
6
1
3
2
8
5
1
4
2
3
5
3
1
3
1
7
4
9
5
0
6
0
8
5
3
9
4
2
4
0
7
5
7
3
4
6
0
3
2
6
5
6
3
2
7
5
0
2
8
0
4
4
0
2
2
2
5
8
6
2
4
2
2
2
17 18 19 20
21 22 23 24
25 26 27 28
3
3
6
2
8
8
2
7
8
0
0
1
5
1
2
4
6
6
0
5
1
2
4
5
9
4
5
5 < .
5
0
9
5
2
6
8
0 1
3 7
2 3
4
3
9
4
9
8
7
6
2
7
8
5
5
8
5
8
6
8
3
9
9
2
9
9
4
0
8
4
3
8
0
1
8
3
1
3
3
3
9
6
1
0
6
0
8
5
3
8
4,
9
2
9
6
0
0
2
7
7
0
4
7
0
7
8
4
0
3
1
6
7
9
4
9
9
4
7
2
5
8
0
6
0
5
8
2
3
3
5
9
5
4
7
7
2
8
4
2
4
6
7
5
7
9
9
6
7
1
3
1
3
9
9
8 ‘
3
3
2
7
3
3
5
8
9
8
9
8
4
9
9
7
0
3
3
1
7
4
0
9
6
1
2
9
6
3
9
2
1
3
3
5
0
7
1
1
8
3
3
9
1
6
9
8
1
6
4
1
5
2 4
9
7
9
3
1
1
8
1
8
0
1
5
4
9
3
9
9
7
3
4
6
2
4
5
8
1
7
9
6
1
6
8
1
8
8
9
2
3
8
3
7
7
8
3
0
8
7
0
3
4
3
8 ‘7
4
8
2
6
7
2
7
1
1
1
0
9
2
9
3
8
9
7
9
9
5
5
0
6
8
9
3
0
4
1
6
0
2
9
2
2
4
0
1
0
7
5
7 •
9
2
6
0
5
8
7
9
4
8
0
0
1
9
9
1
9
9
6
5
2
5
9
3
9
5
3
9
4
2
8
4
5
3
0 /
9
1
4
8
4
8
7
0
7
4
9
0
9
7
5
9
9
0
8
4
5
9
5
8
0
0
3
9
3
1
3
3
9
4
2
9
6
0
2
9
2
0
0
4
7
5
7
3
8
4
9
6
5
3
0
7
8
3
7
4
7
5
5
7
2
6
9
2
5
0
8
5
1
6
5
4
8
7
6
1
9
1
8
8
5
8
9
8
0
4
9
1
0
7
5
8
6
7
4
3
2
4
3
3
1
5
1
29 30 31 32
8
5
4
3
1
9
6
1
5
6
4
8
1
2
1
1
0
8
5
1
6
5
9
2
2
6
6
8
1
1
6
2
1
Annex 11.2
Module 11
FORMULAS FOR CALCULATING SAMPLE SIZE*
The formulas for calculating required sample size are divided in two groups:
1.
For studies aiming to estimate an amount (quantity) of interest with a previously specific precision.
2.
For studies aiming to demonstrate a significant difference.
1.
PRECISION
Formula for minimum
sample size
s2
1.1
Single mean
sample size
n
n
e2
s
e
standard
deviation
required size
of standard
error
r
1.2
Single rate
r
rate
n
e2
n, e
as in 1.1
P
proportion
p(1-p)
1.3
Single proportion
n
e2
1.4
Difference between
two means (sample
size in each
group)
n, e
as in 1.1
Si,s2
standard
deviations
as in 1.1
n, e
S!2 + s22
n
e2
ri + r2
1.5
Difference between
two rates (sample
size in each
group)
ri>r2
n, e
rates
as in 1.1
184
n
e2
p/l-p,) + p2(1-p2)
1.6
Difference between
two proportions
(sample size in
each group)
Pr P2
n, e
n
proportions
as in 1.1
e2
2. SIGNIFICANT RESULT
In the formulas below, u and v have the following meaning:
u
one-sided percentage point of the normal distribution, corresponding to 100% - the power. The
power is the probability of finding a significant result. Eg, if the power is 75%, u = 0.67
v
percentage point of the normal distribution, corresponding to the (two-sided) significance level. Eg,
if sign, level is 5%, v = 1.96
(u+v)2(s12 + s22)
2.1
Comparison of two
means (sample size
of each group)
n
n
sample size
mv m2 means
standard
svs2
deviations
see above
u, v
(m1 - m2)2
u + v)2^ +
2.2
2.3
Comparison of two
rates (sample size
of each group)
rrr2
n
u, v
rates
see 2.2
see above
Comparison of two
proportions
(sample size of
each group)
Pi»P2
n
u, v
proportions
see 2.2
see above
n
(I-, - l-s)2
(u + vHp/l-pJ + p2(1-p2)}
n
(P, - P2)2
★
Modified from Kirkwood B. Essentials of Medical Statistics. Oxford: Blackwell Scientific
Publications, 1988.
185
Module 11
Annex 11.3
EXPLANATION ON SAMPLE SIZE CALCULATIONS GIVEN IN THE
TEXT OF MODULE 11
1.
p.7
Prevalence of Goitre
District A:
proportion is 1 % = 0.01 ,as the 95% confidence interval is the proportion
+ 2 x standard error, the standard error is 0.25% = 0.0025.
Formula used:
n
P(1-P)
(formula no. 1.3 in annex 11.2)
e2
0.01 xO.99
n
= 1600
(0.0025)2
District B:
proportion is 0.1% = 0.001
Standard error is 0.025% = 0.00025
0.001 x 0.999
16000
n
(0.00025)2
2.
p8
Prevalence of malnutrition
Country A:
p, is 10%
0.1
p2 is 30% = 0.3
probability of significant result: 75%
Formula used:
(u + V^fp^l-p,) + P2(1-P2)}
(formula nr 2.3 in
annex 11.2)
n
(PrP2)2
(1.96 + 0.67)2(0.1 x 0.9 + 0.3 X 0.7)
52,say 60
n
(0.1 -0.3)2
186
Country B:
p1 is 10%
0.1
p2 is 15% = 0.15
probability of significant result 75%
(u + v)2{p,(1-p,) + p2(1-p2)}
n
(PrP2)2
(1.96 + 0.67)2(0.1 x09 + 0.15x0.85)
= 602, say 600
n
(0.1 -0.15)2
Country B:
p, is 10% = 0.1
p2 is 15% = 0.15
probability of significant result 90%
(u+v)2{p,(1-p,) + P2(1-P2)}
n
(PrP2)z
(1.96 + 1.28)2(0.1 xO.9 + 0.15x0.85)
= 913, say 900
n =
(0.1 -0.15)2
187
Annex 15.1
Module 15
SUMMARY OF POINTS TO BE CHECKED
DURING A PRE-TEST OR PILOT STUDY
The following points may
be checked during your
pre-test:
Acceptable
1. Reactions of respondents to your research procedures
Availability of sample needed for full
study
Work schedules of population that
may affect their availability
Desire of population to participate
Accebility of questions
Clarity of the language used
2. The data collection tools
Are the data being collected providing
the information you need and are they
reliable?
The time needed for administering
each one of the data collection tools
Presentation and format of question
naire
Pre-coding of questions
Accuracy of translation (adapted to
the local language)
188
Not
Acceptable
Suggestions
Acceptable
Organization of content
Handling and administering the tools
3. Sampling procedures
Are instructions to obtain the sample
used uniformly by all staff?
How much time is needed to locate
the individuals to be included in the
study?
4. Preparation and effectiveness of research team
Adequacy of staff training
Output of each team member
Team dynamics
Reliability of tools when administered
by different team members
Accuracy of interpretation
Appropriateness of plan for supervi
sion
189
Not
Acceptable
Suggestions
Acceptable
5. Procedures for data processing and analysis
Use of data master sheets
Effectiveness of data quality control
Appropriateness of statistical proce
dures
Ease of data interpretation
6. Schedule for research activities
Amount of time allowed for:
- field trips for data collection
- supervision
- administration
- analysis of data
Sequence of activities
190
Not
Acceptable
Suggestions
Module 16
Annex 16.1
OUTLINE OF A SAMPLE RESEARCH REPORT
Presented below is a sample report outline. The format may not be appropriate for every research report
and should be viewed only as a sample framework.
Project Identification
Title
Organization and implementation of the project
Principal investigator
Sponsoring agency
Project timeframe (beginning and completion dates)
Budget (planned and used)
Summary (2-4 pages)
Why the research was conducted
The research objectives and methodology
The major contraints of the study
The major findings
Recommmendations for action
Introduction and Statement of the Problem
Statement of the problem
Background of the problem
Importance of the study
Prior research and information on the subject
(This might be a separate chapter, if literature review is a major part of the study).
Outline of the rest of the report
Objectives
Purpose and objectives of the research
Questions to be answered by the research
Assumptions
Hypotheses (if any)
191
Methodology
Methodology used in the research:
Major variables
Data collection techniques used and their limitations (present questionnaires, if applicable,
in annex)
Sampling methods and sample size
Data analysis
Research project staffing, administration
Work plan (original and adapted)
Plan for and outcome of monitoring and evaluation
Findings
Findings
Discussion of relevance and importance of findings
Recommendations for action
Specific recommendation stemming from the study
Indications of areas which require further study
Suggestions concerning specific steps to be taken in order to implement recommendations
Plans for dissemination of findings (publication, etc.).
Bibliography
Appendices
Technical information
References
192
Annex 16.2
Module 16
STEPS IN PROJECT PLANNING AND IMPLEMENTATION
RELATED TO UTILIZATION AND DISSEMINATION OF RESULTS
STAGE 1: DEVELOPMENT OF RESEARCH PROPOSAL:
Consultation with potential users on selection of topic
Involvement of potential users in project design
Development of tentative plan for dissemination and utilization of results
STAGE 11: IMPLEMENTATION OF STUDY
Maintenance of interest of potential users through reports, workshops
Involvement of pontential users in project implementation through advisory committee, participation in
research (if appropriate)
STAGE III: COMPLETION OF STUDY
Preparation of distribution list for final report
Review of tentative findings with group or community studied
Preparation of Final Report
Gear it towards target audiences
Make it simple and clear
Provide an abstract or summary
Include recommendations for action
Distribution of final Report
Remember those studied and key potential users
Promotion of Utilization of Results
Meet with key decision makers on how to put results into action
Continued Disseminatin of Results
Oral presentations
193
Annex 19.1
Module 19
SAMPLE BUDGET
Utilization of family planning methods in district A
1.
