Integrated Management of Neonatal and Childhood Project Proposal Illnesses: Baseline assessment of childhood morbidity & mortality in selected districts in India
Item
- Title
-
Integrated Management of Neonatal and Childhood
Project Proposal
Illnesses: Baseline assessment of childhood morbidity
& mortality in selected districts in India - extracted text
-
dis
SDA-RF-CH-1A.20
integrated Management of Neonatal and Childhood
Illnesses: Baseline assessment of childhood morbidity
& mortality in selected districts in India
Project Proposal
Project Coordinator
Dr Narendra K. Arora
Professor of Paediatrics
Team Leader, IndiaCLEN Program Evaluation Network
Clinical Epidemiology Unit
All India Institute of Medical Sciences
New Delhi
Draft Proposal
August 5, 2006
1
u
-
Integrated Management of Childhood Illnesses: Baseline
assessment of childhood morbidity & mortality in selected
districts in India
Investigating Team
Principal Investigator
Dr. Narendra K. Arora
Professor, Division of Gastroenterology, Hepatology & Nutrition,
Department of Paediatrics, AllMS
Team Leader, IPEN
Co-Principal Investigators
1
Dr C. S. Pandav
Professor and Head,
Department of Community Medicine,
All India Institute of Medical Sciences,
New Delhi.
2
Dr R.M. Pandey
Additional Professor,
Department of Biostatistics,
All India Institute of Medical Sciences,
New Delhi.
3
Dr Rema Devi
Associate Professor,
Department of Community Medicine,
Trivendram Medical College, Thiruvanantpuram.
4.
Dr. Rakesh Lodha
Assistant Professor,
Department of Pediatrics,
All India Institute of Medical Sciences,
New Delhi.
Draft Proposal
August 5', 2006
2
CONTENTS
I. Introduction
4
II. Aims and Objectives
7
III. Methodology
9
IV. Network dynamics
27
V. Expected Outcome
33
VI. Limitation
34
VII. Timeline
35
VIII. Annexure
36
Draft Proposal
August 5', 2006
3
I. INTRODUCTION
Every year some 2.4 million children die in India before reaching the age of 5 years.
Infant Mortality makes up for 70% of under-five deaths and currently stands at 63 per
1000 live births (SRS, 2004). Within the infant mortality, neonatal mortality component
constitutes about 65% of the total IMR. While IMR has halved in the past 30 years, that
progress is halting during the past decade. Malnutrition rates have also declined in past
30 years but remain high.
Main causes of Under-5 deaths (U5MR) death in sequence of importance are: diarrhea
(23%), pneumonia (23%), birth asphyxia (10%), prematurity (9%), severe infections in
newborn (9%), neonatal tetanus, other neonatal complications, measles and malaria.
Research estimates have shown that full coverage of key interventions can prevent 57%
of these annual 2.4 million deaths. The interventions include early & exclusive
breastfeeding, adequate complementary feeding, clean delivery, immunization against
tetanus toxoid, measles & HiB, zinc & vitamin A supplementation, newborn temperature
management, ORS use for diarrhea, antibiotics for sepsis and pneumonia.
•
Nutrition interventions (breastfeeding, complementary feeding, vitamin A and
zinc supplementation) can prevent 25% of deaths.
•
Case management interventions (ORT, antibiotics for diarrhea, pneumonia,
neonatal sepsis, and anti-malarial) can prevent 30% of deaths.
In India, as many as 1.72 million children die each year before reaching their first
birthday and of these 72% die during the neonatal period. NFHS data shows that only
16.5% of women were visited in the first two days of birth and 42% of deliveries were
conducted by skilled personnel.
The millennium development goal (MDG) for India envisages reducing infant and child
mortality by two- thirds between 1990 and 2015 (World Bank 2004). Research has shown
that low technology solution combined with political will and financial commitment
could save the vast majority of lives of mother and babies {State of the World's Mother
2006, Report ofSave the children).
Draft Proposal
August 5', 2006
4
Integrated Management of Neonatal Childhood Illnesses (IMNCI), the adapted version of
IMCI in India, involves putting these interventions in place, both at the family as well as
at the health facility level. IMNCI has emerged as a promising approach to deal with the
issues related to child survival in a more holistic manner and has shown promise in other
countries, where it has been implemented. This strategy was designed to include
coordinated activities within three components:
(1) Improving health worker skills,
(2) Improving community practices related to child health and development
(3) Strengthening of health system supports for child health activities.
The Ministry of Health, Government of India has agreed in principle to implement the
IMNCI strategy as part of the Reproductive and Child Health Programme - II (RCH-II).