Personnel
MOH
Principal investigator (1)
20% time x 6 months x $1000/month
Per diem 30 days (no nights out)
Donor
Total
$150
$1200
$150
$1200
Supervisor (1)
100% time x 2 months x $400/month
10% time x 5 months x $400/month
Per diem 40 days (no nights out)
$200
$800
$200
$150
$600
$500
$1000
$500
$800
$200
Interviewers (5)
100% time x 1 month x $200/month x 5
Per diem 5 x 20 days (no nights out)
$400
Clerks (2)
40% time x 4 months x $250/month
$400
$400
Local consultants
100% time x 8 x 1 day x $50/day
$400
$400
1. SUBTOTAL
$2200
$2650
$4850
Field work interviewers + supervisor
(interviews! 00 km a day each
using questionnaires)
4 motorcycles +
1 car travelling
for a month
$400
$1200
$1600
Field work principal investigator
& supervisor
(group interviews and
unstructured interviews)
1 car travelling
100 km a day for a
month
$200
$600
$800
$600
$1800
$2400
2.
Transport
2. SUBTOTAL
194
3.
Stationery
Questionnaires, paper & pens
Photocopying
$100
$50
$100
$50
$150
$150
$100
$400
$500
$50
$50
$100
$100
$150
$800
4.SUBTOTAL
$200
$600
$800
GRAND TOTAL
$3000
$5200
$8200
3.SUBTOTAL -
4.
Workshops (excluding wages)
Training before field work (1 week)
Dissemination of results
- community and local government (1 day)
- health workers (1 day)
195
A
predominant in AP even during 1980s. It accounted for about 70 percent of
inpatient care in rural areas and 62 percent in urban areas, which was highest in
the country. By mid 1990s, its share further increased to 77.5 percent in the
rural areas, and a marginal increase in the urban areas. In the provision of
outpatient care, the private sector was predominant in most of the states in
1990s as well as in 1980s. It accounted for about 70 percent of outpatient care
in 1980s and 80 percent during 1990s (Table2). Similarly, the estimates of
National Family Health Surveys on the type of medical attendance at the time
of delivery show growing predominance of private medical facilities in A.P.
(UPS, 1992-93 and 1998-99). At the national level, the per cent of births
taking place at the health institutions (public and private hospitals) had
p;
i
ex
ah
pa
Andhra Pradesh Development
342
■T:-:
1>ri
Table 1. Percentage Distribution of Hospitalised (Inpatient)
_____________ Cases by Type of Hospital_____________
i aht
State
-
<,
■
y
rcid
■
________ Urban_____________
Rural ___________
Private_________ Government_______ Private________ Government
1986-87 1995-96 1986-87 1995-96 1986-87 1995-96 1986-87 1995-96
37.98
36.2
63.8
62.02
29.91
22.5
70.09
77.5
65.2
34.8
82.33
90.02
17.67
73.8
26.2
9.98
34.6
65.4
45.71
54.29
49.86
24.7
50.14
75.3
59.21
36.9
32.1
63.1
48.96
40.79
51.04
67.9
55.31
62.7
37.3
44.69
50.96
30.5
49.04
69.5
48.90
29.8
70.2
51.10
45.8
58.02
54.2
41.98
38.4
61.6
55.65
43.38
40.1
44.35
56.62
59.9
56.0
76.98
44.0
23.02
53.3
46.7
79.23
20.77
31.8
46.23
68.2
31.2
53.77
68.8
43.57
56.43
81.0
81.48
18.52
19.0
90.6
9.4
88.06
11.94
27.6
72.4
48.77
39.4
51.23
60.6
47.49
52.51
85.62
73.1
26.9
14.38
80.01
64.9
19.99
35.1
58.04 , 35.7
41.96
64.3
‘41.1
58.9
56.15
43.85
39.8
60.2
59.25
40.75
47.1
55.37
52.9
44.63
72.1
73.90
26.10
27.9
91.62
82.0
18.0
8.38
43.1
60.26
39.74
56.9
59.74
45.3
40.26
54.7
A.P.
Assam
Bihar
Gujarat
Haryana
Karnataka
Kerala
M.P.
Maharashtra
Orissa
Punjab
Rajasthan
Tamilnadu
U.P.
W. Bengal
All India
Source: 1. NSSO (1992).
2. NSSO (1998).
increased from about 26 percent in 1992-93 to 34 percent in 1998-99
(Table3). In A.P., it increased from about 34 to 54 percent during the same
period. However, the private sector accounted for a high proportion of
hospitalised births in AP. Its share increased rapidly from 58 percent in 199293 to 75 percent in 1998-99. Similar trend was manifested at all India level but
it is less pronounced. The share of private sector increased from 43 to 52
percent.
Size and Nature ofHealthcare System
343
In sum, the national surveys on utilisation of medical facilities indicate
the rapid growth of private sector in the country during 1990s. The private
sector seems to be rampant particularly in AP.
Table 2. Percentage Distribution Of Treatments (Outpatient) Over
__________________ Source Of Treatment_____________________
______ Rural___________________________ Urban_____________
_____ Private_ Government_______ Private_________ Government
_________ 1986-87 1995-96 1986-87 1995-96 1986-87 1995-961986-87 1995-96
A.P.
78.00
20.01
79.99
76.7
23.3
88.5 22.00
11.5
Assam
55.0
53.01
45.0
70.40
46.99
65.1 29.60
34.9
Bihar
82.00
83.13
94.9
16.87
80.9 18.00
5.1
19.1
67.50
Gujarat
18.4
81.44
81.6
32.5
81.8 18.56
18.2
83.10
Haryana
13.4
86.6
16.9
78.31
10.1
89.9 21.69
63.60
Karnataka
68.4
36.4
31.6
68.70
81.4 31.30
18.6
65.48
Kerala
70.1
34.52
63.68
29.9
69.5 36.32
30.5
25.8
M.P.
67.24
74.2
32.76
68.00
80.0 32.00
20.0
Maharashtra
26.32
73.68
88.1
74.98
10.1
11.9
89.9 25.02
Orissa
62.2
37.8
32.4
47.31
52.69
52.13
67.6 47.87
Punjab
93.0
7.0
89.08
92.8 10.92
7.2
87.23
12.77
43.88
Rajasthan
56.12
59.3
40.7
42.49
59.1 57.51
40.9
Tamilnadu
62.96
37.04
30.1
64.93
69.9
24.3
75.7 35.07
95.8
4.2
83.80
U.P.
89.7 16.20
10.3
W. Bengal1
19.41
13.8
80.59
87.1
12.9
76.77
86.2 23.23
18.2
All India _______ 74.42
17.4
81.8
25.58
72.85
82.6 27.15
Source: 1. NSSO (1992).
2. NSSO (1998).
State
Size of Public and Private Health Care Facilities
Occasional surveys of medical facilities also indicate the growing
predominance of private medical care in AP. The GOAP survey of medical
facilities had shown that the private hospitals accounted for 59 percent of total
hospital beds in the state, leaving 35 percent in the public sector and 6 percent
in the voluntary sector (Table 4). However, there were considerable regional
variations in the relative strength of these sectors. The private sector was very
developed in the Coastal districts which are more advanced in the socio
economic development. Within the Coastal region, the backward north coastal
districts (Srikakulam, Vizianagaram and Vishakhapatnam) had a lower percent
of private hospital beds. Simiterly, the developed districts like Nizamabad>Nalgonda, Karimnagar and Khammam in Telangana region had a larger
private sector while the public sector was still predominant in the backward
districts like Adilabad and Mahbubnagar.
■
W— A
I
344
pi
_____ Table 3. Place of Delivery
I : piin
AP/India_______
Public
i "
A.P
1992-93
252
353
605
(%)
13.7
19.2
32.9
396
35.1
49.8
!
I
de
ah*
► c
fc-
<
The
sign
and
add.
the
mak
broa
that
prov
form
for
read
disad
regio
. . -
:-7\.
Total
Home Others Total
No. of
Births
1207
65.7
28
0.15
1838
100
557
49.3
10
0.9
1129
100
494
49369
1.0
100
141
25
(%)
12.5
2.2
India
1992-93
7208
5381
12589
(%)
14.6
10.9
25.5
36286
73.5
1998-99
5284
227
5410
10921
21185
324
32393
(%)
16.2
0.7
33.6
16.7
Distribution of Institutional Deliveries
65.4
1.0
100
Th
T
Health Facility/Institution
NGO/Trusr
Private
1998-99
Pn
►
Size and Nature of Healthcare System
s ■
i 3
I
Andhra Pradesh Development
Private Hospitals Voluntary Hospitals
A.P
1992-93
353
(%)
58.3
562
Public Hospitals
Total
252
41.7
605
100.0
1998-99
396
25
141
(%)
70.5
4.4
25.1
562
100.0
India
1992-93
5381
(%)
42.7
7208
57.3
12589
100.0
1998-99
5410
in
5284
(%)
49.5
2.1
48.4
10921
100.0
w
it -
345
Table 4. Hospital Beds in Public, Private and Voluntary Sectors
Voluntary
Total CMIE
District
Public
Public
Private
Beds Index
Beds
Beds
% Beds
%
%
65.0
37224
2787
10244 27.5
24193
7.5
Coastal Andhra
40
3.6
1102
56
46.3
50.1
510
Srikakulam
552
1.
0.0
404
65
0
20.5
Vizianagaram
321
83
79.5
2.
92
160
3.8
4180
1602
Visakhapatnam* 2418
57.8
38.3
3.
93
404
70.1
East Godavari’ 1624
5.9
6793
23.9
4765
4.
0.0
0
87.1
4935 115
West Godavari 638
12.9
4297
5.
58.2
826
3182
5471 119
Krishna*
26.7
15.1
1463
6.
942
13.6 6902 114
4196
60.8
25.6
Guntur*
1764
7.
3.2
86
100
3157
82.4
Prakasam
14.4
2601
456
8.
4280
96
7.4
1008
23.6
69.1
315
Nellore
2957
9.
11.4 8507
37.0
51.6
973
Rayalaseema
3145
4389
84
33.2
18.7 2808
932
525
48.1
1351
10. Chittoor’
17.2
85
31.4
313
1825
939
51.5
573
11. Cuddapah
0.0
1564
92
30.6
0
69.4
479
12. Ananthapur
1085
2310
83
34.4
5.8
1380
135
59.7
795
Kurnool*
13.
2.2 33257
19212
57.8
738
Telengana
13307 40.0
90
5.0
1794
41.0
968
53
54.0
14. Mahbubnagar 736
42
2256
1.9
31.4
66.7
75
709
1505
15. Ranga Reddy
0.0
0
16099 299
56.9
Hyderabad’
6946
43.1
9153
16.
0
86
0.0
1227
52.9
649
47.1
578
17. Medak
2.2
1801 104
62.1
40
35.6
642
1119
18. Nizamabad
2.2
59.4
446
1159 66
688
38.5
25
19. Adilabad
210
2468
97
65.5
641
26.0
1617
8.5
20. Karim nagar
80
2463
43.7
97
3.9
1077
21. Warangal*
1289
52.3
5.2
2132 76
110
70.4
24.4
1501
Khammam
521
22.
124
1858
71
63.3
6.7
30.0
1177
557
23. Nalgonda
46550
58.9 4498
35.4
5.7 78988
Total_______ 27940
S.No.
-
1
IT
4?
Source: IHS (1996).
Note: ’ Districts with Govt Medical Colleges.
Source: UPS, National Family Health Surveys, 1992-93 and 1998-99.
IsP
The relative size of public and private sectors in some districts was
affected by the presence of government medical colleges. The relative size of
private sector was small in Kurnool district in Rayalaseema and Warangal
district in Telengana due to presence of government teaching hospitals. But in
the most developed districts like East Godavari, Krishna and Guntur in the
Coastal Andhra, the private sector had become very predominant despite the
presence of large teaching hospitals in the government sector. The growth of
private medical care within the state is thus closely related to the overall
development at district level.
f
-4
The per capita availability of beds in private sector also varied with
overall development of districts (Table 5). In the public sector the per capita
bed availability was more uniform across the districts. It was marginally high in
the districts with the government teaching hospitals. However, the per capita
availability of beds in the private sector tended to be higher in the developed
districts. The number of hospital beds per 1000 population was around 1.0 in
the developed south Coastal districts. In the backward north Coastal districts,
Rayalaseema and in most of the Telengana districts it was about 0.5 bed per
thousand population in the private sector.
Andhra Pradesh Develot
346
znt
Voluntary
0.22
0.24
0.02
0.48
Vizianagaram
Visakhapatnam
0.04
0.15
0.74
0.05
0.19
1.28
East Godavari
West Godavari
0.36
0.18
0.09
Krishna
1.05
1.25
0.86
0.4
0.22
Guntur
1.02
0.43
0.94
0.17
0.42
0.23
0.04
1.68
Prakasam
0.13
1.79
0.16
0.14
0.87
0.8
Coastal Andhra
Srikakulam
1.
2.
3.
4.
5.
6.
7.
8.
0.49
Nellore
9.
Rayalaseema
1.24
10.
Chittoor
11.
Cuddapah
Ananthapur
0.29
0.41
12.
Total
Public
Private
13. Kurnool
T elengana
14. Mahbubnagar
0.15
0.36
0.42
0.25
0.34
1.5
1.43
1.48
1.15
0.49
0.63
0.06
1.05
0.03
0.02
0.59
0.89
5.12
0.32
0.24
Ranga Reddy
0.59
Hyderabad
medak
2.91
0.28
2.21
Nizamabad
0.29
0.56
0.25
0.32
0.02
0.9
19.
20.
Adilabad
0.22
0.34
0.01
Karimnagar
0.53
0.21
0.07
0.57
0.81
21.
Warangal
0.38
0.46
0.03
22.
Khammam
0.24
23.
Nalgonda
0.69
0.41
0.2
0.05
0.04
0.71
0.43
0.07
15.
16.
17.
18.
Total__
Source: IHS(1996).
iem
347
Table 6. Rural-Urban Distribution of Hospital Beds (In Per Cent)
Table 5. District wise Availability of Beds per 1000 Population
Districts
Size and Nature ofHealthcare
0.54
0.87
0.98
0.65
1.21
Medical facilities in the private, public as well as voluntary sectors were
located mainly in the urban areas. About 80 per cent of total hospital beds in
the state were located in the urban areas. When compared to the private sector
(78 per cent), the urban concentration was actually more in the public(86 per
cent) and voluntary sectors(93 per cent) (Table 6). As a result, the presence of
private &ctor was much stronger in rural areas (72 per cent) when compared to
urban areas (56 per cent), although in both the areas the facilities were more in
private sector (Table 7).
S.No.
District
Coastal Andhra
Srikakulam
Vizianagaram
2.
Visakhapatnam
3.
East Godavari
4.
West Godavari
5.
Krishna
6.
Guntur
7.
Prakasam
8.
Nellore
9.
Rayalaseema
Chittoor
10.
Cuddapah
11.
Ananthapur
12.
Kurnool
13.
Telengana
Mahbubnagar
14.
Ranga Reddy
15.
Hyderabad
16.
medak
17.
Nizamabad
18.
Adilabad
19.
Karimnagar
20.
Warangal
21.
Khammam
22.
Nalgonda
23.
Total______
Source: IHS(1996).
Private_________ Public______ Voluntary_______Total
Rural Urban, Rural Urban Rural Urban Rural Urban
77.9
92.7 22.1
85.2
7.3
73.1 14.8
26.9
100.0
10.3
89.7
0.0
80.4
19.6
98.8
1.2
80.2
0.0 19.8
0.0
75.1
0.0 100.0 24.9
95.4
100.0
4.6
0.0
94.0
6.0
97.1
2.9
78.1
94.1 21.9
93.4
5.9
6.6
28.4 71.6
54.5
0.0 45.5
0.0
2.5
37.7 62.3 97.5
97.0 22.5
77.5
3.0
90.4
9.6
66.5
33.5
81.6
88.3 18.4
94.8
11.7
5.2
74.5
25.5
76.5
100.0 23.5
0.0
84.9
74.2 15.1
25.8
79.4
85.7 20.6
14.3
84.5
76.9 15.5
23.1
80.8
93.5 19.2
89.8
6.5
35.7 64.3 10.2
96.2 10.9
89.1
3.8
8.9
91.1
82.3
17.7
93.4
6.6
86.3
93.0
13.7
7.0
96.1
3.9
83.8
16.2
88.3
73.7 11.7
26.3
58.6
100.0 41.4
0.0
88.3
0.0 11.7
100.0
84.8
92.1 15.2
7.9
85.5
84.0 14.5
16.0
50.8
100.0 49.2
0.0
73.4
29.1 26.6
70.9
61.6
100.0 38.4
0.0
13.8
83.1 86.2
16.9
0.0 100.0
0.0
0.0
0.0 100.0
0.0 100.0
0.0 28.4 71.6
0.0
68.6 25.1
74.9
31.4
66.9
100.0
33.1
0.0
79.8
20.2
58.4
41.6
84.1
100.0 15.9
0.0
73.3
0.0 100.0 26.7
100.0 25.6 74.4
0.0
66.5
74.3 33.5
25.7
100.0 32.6 67.4
0.0
86.2
41.9 13.8
58.1
85.2
100.0 14.8
0.0
63.0
91.8 37.0
8.2
76.2
53.2 23.8
46.8
85.6
74.2 14.4
25.8
82.1
92.8 17.9
7.2
86.2
21.7 78.3 13.8
A survey undertaken by the Centre for Economic and Social Studies
confirmed these trends in the size of private sector and identified characteristics
of medical care at different stages of development (Narayana, 1998). It was a
comparative study of public and private hospitals in Eluru, Cuddapah and
Mahbubnagar towns that are head quarters of West Godavari, Cuddapah and
Mahbubnagar districts respectively. Excluding Hyderabad, West Godavari was
ranked second on the index of infrastructure development by the Centre for
Monitoring Indian Economy (CMIE, 1992). In contrast, Mahbubnagar was
the least developed district with the lowest rank on the CMIE index. Cuddapah
was at a medium level with its rank of 15 among the 23 districts in the State.
An earlier study on ranking of districts on the basis of overall development also
gave similar ranks to these districts (Iyengar and Sudarshan, 1982). A sample
of three towns located in three regions (Coastal Andhra, Rayalaseema and
WfflM
■ Ti.’«
3
■
'I
1
Ito J
toto®h
ri
t'fi :
it®
■wi
to
Andhra Pradesh Development
348
Telengana), which are at widely different levels of development, was
purposively selected to observe the emerging pattern in the medical care in the
process of development.
Table 7. District wise Rural-Urban Distribution of Hospital Beds
S.No.
District
Coastal Andhra
1.
Srikakulam
2.
Vizianagaram
Visakhapatnam
3.
4.
East Godavari
West Godavari
5.
6.
Krishna
Guntur
7.
8.
Prakasam
Nellore
9.
Rayalaseema
10. Chittoor
11. _ Cuddapah
12. Ananthapur
13. Kurnool
Telengana
14. Mahbubnagar
15. Ranga Reddy
16. Hyderabad
17. medak
18. Nizamabad
19. Adilabad
20. Karimnagar
21. Warangal
22. Khammam
23. Nalgonda
Total________
Source: HIS (1996).
____________Rural_____________________ Urban________
Private Public
Voluntary Private Public Voluntary
5.3
0.0
24.4
91.2
72.3
86.6
84.1
90.7
77.2
94.7
100.0
75.6
7.2
27.7
11.4
7.2
9.3
17.7
54.1
30.8
49.8
83.2
39.3
33.3
50.2
16.8
77.8
29.4
22.2
70.6
58.5
78.2
0.0
65.9
77.9
38.9
68.8
71.7
41.5
21.8
100.0
34.1
22.1
61.1
18.1
26.1
1.6
2.0
8.6
5.1
6.6
35.8
13.1
2.2
51.0
25.6
39.0
64.3
99.4
49.9
55.5
79.8
67.0
44.9
74.4
57.0
28.6
0.6
31.2
29.7
16.0
25.1
4.0
0.0
4.0
7.2
0.0
18.9
14.8
4.1
7.9
30.6
52.9
26.9
49.2
31.3
73.1
90.0
20.2
15.8
30.9
89.9
56.9
50.7
54.2
45.7
65.4
27.2
75.9
61.7
55.9
59.2
7.1
43.1
49.3
42.5
51.7
23.2
67.0
18.1
33.7
37.5
9.9
3.0
3.3 ’
2.6
11.4
5.8
6.1
4.7
6.5
In terms of number of doctors the size of private sector was larger than
the public sector in all the towns (Table 8). About 75 per cent of doctors in
Eluru and Cuddapah, and a little more than half (57 per cent) of them in
Mahbubnagar were in the private sector. In addition, the private sector includes
the private practice of doctors in the government service. About 85 per cent of
government doctors in all the towns have private practice, mostly clinic based.
For instance, the government doctors accounted for about 32 per cent of
private clinics and 25 percent of outpatients in Eluru. The proportion of
government doctors in the private sector was higher in the less developed
Size and Nature ofHealthcare
349
iem
towns In the less developed Mahbubnagar town the share of government
doctors was about 45 percent of private chmcs and outpatients (Table 9) In
in
terms of hospital beds, the government sector was still predominant m
Mahbubnagar (55 per cent) but its share was reduced to 45 per cent in
Cuddapah and 40 per cent in Hum (Table 8). Thus there was a gradual shift
Table 8. Public and Private Mixjn Medical Care____________
Cuddapah________ Mahbubnagar
----------------------- Eluru
Govt Private Total Govt Private Total Govt Private Total
34
18
16
94^ 17
54
71
73
21
MBBS
12
7
5
23
17
6
26
17
9
MS
12
6
6
17
12
30
5
25
5
MD
6
BDS
I
MDS
Total Doctors
Nurses
a) .Trained
b) .Un Trained
Male Assistants
Ayas & Sweepers
Other Employees
Lab Technicians
X-Ray Technicians
Beds (No.)
Inpatients (per day)
Out Patients (per day)
1
0
36
5
3
123
6
3
159
2
0
30
5
3
91
7
3
121
1
0
28
5
0
35
0
63
72
20
52
187
82
105
158
43
43
0
77
77
0
146
74
69
5
91
82
9
144
61
26
35
68
62
6
82
30
113
83
172
39
160 133
28
26
2
67
64
3
55
10
13
3
14
17
34
3
380
165
786
215
434
352
849
161
72
89
417
466 210 207
1190
2081
2722
891
2001 3021 795 1927
In Percent
Mahbubnagar
Cuddapah____
Eluru
Govt
Private Total
Govt Private Total Govt Private Total
52.9 100.0
47.1
77.7 100.0 23.9 76.1 100.0
22.3
58.3 100.0
41.7
73.9
100.0
100.0
26.1
65.4
34.6
50.0 50.0 100.0
70.6
100.0
100.0
29.4
83.3
16.7
83.3 100.0 28.6 71.4 100.0 16.7 83.3 100.0
16.7
0.0
0.0
0.0
0.0 100.0 100.0 0.0 100.0 100.0
44.4 55.6 100.0
22.6 77.4 100.0 24.8 75.2 100.0
73
0
28
25
91
5
3
340
230
1020
85
146
116
57
69
45
29
509
236
MBBS
MS
MD
BDS
MDS
Total Doctors
Nurses
53.8
a). Trained
46.2
100.0
b). Un Trained
80.6
19.4
Male Assistants
69.5
30.5
Ayas&Sweepers
43.1
56.9
Other Employees
81.8
9.1
Lab Technicians
85.3
8.8
X-Ray Technicians
60.0
40.0
Beds (No.)
50.6
49.4
Inpatients
61.7
38.3
Out Patients
100.0 43.9 56.1 100.0 72.2 27.8 100.0
100.0 100.0
100.0 100.0
100.0
93.2 100.0
6.8
100.0
90.1
9.9
100.0
100.0 8.8 91.2 100.0 42.6 57.4 100.0
100.0 77.3 22.7 100.0 73.5 26.5 100.0
7.1 92.9 100.0
100.0 4.5 95.5 100.0
100.0 17.6 82.4 100.0 23.1 76.9 100.0
100.0 44.8 55-2 100.0 56.6 43.4 100.0
100.0 50.4 49.6 100.0 55.3 44.7 100.0
100.0 31.7 68.3 100.0 46.7 53.3 100.0
gi
■
350
Andhra Pradesh Development
in the medical infrastructure towards the private sector with increasing levels of
development.
In all the towns, the diagnostic facilities were Ilargely
o’v concentrated in
the private sector. More than 80 per cent of’ the laboratory andI x-ray
technicians were in the private sector (Table 8). A relatively higher proportion
of diagnostic facilities when compared to their share in the number of hospital
beds and inpatients imply excess capacity and scope for unnecessary testing in
the private sector.
Table 9. Number of Clinics Owned by the Govt and Pvt Doctors
---Huru_______________ Cuddapah________ Mahbubnagar
Govt
Private Total Govt Private Total Govt Private Total
Doctors Doctors
Doc- Doctors
Doc- Doctors
tors
tors
Doctors
1. MBBS
2. MS
3. MD
4. BDS
5. MDS
Total
Outpatients
(Per Day)
14
5
2
2
0
23
295
31
3
8
5
3
50
882
12
45
8
5
10
5
2
7
3
0
24
73
1177 335
27
7
5
5
3
47
795
39
12
10
7
3
71
1130
____________________ In Percent
________ Eluru_____________ Cuddapah_____
Govt
Private Total Govt Private Total
Doctors Doctors
Doc- Doctors
_
tors
I. Doctors
1. MBBS
2. MS
3. MD
4. BDS
5. MDS
Total
Outpatients
(Per Day)
31.1
62.5
20.0
28.6
31.5
25.1
68.9
37.5
80.0
71.4
100.0
68.5
74.9
100 30.8
100 41.7
100 50.0
100 28.6
100
100 33.8
100 29.6
69.2
58.3
50.0
71.4
100.0
66.2
70.4
100
100
100
100
100
100
100
11
4
5
1
0
21
385
14
2
4
4
0
24
490
5
6
9
5
0
45
875
Mahbubnagar
Govt Private Total
Doc- Doctors
tors
44.0
66.7
55.6
20.0
56.0
33.3
44.4
80.0
46.7
44.0
53.3
56.0
100
100
100
100
100
Size and Nature ofHealthcare
351
_.em
positive indicator for the quality of care in the private hospitals. Excepting the
"Other Employees" consisting of clerical staff, a majority of other menial staff
■ -‘ir /
(Male Assistants, Ayas etc.) are in the private sector.
Table 10. Size wise Distribution of Private Hospitals (In Percent)
---------------------------- Eluru
Hospitals
Bed Size
<=10
11-20
21-30
Sub Total
31-40!
41-50
51-60
Total
55.9
26.5
8.8
91.2
2.9
2.9
2.9
100.0
Beds
25.9
29.1
15.5
70.5
7.9
9.8
11.8
100.0
Cuddapah_________ Mahbubnagar
Hospitals
Beds
Hospitals
Beds
48.3
37.9
10.3
96.6
26.7
42.2
19.6
88.5
3.4
11.5
100.0
100.0
50.0
30.0
10.0
90.0
18.8
29.7
15.2
63.6
10.0
100.0
36.4
100.0
No Preference for Private Sector
However, it may be noted that the utilization pattern of public and
private hospitals, as indicated by the NSSs and the NFHSs, does not imply any.
specific preference for private hospitals in AP. The proportion of people going
to public and private hospitals broadly corresponds to their bed strength in the
state The private hospitals accounted for about 72 per cent of beds in the rural
areas and 56 per cent in the urban areas in 1993(Table 7). The NSS (42nd
round) had also shown that the private hospitals accounted for 69 per cent of
inpatient care in the rural areas and 58 per cent in the urban areas. Similarly,
the NFHS (1992-93) revealed that 58.5 per cent of total births in the hospitals
have taken place in the private sector. It is almost equal to the share of the
private sector (59 per cent) in the hospital beds in Andhra Pradesh. It is
therefore erroneous to cite the NSSs and NFHSs as an indication of people s
preference for private sector. It is lack of facilities and poor quality of services in
the public sector that may be driving even the poor to the private sector. It
doest not imply their preference for and ability to pay for the services in the
private sector.
100
i. Eluru as
A majority (55 per cent) of trained nurses/paramedicals in
well as Cuddapah were in the private sector. In the less developed
Mahbubnagar, only about 28 per cent of them were in the private sector.
However,* when compared to its share in number of doctors and beds, the
private sector had a low per cent of trained nurses in all the towns. The private
hospitals depend more on untrained paramedical personnel, which is not a
Nature of Private Sector
Although the private sector accounted for a higher per cent of medical
facilities in the state, the size of private hospitals was relatively small. In all the
towns, the average bed strength of hospitals was about 15 only (Table 12).
And, about 90 per cent of the hospitals had bed strength less than 30 (Table
10) Together, the hospitals with less than 30 beds accounted for more than
two-thirds of hospital beds in the private sector. The GoAP survey also showed
ty
in
-r,
•c
‘g
L1:
itl
iy
id
iis
of
al
4h
fils
•a"
lie
ent
Uy.
He
ral
.nS
F
‘
Andhra Pradesh Deve
merit
353
em
Size and Nature of Healthcare
352
- cent of beds were located in
this pattern at the state level. About 66 per
less (Institute of Health Systems,
hospitals
the bed strength
of 30 or
—
r----- --with
------------------------o--------spitals treat all types of cases and
1996). In addition, a majority of the private o^P
^fious SpeciaiisationSj the
a few were confined to a single speciality- A^isa^ons was the single largest
obstetrics in combination with other speci
specialisation in the private sector.
f doctors in the private sector,• > a
Apart from having a higher number o
,,
. , ,
,
had hospital-based practice
majority (about 60 per cent) of the private
half (47.8 per cent) of the
in
in Eluru
Eluru (Table
(Table 11).
11). In
In Cuddapah
Cuddapah also
also nea^^e sector in Mahbubnag;
igarwas
doctors were hospital based. In contrast, the pn
....
i doctors had hospital-based
mainly clinic based. Only 31 percent of the P^1V^ pro,
.m clinic to hospital-based
practice. In other words, there was a gradual s
practice with increase in levels of developmentdegrees
tended to have
In all the towns, the doctors with p° ^oriSpjcuous at higher levels of
hospital-based practice but the trend was more^
were p0Stgraciuates
development. On average about 40 per cent
all the towns but the per cent of postgradua^
with hospital-based
ed to Cuddapah (58.6) and
practice was very high in Eluru (75-6) when co
Mahbubnagar (53.8). Hence in Eluru, the
sector is predominantly hospital based and
jeveioped town, the private
oriented.
medical sector were also
The differences in the composition o^^eatej at hospitals and clinics,
reflected in the average number of outpatients
outpatients in
towns
The private sector accounted for a high per cen
$
Cuddapah
but the proportion of outpatients treated at ^^^ubnagar (26.5 per cent).
(41 per cent) was higher when compared to an
Thus, there was a gradual shift in the outpatien
to hospitals
h;^her share of hospitals in
reported in the NSS 42nd
with the increase in the levels of development
the outpatient care within the private sector 1
survey also.
- • these towns were family owned
About 90 per cent of the hospitals in
j-jowever, the per cent of
increasing with increasing
and headed by a single doctor or doctor-cou^
c----hospitals jointly headed by the doctor-coup
3UL u 41 per cent of hospitals were
levels of development
(Narayana,
1998)*
.
' 1-cent in Cuddapah and 10 per
headed by the doctor-couples as against 31
private hospitals are turning
cent in Mahbubnagar. Perhaps it indicates that
regions>
into family enterprises in the small towns in
Corporatisation ofMedical Care
requiring huge capital
The growing capital intensity in medical
•dical teC
te< ^n of medical care. The private
investments, has been encouraging corporatisa^^gntrepreneurs
in
nursing homes owned by the individual
0
Table 11. Nature of Private Sector
Mahbubnagar
Eluru
Cuddapah
Dr’s TotalHosp Clinics D.C’s Total Hosp Clinics D.Cs Total
Hosp Clinics
itals
itals
itals
18
0
14
4
54
0
27
27
73
0
31
42
0
7
2
5
0
15
7
17 8
0
3
14
6
0
4
2
12
0
5
25 7
0
8
17
4
0
4
0
6
0
5
1
0
5
5
0
0
0
0
0
3
0
3
0
3
0
3
0
0
35
24
11
90
0
47
0 123 43
50
73
MBBS
MS
MD
BDS
MDS
Total
Nurses
12
73
Trained
69
Un Trained 77
63
Male Assistants 53
52
5
Ayas&Sweepers
Other
14
39
Employees
0
5
Lab Technicians
X-Ray
0
3
Technicians
0
509
Beds(No.)
Out Patients 824 1177
(Per Day)
0
0
0
0
85 105
146 18
116 22
57 58
0
59
49
4
0
0
11
0
105
77
82
62
19
14
30
19
1
29
39
16
0
0
0
0
20
43
69
35
16
40
69
45
28
30
11
8
0
26
39
64
5
3
20
0
5
23
30
26
-S.! L
5
0
0
0
14
2
434 165 0
1927 315 875
8
0
0
10
165
1190
|
1
26 29 8
0 509 434 0
0 2001 797 H30
In Percent ___________
- ---------- E1 ru
Cuddapah--------------- Mahbubnagar-----Hosp Clinics D.C’s Total Hosp Clinics D.Cs Total Hosp Clinics D.Cs Total
_____
itals------------------______
itals
i«ls
1^
--100
100 22.2 77.8
100
50.0
50.0
57.5 42.5
MBBS
100
100 71.4 28.6
100 53.3 46.7
82.4 ' 17.6
MS
100
100 33.3 66.7
100 58.3 41.7
68.0 32.0
MD
100
100
100
100 16.7 83.3
100
BDS
100
100
100
■
100
MDS
100
100 31.4 68.6
100 47.8 52.2
Total Doctors 59.3 40.7
Nurses
100
100 95.0 5.0
100 100.0
85.9 14.1
Trained
100
100
32.6
67-4
100 23.4 76.6
Un Trained 52.7 47.3
100
100 43.5 56.5
100
26.8
59.8
13.4
Male Assistants 45.7 54.3
100
100
54.3
45.7
100
93.5
6.5
Ayas&Sweepers 91.2 8.8
Other
100 16.7 66.7 16.7 100
56.5 20.3 23.2 100 71.8 28.2
Employees
88.5 100
100 11.5
100
46.9
12.5
40.6
88.9
Lab Technicians 11.1
X-Ray
80.0 100
89.7 100 57.1 7.1 35.7 100 20.0
10.3
Technicians
100
100
100
100 100
100
Beds (No.)
100
100 26.5 73.5
100
41.4
58.6
41.2 58.8
Out Patients
.
■
...
■■
ril
i
354
Andhra Pradesh Development
a position to generate necessary investments. It required a large-scale
mobilization of resources from financial institutions, public etc., and the
corporate hospitals are mobilising resources through promoters’ share capital,
borrowing from financial institutions and the snare market. The state also
played a very active role in the corporatisation of medical care by extending
financial and other benefits.
The corporate phenomenon in medical care in Andhra Pradesh began
with the establishment of a multi-specialty diagnostic centre (Medinova) with
outpatient consultancy by a local pharmaceutical company in 1985. In the
hospital sector, the corporate phenomenon began with the establishment of
Apollo Hospital in the state capital in 1989 by the Non-Resident Indian (NRI)
doctors from USA. The state government encouraged the corporatisation of
medical care by providing government land while the central government
offered tax concessions on import of medical equipment. The existing
corp'&mendspitals are opened by the doctors (either NRI or local) and local
business families. All of them are promoted by the local dominantjigricultiiral
castes (Kamma, Raju, Reddy and Velama) and most of them are family
controlled and managed by their family members who are doctors by
profession.
There is no study yet on the exact size and nature of the corporate
sector. There are about 6 big corporate hospitals in Hyderabad. In recent
years, the corporate phenomenon is spreading fast through acquisition of
medium level private hospitals by the corporate hospitals. For instance, the
Apollo hospital has acquired four private hospitals located in different parts of
Hyderabad. Similarly, the CDR hospital has established a number of
emergency medical centres in different localities in Hyderabad. It is basically
intended to widen the catchment area to ensure steady flow of patients for its
main super speciality hospital for high tech treatment. The corporate hospitals
are spreading their network also by giving franchise to small diagnostic centres,
speciality clinics and pharmaceutical stores. Similarly, the Medinova diagnostic
centre has established its branches in a large number of towns in the state. The
corporate hospitals have also come up in agriculturally prosperous towns like
Vijayawada and in the port city of Visakhapatnam. They too have their own
referral centres in the neighbouring towns. In addition, a good number of
specialists from Hyderabad visit clinics and hospitals in the towns every
weekend and refer complicated cases to the super speciality hospitals with
which they are associated in the city.
There is also a trend towards trans-nationalisation of corporate
hospitals’. For instance, the Apollo hospital has a tie-up with a Singaporean
Health Group in a joint-venture company to undertake management of
hospitals in the south and Southeast Asia. While the Parkway Health Group of
Singapore takes the responsibility to secure the contracts for hospital
Size and Nature ofHealthcare System
355
management, the Apollo would supply doctors, nurses, engineers etc. to
manage hospitals that would be under their contract. The basic objective is to
take advantage of India’s medical manpower that is far less expensive when
compared to US and Europe. Recently, the Apollo group has established a
hospital in Colombo, capital of Srilanka.
2. ORIGINS AND GROWTH OF PRIVATE SECTOR
Although the origin of private sector in AP can be traced even to pre
independence days in the state, the real spurt in the growth of private hospitals
occurred mainly in the 1980’s and the 1990’s. About 75 to 90 per cent of
existing hospitals in Eluru, Cuddapah and Mahbubnagar were established
during rhe last two decades (Table-12). The growth of private diagnostic
centers is an even more recent phenomenon; most of them were established
only during 1990s.
_
The government has been responsible for encouraging the growth of
private sector indirectly by neglecting the quality and quantity of services in the
government hospitals and directly by providing subsidies, tax concession etc. to
the private hospitals. During 1980s there was a marked shift in the government
policy in the health sector. In the Health Policy Document (1982), the state
had given up its proclaimed objective to meet the health care targets on its own
(Ministry of Health and Family Welfare, 1981). Instead, it started emphasizing
the role of private sector and non-government organisations in achieving the
objectives in the health sector. The main factors responsible for the rapid
growth of private hospitals in the state are analysed below.
Stagnation in Expansion of the Public Hospitals
Expansion of facilities in the public sector did not keep pace with the
increase in population. It resulted in excess load on the public health care
system and decline in its quality. The shortfall in the quantity as well quality of
services had provided tremendous opportunities for the growth of private
secto r.
The availability of hospital beds in the public sector had declined from
6.2 beds per 10,000 population in 1961 to 4.8 beds during 1970s and 1980s
and 5.1 beds in 1990s (GoAP, Annual Senes). The inadequacy of facilities in
the public sector was further worsened by the lopsided distribution of medical
facilities and public expenditure between the primary, secondary and tertiary
hospitals. Keeping in view the inpatient load at different levels, the Planning
Commission (the seventh five-year Plan) suggested 70 per cent of hospital beds
at the secondary level and 15 per cent each at the primary and the tertiary
hospitals (Government of India, Planning Commission, 1985). However, the
secondary level hospitals have only 30 per cent of total beds.
5
Bi
y
a
is
jf
al
th
as
ihi
©
•it
. icn
•al
>lic
ent
■ity.
He
era!
jin$
01
if-’
W
S-gf ■..
356
Andhra Pradesh Deve
lent
Size and Nature ofHealthcare
*em
1960s
4
65
1
50
1970s
3
72
5
86
1
25
1
2
overcrowding of the hospitals in the urban centres which had created vast
demand for the private hospitals/ nursing homes from the urban middle classes.
Decrease in Allocations to Health Sector
The overall decline in the share of the health sector in the government
budget, particularly after the initiation of Structural Adjustment Programme
(SAP), has further worsened the scarcity of resources in the health sector^
The share of the health sector in the state budget was highest (6.5 per
cent) in the 5* FY Plan. Thereafter it declined continuously and fell to 5.2 per
Sub Total
7
137
7
136
1
25
2
2
cent in the 8'h FY Plan(Table 13). The
1980s
12
215
7
103
6
101
7
8
2
1990s
15
157
15
191
3
39
17
41
11
Total
34
509
29
434
10
165
24
51
15
_____ Table 12, Establishment of Private Medical Facilities__________
Period______________ Hospitals and Beds
Eluru
Cuddapah
Diagnostic Centers
Mahbubnagar
Eluru Cuddapah Mbnagar
Hospitals Beds Hospitals Beds Hospitals Beds
1950s
4
1
Year
In Percent
Period
Hospitals and Beds
Cuddapah
Mahbubnagar
Eluru
Diagnostic Centers
Eluru Cuddapah Mbnagar
Hospitals Beds Hospitals Beds Hospitals Beds
1950s
3.4
0.9
1960s
11.8
12.8
3.4
11.5
1970s
8.8
14.1
17.2
19.8
10.0
15.2
2.0
13.3
2.0
13.3
15.7
13.3
Sub Total
20.6
26.9
24.1
31.3
10.0
15.2
1980s
35.3
42.2
24.1
23.7
60.0
61.2
29.2
1990s
44.1
30.8
51.7
44.0
30.0
23.6
70.8
80.4
73.3
Total
100.0 100.0 100,0
Source: K.V.Narayana (1998).
100.0
100.0
100.0 100.0
100.0
100.0
The government has been responsible for encouraging the growth of
private sector indirectly by neglecting the quality and quantity ofservices in the
government hospitals and directly by providing subsidies, tax concession etc. to
the private hospitals. During 1980s there was a marked shift in the government
policy in the health sector. In the Health Policy Document (1982), the state
had given up its proclaimed objective to meet the health care targets on its
own. Instead, it started emphasizing the role of private sector and non
government organisations in achieving the objectives in the health sector. The
main factors responsible for the rapid growth of private hospitals in the state are
analysed below.
The problem of overcrowding was further worsened by the failure of
referral mechanism in the public health care system. Due to uncertainty in the
availability of staff and medicines at the Primary Health Centres (PHCs), the
patients go directly to the first referral (district and area/sub-divisional
hospitals) and teaching hospitals even for the minor health problems. It is the
5 th FY Plan
(1974-79)
1979-80
6th FY Plan
(1980-85)
7th FY Plan
(1985-90)
1990- 91
1991- 92
Sth Plan
(1992-97)
1997- 98
1998- 99
1999- 00
iar
cis
: o
hi
an<
mi.
Hi
ica
su
357
Table 13. Share of Social and Economic Services
in Public Expenditure (In Percent)
Total
Health
Economic
Social
Expenditure
Sendees
Services
Services
6.5
34.2
42.9
100
5.9
6.2
35.6
40.1
42.7
37.9
100
100
5.8
38.0
38.4
100
5.6
5-6
5.2
36.2
35.2
34.3
37.5
37.3
37.5
100
100
100
33.3
35.5
5.4
29.7
39.2
5.7
28.0 •
37.7
5.8
33.8
32.2
5.0
2000- 01____________ ______________________
Source: Govt of AP, Budget Estimates, Detailed Demands of Grants.
s
*rt)
; />
ber
itly
ind
his
•ral
yth
vas
Jhi
ch,
ten
100
100
100
al
lie
100
:nt
share of the health sector declined even in the and 7 FY Plans in which the
aggregate share of social services increased. In fact, the share of social services
was more than the share of economic services in the 7 FY Plan. The allocation
ty.
He
ral
ng
on
to health sector remained low in the 1990s also.
At constant (1993-94) prices, the per capita public expenditure on
health care was about Rs.58 during the Fifth five-year Plan (Table-14) It
increased by about 45 percent in the next two five-year plans and reached
Rs.85 per capita during the seventh five-year Plan (1985-1990). However, ^he
1990s witnessed a sudden fall in the per capita public expenditure on health
care and other services. It declined from Rs.85 in the 7 plan to Rs.73 in the 8
p^Wn compared to the economic services, the social services experienced
a steep fall in public expenditure. Within the social services, the decline was
di ~
f
f gil
IB
358
Andhra Pradesh Development
more in the case of health care. It clearly indicates adverse impact of the SAP
on the provision of health and other social services.
Moreover, there was a continuous increase in the share of salaries in
the hospital budgets from about 52 per cent in 1974 to 80 per cent in 199?
which left very meagre resources for the supply of materials and maintenance of
the infrastructure (Table-15). The decline in allocations to
Table-14. Per Capita Net SDP and Public Expenditure in AP
at 1993-94 Prices
Year
5 th FY Plan
(1974-79)
1979-80
6th FYPlan
(1980-85)
7th FY Plan
(1985-90)
1990- 91
1991- 92
Sth Plan
(1992-97)
1997- 98
1998- 99
1999-00
2000- 01
Health
Sendees
58
Social
Sendees
304
Economic
Services
381
Total
Expenditure
888
66
75
402
480
481
454
1128
1197
(In Rs.)
SDP
5306
5508
5999
85
558
565
1470
6526
75
72
73
488
463
487
506
492
533
1350
1318
1420
7050
7164
7695
79
88
93
97
515
483
1451
607
453
1550
1610
605
451
____________
627
658____________1945
7964
8790
8980
9386
Source: 1. Govt of AP, Budget Estimates, Detailed Demands of Grants.
2. Bureau of Economics and Statistics (AP), Estimates of SDP.
Table 15. Public Expenditure on Secondary Hospitals (In Percent)
Plan/year__________
Salaries
Vth Plan( 1974-78)
51.7
(1979-80)
54.1
VIth Plan(l 980-84)
59.2
Vllth Plan(1985-89)‘
61.4
1990- 91*
67.8
1991- 92
68.3
1992- 93
73.1
1993-94
74.4
1994- 95
78.4
1995- 96 *_______ _
75.5
* Excluding Grants to APWP.
Materials ^Supplies_____ Others______ Total
100.0
15.8
32.5
100.0
30.9
14.9
27.8
100.0
13.0
100.0
30.2
8.4
100.0
17.8
14.3
100.0
18.6
13.1
100.0
18.2
8.7
100.0
15.1
10.5
100.0
10.0
11.7
100.0
13.2
11.3
359
Size and Nature ofHealthcare System
the medical supplies and maintenance has contributed to faster degeneration in
the quality of treatment in the public hospitals.
Prior to the Structural Adjustment Programme (SAP), there was
increase in the ratio of public expenditure on the health sector in the State
Domestic Product (SDP), in spite of decrease in allocations to health in the
state budget, because of overall increase in the size of state budget. The share of
health sector in SDP increased from 1.09 per cent in the 5 plan to 1.29 per
cent in the 7th plan (Table 16). During the 6th and 7th FY Plans the share of
Social Services was actually higher than the share of Economic Services.
However, with the beginning of SAP in 1990s there was a sudden decline in
the share of social sectors including health. The proportion of public
expenditure on the health sector to the SDP declined from 1.29 per cent in the
7d' FY Plan to 0.94 in the 8th FY Plan. Thus, after the beginning of SAP, there
was decline in the allocations to health services despite all the talk of safety net.
Table 16. Share of Public Expenditure in SDP (In Percent)
Year______ Health Sendees
1.09
5th FY Plan
(1974-79)
1.20
1979-80
1.25
6th FY Plan
(1980-85)
1.29
7th FY Plan
(1985-90)
1.06
1990-91
Social Services
5.7
Economic Services
7.2
Total Expenditure
16.8
7.3
8.1
8.7
7.6
20.5
20.2
8.5
8.5
22.3
6.9
6.5
1.00
1991-92
6.3
0.94
Sth Plan
(1992-97)
6.5
0.99
1997- 98
6.9
1.00
1998- 99
6.7
1.04
1999- 00
6.7
________________
2000- 01
1.03___________________
7.2
6.9
6.9
19.1
18.4
18.4
6.1
5.2
5.0
7.0
18.2
17.6
17.9
20.7
-3 3J;
■
ir
Y -
g
Ob
I
^7
■
...
;•
■Il
.3-'
Source: l.Govt of AP, Budget Estimates, Detailed Demands of Grants.
2. Bureau of Economics and Statistics (AP), Estimates of SDP.
Encouragement to Private Sector
The growth of private hospitals and diagnostic centres was encouraged
directly by the Central and State governments by offering tax exemptions, land
at concession rates etc. in return for provision of free treatment for the poor,
free treatment to at least 40 percent of outpatients and 10 percent of inpatients.
Apart from availing the tax concessions, some hospitals received special benefits
irice. For
like allotment of government land at an insignificant part of market pric^.
instance, in 1989 a corporate hospital in Hyderabad was given 30 acres of
fat
360
Andhra Pradesh Development
government land in a posh locality at a throw away price on the condition of
reserving additional 15 percent of beds for the poor. In 1995, the A P Legislative Assembly had appointed a House
Committee to find out whether or not the corporate hospitals, private nursing
homes and other medical centres honoured the conditions (i.e., free treatment
to the poor) imposed for granting tax concessions, exemptions and other
benefits (Andhra Pradesh Legislative Assembly, 1996). The committee was also
to examine the efforts made by the concerned authorities in seeking the
compliance of these conditions and to make suitable recommendations to
regulate the private medical establishments in future.
The Committee found that "most of the Corporate Hospitals who
have availed the benefit of customs duty exemption were not treating the poor
patients free of cost". The State Government was equally lax about
implementing the conditions on the private hospitals. "The Committee
observed that there was no check as to whether the Apollo Hospital was
adhering to the condition of 15 percent of the beds to be reserved to the poor
for free treatment. All these years, there was no check and no one was made
responsible. The committee was at a loss to understand on the failure of the
State Government in implementing the conditions." It appears that neither the
private hospitals nor the government were serious about the provision of free
service to the poor. It was only a ploy to dole out these benefits to the private
sector.
Apart from subsidies, the private corporate hospitals receive huge
amounts of public funds in the form of reimbursements from the public sector
undertakings, the state and the central governments (eg.CGHS) for treating
their employees. Another major source of revenue is the inflated and
fraudulent claims from the public insurance companies under various medical
insurance schemes.
Lack of Regulation
The lack of government or any other regulation on the minimum
standards in terms of staff, facilities etc. and on the pricing of services in the
private hospitals has contributed to the rapid expansion of private sector in
medical care. Having no controls on its size, quality and pricing, the medical
care became a very attractive outlet for private investment.
Before the era of reform that began in 1991, the public sector
dominance and pervasive controls over the private sector were the key elements
in the development strategy of India. Yet, the private medical care remained
beyond the purview of any regulation for more than 4 decades during which
the private sector achieved “Commanding Heights” of medical care industry in
rhe country.
Size and Nature ofHealthcare
361
Realising that the State Government has no control over the private
hospitals, the House Committee recommended a comprehensive legislation on
nursing homes, diagnostic centres etc. To curb the practice of unnecessary
investigations and prescription of drugs in the private sector, the Committee
recommended preparation of "Clinical Manuals" in consultation with the
Medical Council of India and other professional bodies. The Committee also
suggested a law prescribing minimum standards in terms of buildings,
equipment, medical staff etc. for private hospitals and a mechanism to enforce
those standards. The legislation was also intended to ensure that all private
hospitals maintain inpatient and outpatient registers and case sheets which
should be given to the patient at the time of discharge.
Recently, the state legislative assembly passed a bill to regulate the
private sector in medical care but not yet operationalised it. It appears that the
government does not have either the will or ability to regulate the private
practice in medical care.
Role of Public Sector
Like in other areas of economy, the establishment of public sector laid
foundation for the growth of private sector in medical care. The establishment
of public hospitals generally preceded the growth of private medical care and in
many towns its origin can be traced to the private practice of government
doctors. While most of the government doctors have private clinics, some of
them are also involved in establishment of private hospitals and diagnostic
centres. A good number of private hospitals in small towns still belong the
doctors who were or are in the government service.
In addition, the public investment in free medical education has
created a vast network of medical professionals who provided basis for the
expansion of the private sector.
Impact of Urbanization
Rapid urbanization, expansion of middle class with higher purchasing
power and growing health awareness have together provided necessary
conditions for the rise of private sector in medical care. Urbanisation,
accompanied by the increase in size of middle classes, created market for the
medical care.
3. IMPACT OF PRIVATE SECTOR ON PUBLIC HOSPITALS
The state’s patronage to the private sector is sometimes justified on the
ground that it would ease out pressure on the government hospitals. But in
reality, the private hospitals are replacing rather than complementing the public
hospitals by weaning away the resources from the government hospitals. The
f
d
h
is
hi
h,
cn
al
lie
:nt
ity.
He
iral
jing
on
362
Andhra Pradesh Devei
>nent
adverse impact of privatization on public hospitals is identified in the following
subsections.
Lack of Patronage to Public Hospitals
>
With the large-scale expansion of private health care in 1980s and
1990s, the rich and middle classes no longer go to the government hospitals.
The public hospitals are used mostly by the poor at present (Mahaptra et al,
2002: P.43 and GoAP, 2001). While the rich have deserted the government
hospitals in favour of the corporate hospitals, the middle classes rely on the
small private hospitals. As a consequence, the public hospitals are facing unfair
competition in mobilising resources since the state (politicians and bureaucrats)
is more interested in extending facilities to the private, mainly corporate,
hospitals.
Apart from depravation of resources, the day-to-day maintenance of
public hospitals is also neglected. Whenever public services (hospitals, schools
etc) are used exclusively by the poor, the quality of services tend to degenerate
due to lack of resources and effective supervision
Since public hospitals are used exclusively by the poor, there is little
scope for generating additional resources through user charges. The growth of
private corporate hospitals was found to have adverse impact on the demand
for paid wards in the government hospitals, particularly the teaching hospitals.
The bed occupancy in the paying wards was only about 35 per cent in the
Osmania Hospital (Prasad, 1995). In the past, there used to be waiting lists for
paying wards in the same hospitals.
Luring away Specialists from Public Hospitals
A study of public and corporate hospitals found that the popularity
and use of corporate hospitals depend mainly on the reputation of individual
doctors (Prasad, 1995). The patients still choose the doctors, not the hospitals.
As a result, there is severe competition between the corporate hospitals to
attract the reputed specialists by offering higher remuneration, better service
conditions etc. In the process, many reputed specialists are lured away from the
government teaching hospitals and the shortage of specialists has further
adverse impact on the image of government hospitals.
While the big corporate hospitals offer attractive service conditions for
the senior specialists, the junior doctors, nurses and other staff have low salaries,
long working hours and no job security. The junior doctors prefer government
service on account of job security, better working conditions, diversified
experience i.e. greater exposure to a variety of cases. Once they gain enough
experience in the government hospitals, the corporate hospitals lure them away
with better salaries and service conditions. The government sector has thus
become a cheap source of well-trained specialists for the corporate hospitals.
Size and Nature ofHealthcare System
363
Transfer of Critical Cases from Private to Public Hospitals
Public hospitals are used by the private hospitals for -dumping
unwanted cases. The private hospitals transfer serious/terminal cases to the
government hospitals to avoid medico-legal problems and to maintain the
image of being efficient. Even the super-specialty corporate hospitals transfer
terminal cases, AIDS patients etc. to the government hospitals. The problem
has become widespread after the extension of Consumer Protection Act to the
medical care in recent years. As a consequence, the public hospitals face the
problem of adverse case mix, apart from overcrowding.
Growing Capital Intensity, Cost Escalation and Supplier Induced Demand
With privatization and corporatisation, the medical care became very
capital intensive and added to escalation of cost inflation in the medical sector.
Because of growing competition in the private sector, there is excess investment
in diagnostic and other equipment to attract patients. The government policy
of tax exemptions on import of medical equipment and easy loans from public
financial institutions has further encouraged the capital intensity and excess
supply in the private sector. This in turn led to unnecessary tests and treatment
to recover huge investments on the equipment. It was found that between the
public and private hospitals, the latter have a higher proportion of Caesarian
deliveries. In Eluru, Cuddapah and Mahbubnagar towns, the incidence of the
Caesarian operations in the private hospitals was about 16 per cent higher than
that in the public hospitals (Narayana, 1998). The problem of supplierinduced demand has become so rampant that the state unit of Indian Medical
Council publicly warned the private hospitals against the unnecessary tests and
surgeries.
4. REFORMS IN PUBLIC HEALTH SECTOR
With the avowed objective to improve the financial viability and
quality of care in the public hospitals the State Government initiated a series of
reforms, mainly at the secondary and primary levels. To begin with the reforms
were introduced on its own but later on the reforms were initiated and guided
mainly by the World Bank. The following sections describe reforms in the
public hospitals at the secondary and primary levels.
The Andhra Pradesh Vaidya Vidhana Parishad
The scarcity of funds and declining standards in the government
hospitals led to the creation of the Andhra Pradesh Vaidya Vidhana Parishad
(APWP), an autonomous Commission for the management of secondary level
hospitals in 1986. The objective was to grant financial and administrative
autonomy to the secondary hospitals. Autonomy to the hospitals was aimed at
r'~ B. :
nj
364
Andhra Pradesh Development
reducing the financial burden on the government through more efficient use of
existing resources and mobilising additional resources. It is also expected to
improve the quality of care and patient satisfactigp in the public hospitals.
As a part of financial autonomy, the APWP is empowered to receive
funds and donations from general public and financial institutions. Apart from
receiving grants-in-aid from the state government, it is empowered to take
loans from domestic and foreign financial institutions and collect user charges
for diagnostic and treatment services. It can even construct and maintain
commercial complexes in the hospital premises, and organise lotteries to
achieve financial self-sufficiency.
To provide administrative autonomy, the Commission is empowered
to make its own rules and regulations in running the secondary hospitals.
However, all the new rules and regulations are to be placed before the State
Legislative Assembly for its ratification and in case of any dispute between the
Government and Commissionerate, the former would prevail. In practice, such
clauses nullified the very objective of autonomy and reduced the APWP to
one among many other government agencies.
In spite of limited autonomy, the APWP seems to have achieved
considerable improvement in the functioning of public hospitals at the
secondary level (Chawla, no date). Creation of a single apex agency is credited
with elimination of bureaucratic delays in getting sanctions from the State
Secretariat. It is also supposed to have improved the medical supplies and
maintenance of diagnostic equipment and other infrastructure facilities.
However, the APWP failed in achieving two of its basic objectives:
recruitment of independent staff on contract basis and financial independence.
One of the objectives of APWP was to introduce flexible manpower policy
based on contract system. The APWP has legal authority to recruit its staff,
independent of the Department of Health, Medical and Family Welfare
(DHMFW), but in practice it faced many hurdles. The APWP tried to recruit
staff on contract basis but could not sustain due to internal resistance. The
doctors complained about the feeling of insecurity, discrepancy in the service
conditions between the existing and newly recruited staff etc. Finally, the newly
recruited doctors were absorbed by the DHMFW and the APWP does not
have any doctors on its own.
A major objective in establishing the APWP was to introduce user
charges. However, the matter was so sensitive politically that the first
commissioner of APWP was sacked within three months by the government
for announcing the introduction of user charges which created furore in the
State Legislative Assembly. Other financial proposals also did not materialise.
As a result, the revenue receipts remained an insignificant part of its
expenditure that is financed mostly from the block grants from the state
Size and Nature ofHealthcare System
of
365
OS
doubtful
government. The financial self-sufficiency of the APVW
because there is very limited scope for raising revenues through user charges.
ic
is
a/
in
,<1
•ry
in
:r,
Andhra Pradesh First Referral Health Systems Project (1995-2002)
With the aid from World Bank, the Andhra Pradesh First Referral
Health System (APFRHS) Project was launched to introduce reforms at the
secondary level hospitals that are under the management of APWP (World
Bank 1994). Apart from renovating and upgrading the secondary level
hZi’tat through huge investments (US $159 million/RsAOS billion, 9
percent of which is funded by the IDA) in civil works, medical equipment,
additional staff and medical supplies, the project aims at introducing major
cc
changes in the health policy and institutional structures
The major policy changes proposed in the health policy include the
-
introduction of user charges at secondary hospitals, contracting out supportive
services and shift in allocations from the tertiary to secondary level hospitals.
To strengthen the policy development capacity it proposed a Strategic Planning
Cell in the Health Department to organise workshops and seminars, and
commission studies on key issues like implementation of cost recovery
mechanisms in public hospitals, and cost effectiveness of public health
ily
-?• ■■
intervention.
.
r ,. .
r i D
Under the institutional development, the basic of
World Bank Project is to secure complete autonomy to the A1VV1 in
introducing user charges, contracting out suppornve services recruitment of
staff on contract basis etc. As a part of the project, the APWP is rev.smg all
employment regulations and seeking the government approval to implement
them. The project envisages that the state government would revise all cadre
ree;ulations and transfer all staff to APWP.
The APFRHS project stipulates introduction of user charges to recover
at least the non-salary component of incremental recurrent costs. It is estimate
that the expansion of beds and improvements in the quality of care would
result in additional recurrent costs of Rs.300 million per annum It is aimed at
recovering Rs.45 to Rs.55 million per annum or 15-18 percent of incrementa
recurrent costs. It is also proposed to decide the extent and level of user charges
ig
nd
us
of
ral
*th
’as
ihi
A
i
cn
al
lie
snt
ty.
He
ral
ng
on
on the basis of cost effectiveness of medical interventions: to impose user
charges on the services which are low in cost-effectiveness and subsidise the
ones high in cost-effective procedures.
The project approach represents two major shifts in the health policy.
Firstly, the shift in emphasis from the incidence of disease to economic burden
of disease in setting health sector priorities. The cost-effectiveness of health
interventions was assessed in terms of gains in the Disability Adjusted Life
Years (DALYs). In other words, the economic value of people is taken into
consideration in the provision of health services. Any emphasis on investment
3
fell
Chart 1.5
Key Participants in the Health Care System in the United States
Private Insurers
----------------Budget/Trust Fund
Federal
1
Government
T
Stale Medicaid
Budgets
Premia
Medicare
TradMonal
Managed
Insurers
CareEnNes
Programmes
State
<
FFS (TradMonal) /________
Various Contacts (Managed)
Budget T
Governments
(/)
>
cr
I
CPC (DRGs)
co
a.
Negotiated Prices
FFS (Traditional) /________
Various Contacts (Managed)
1
FFS/Formularies
Pharmacists
i
i
i
i
Discounted FFS
Primary
and CPC
Physicians
General TaxaborVPaytol Taxes
---- Service Flows
h >i
>»
2
Copayments A OOP
Copayments & OOP
ill
General Taxation
------ ► Financial Flows
Hospitals
8'
;
i
i
i
i
I
If
ReimtMjrsment tor OOP A Copayments (eg Medlgap)
♦ otter benefits (eg pharmaceuticals)
Patients
Taxpayers
3
<n
-3
Consumers
(Mosfy) Risfc Related Premia
Chart 11.3 The reformed health care system in the Netherlands
(publicly financed sector)
Closed,
compulsory,
income-related
contributions
Transferable
risk-related —
contributions
Central
iund
Competing
insurers
i
i
i
i
i
i
i
i
i
i
i
i
i
i
I
I
Open, voluntary, flat rate,
insurer-specific
contributions
i
i
i
i
I
I
i
i
I
I
I
i
I
I
I
Services
Competing
first-level
providers
Payments
to
providers
i
i
1
i
i
i
i
Population
and
enterprises
i
Referrals
i
i
i
i
Patients
i
Services
i
i
i
Competing
second-level
providers
i
i
Chart 7.1 The health care system in the Netherlands in 1987
Private insurers
32 % of
population
Voluntary _
premiums
i
i
i
Gid 7.4 b.
Insurance for
public servants
6 % of population
1 Compulsory
I pz-uMriKi ifirinc
contributions
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
i
Reimbursement of patients
i
Reimbursement of patients
Gid 1.4 b.
i
1 Compulsory
1 contributions
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
Exceptional medical
expenses scheme
100 % of population
i
i Tax subsidyi
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
1 Compulsory
1 contributions
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
' Tax subsidy-
Sickness funds
62 % of
population
Gid 9.9 b.
Gid 15.9 b.
I
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i General
i taxation
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
Government
budget
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
Public health
services
Gid 0.4 b.
Pharmacists
Patients
Fee for service
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
___ i
Gid 2.2 b.
Gid 1.2 b.
Gid 0.7 b.
^^\Gld3.5b.
Gid 1.0 b.
Population
and
enterprises
" " Services and
_________ jGId 0.2 b. -►
direct payments
General
practitioners
Mainly private
general hospitals,
+ specialists
Psychiatric
hospitals and
institutions
Mental
handicap
services, etc.
4
I
I
I
I
I
Domiciliary
health
services
I
Reimbursement
of patients
i
i
i
i
i
Central
payments
mechanisms
----- Capitation-------- 1
Gid 1.0 b.------------------- '
“ *<- Global budgets Gid 6.9 b.
_
i
— Fee for service -•
Gid 2.3 b.
Gid 2.7 b.
Gid 2.3 b.
Gid 0.6 b.
Nursing
homes
I
I
I
Gid 3.9 b.
I
Service flows
<-------------
Financial flows
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
si
v>
*
1
I
u.
CPC •nd FFS
st
*
£
a>
£
£
2
o
CL
O
E0)
i
i
i
i
i
i
i
i
i
i
(/)
£
d>
■
£
1 5
o 0)
£
0
I
co
O
I
w
u.
WaigMad
Capitation
■8
o
co
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
£
3
1
«)
o
itx
Wwghted
Capitation
iZ co
CO <D
co
§
CPC/Budg«ts
w
co
CXI O
J3
i
Drug Tariff Priam
I
I
I
I
I
I
I
I
I
I
I
I
•
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Ct
Id'
J
1
o
*
i
i
i
Chart 1.2
Key Participants in the Health Care System in Canada
Federal
Unweighted Capitation
Government
Indexed to real GNP
Provincial
Budgets
Private Insurers
Provincial
Governments
Health Systems
3
E
Various (Care outside Provincial Plan)
Fee For Service
d
o>
■o
m
i
General
Hospitals
Practidoners
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
“l
Pharmacists
i
i
i
CL
i
i
8
•
i
'
ti
I
Reimbursement (pharmaceuticals, care
outside the Provincial Plan)
Patients
Taxation
Risk Related Premiums
T axation
Financial Flows
► Service Flows
Taxpayers
Consumers
8
8
o
c
0)
O)
1 z
Q.
i
C
0)
E
1
<0
I C«
!t=
<i 2 2
n
ft J2
c c
E
o
>>
a<0
(Z)
w
O
'Ji?
:
<0
c
o
E
o
eV; iffi
I
CZ)
(D
03
0>
i
s
3
±5
oCL
■a
I£
t
I
<r
O\
TJ
Z5
a)
h4
OJ
o
3
o
«
INPUT
HEALTH CARE SYSTEM —>
OUTPUT
economic
financing
health status
political
delivery of services
income distribution
social
reimbursement
employment
regulation
competitive power
Exhibit 1 An input-output model of a health care system
i
a
■
: -bi
Figure 5.1 Income and health spending in seventy countries, 1990
Share of GDP spent on health
Percent
13
united States •
11
France
9
o
Lesotho
7 - o
o
o
Jamaica
-°o
5
o,> Kenya
■* a u
o
o
°
•rr—Brazil
o
o
o
o
Japan q
United Kingdom
°OO
0
Philippines
• Egypt
1
0
°4o
° Norway^^as-
o o
o
o
o
3
• Canada
o
Korea,
Rep. of
India
o
5,000
10,000
15,000
GDP per capita (1991 international dollars)
20,000
_t
• . j ■
o
Model Gabay and Wolfe
1
OLDER THAN 65
HEALTH CARE
EXPENDITURES
OCCUPATIONAL RISK
NEGATIVE LIFE STYLE
HEALTH STATUS
. MA ARSE
»
■
u
c
I
KMJ 0H»6d6 H11WH VJJM) M
0011
DON
ONI
OKI
0091
r
“T
T ~i—r T
T
•taros
- MOI
(MUrtHlum
M»«N
aw
OK
OK
T T T T T
1«
ON
OM
tuauow auias
OK
oat
9
L
8
e
01
K
£1
n
..•r CS’N
HOd x ..6900 - ..•‘H =
si
91
$
£
ii *
•ddd Nl ONIQNadS HIIVSH VlldVO H3d SHSHBA
S31Vd H1V3Q INVdNI aivwad
6
DATA PROBLEMS IN INTERNATIONAL HEALTH CARE SYS
TEMS RESEARCH
Data are crucial important both for scientific progress and health
care policy making!
Lack of available data
Lack of reliable data
Lack of comparable data.
Lack of conceptual unity: The apples and oranges problem
National health statistics follow the prevailing pattern of
administrative accounting.
* physician/population ratio in the UK
* share of hospital expinditures in Germany
* cultural element in medical recording
* what is included in the national health expenditures?
Summercourse on Health Care Policy Analysis. University of Limburg, Faculty of Health
Sciences. Maastricht, July 1994
HEALTH AFFAIRS I Fall 1991
26
Exhibit 3
Per Capita Health Spending And Per Capita Gross Domestic Product (GDP) In
Twenty-Four OECD Countries, 1989
Per capita health spending ($PPP)’
3,000
2,500
2.000
I. United State*
7. Norway
13. Italy
19. New Zealand
2. Canada
8. Germany
14. Japan
20. Ireland
k Belgium
9. Luxembourg 15. Australia
2 I. Spain
4. Sweden
10. Netherlands
5. Iceland
I 1. Austria
17. Denmark
22. Portugal
23. Greece
6. France
12. Finland
18. United Kingdom
24. Turkey
16. Belgium
i
1,500
6
105]
LQQO
i
500
0
|22|
[E3
21
.2
4
6
8
10
Per capita GDP (thousands $PPP)
12
14
16
18
20
Sources: C'-.J. Schiehcr and J.P. Poullier, “International Health Spending: Issues and Trends
(Spring IWI): 106-116; and OECD Health Data, 1991
’
22
’’
Note: PCH = pe-r capita health spending. PCGDP=per capita gross domestic product. PCH= -419+107 x
I'CGDP Both the constant term and the regression coefficient are statistically significant at the 01 level
65 (adjusted correlation coefficient squared; also significant at the 01 level)
PurchiiMng power purities.
R- =
r
7
Problems with data on health expenditures
*
fluctuating exchange rates
*
price differences
If a country A spends say 40 percent more on health care
than country B, this does not necessarily mean that the
volume of health care activities in country A exceeds the
volume in country B by 40 percent. Price differences may
(partly) explain the 40 percent difference in health expendi
tures.
Thus: comparability problem — > conversion to a common
denominator is needed —> correction by means of Pur
chasing Power Parities (PPPs). The computation of PPPs is
still rather crude
Effect: variance in health care expenditures declines
CONCLUSION: BE CAREFUL WITH YOUR CONCLUSIONS!!
Summercourse on Health Care Policy Analysis. University of Limburg, Faculty of Health
Sciences. Maastricht, July 1994
Y|
- Media
RF_PH_22_SUDHA.pdf
Position: 219 (10 views)