However since the implementation of the IMNCI requires training a huge number of
health functionaries across the country as well as improvement in the overall health
systems, it is only prudent that the relevance, feasibility and effectiveness of this strategy
is studied in some detail before its countrywide implementation is effected.
IMNCI will be rolled out in a phased manner and the initial phase involves
implementation in 8 districts of the country. These districts will be carefully chosen to
represent major socio-cultural areas of the country so that decision to expand IMNCI in
the entire country could be taken based on the experience of implementation in these
districts.
Evaluation of effectiveness of the IMNCI strategy would require a sound and detailed
baseline assessment of key demographic parameters related to child survival. The
Ministry of Health, GOI and UNICEF - India mission have asked IndiaCLEN Program
Evaluation Network (IPEN). conduct a study to establish the baseline for key indicators
1
"
"
“
■
*" —
~
-
IM
l
related to child survival.
Why IndiaCLEN Program Evaluation Network (IPEN)
Currently, Clinical Epidemiology Units (CEUs) are functioning at eight medical colleges
in
India
(New
Draft Proposal
August 5‘, 2006
Delhi,
Lucknow,
Nagpur
(2),
Chennai
(2),
Vellore
and
5
Thiruvananthapuram). CEU faculty members have been trained in clinical epidemiology,
health social sciences, biostatistics and health economics under the global International
Clinical Epidemiology Network (INCLEN) program. Members of all the eight CEUs
have formed a national body called Indian Clinical Epidemiology Network (IndiaCLEN).
The main objectives of the body include disseminating the knowledge and skills of
clinical epidemiology to other academic and non-academic medical institutions in the
country and participating in policy relevant research activities. Since 1997, IndiaCLEN
expanded itself to partner with medical institutions and NGOs beyond the six CEUs and
form IndiaCLEN Program Evaluation Network (IPEN). In addition to participation in
several public health program evaluation studies, members of the group have had the
benefit of attending workshops on program evaluation, qualitative research methods and
continuous quality assurance.
Over a period of last 8 years, IndiaCLEN Program Evaluation Network (IPEN) has
emerged as a professional and cohesive network of 84 partner institutions (medical
colleges / NGOs / public health institutions). IndiaCLEN Program Evaluation Network
has supported program managers in refining the existing public health programs to make
them client friendly and in harmony with their socio-cultural beliefs. IPEN has
successfully evaluated the Pulse Polio Immunization Program for four consecutive cycles
[1997-98, 1998-99, 1999-2000 and 2000-2001] and was also involved in the evaluation
of three rounds of Family Health Awareness Campaign [1999, 2000, 2002] in the
country. This network has also completed the evaluation of Vitamin-A and Iron folic acid
Supplementation Program(s) (2001-2002). It is heartening to note that policy makers
r*--
have incorporated several key recommendations made by this network in the subsequent
cycles of the respective programs. Most recently, nation-wide Assessment of Injection
Practices and Routine Immunization had been conducted.
Apart from undertaking research activities, IPEN has tried to develop capacity of its
network partners to undertake policy and program relevant studies in their respective
regions and states independently.
Draft Proposal
August 5", 2006
6
II AIMS
•
To determine the baseline morbidity and mortality of children under five years.
(Phase I)
•
To capture the process of implementation and the different strategies adopted by
the state health department in implementation of IMNCI. (Phase II)
•
To evaluate impact of IMNCI package on child survival indicators. (Phase III)
OBJECTIVES
Objectives of Baseline study: Phase I
1.
To assess child survival indicators
i. Mortality rate, establish causes, and pathway analysis of events prior
to death
ii. Morbidity prevalence and pathway analysis of events prior to
improvement / recovery
2.
To assess sickness management practices
i. Household Level
ii. At community
iii. At health facility level (Out Patient)
3.
To assess health behavioral practices
i. Household Level
ii. At community
iii. At health facility level
4.
To assess the skills and care providing competencies of the health care providers
(Public and Private)
i. Doctors
ii. Heath Worker .
iii. AWW/ANM
iv. Other Community level Service Provider, including ASHA (wherever
in place)
5.
To assess the perspective of policy makers, program managers and health
providers at district and state level about sickness
Draft Proposal
August 5' \ 2006
7
6.
To assess health system support for
i. Logistics
ii. Referral mechanism
iii. Intersectoral coordination
iv. Social mobilization
v. Monitoring & supervision
7.
To assess the client and health provider perspective on existing child health
services in terms of their
’
i. Availability,
ii. Accessibility,
iii. Perceived Affordability,
iv. Appropriateness,
v. Quality of care and
vi. Socio-cultural acceptability.
8.
To determine the role of civil society (NGOs, Community Based Organizations,
. community leaders) in promoting, providing, utilizing and monitoring child health
services.
Draft Proposal
August 5', 2006
8
III. METHODOLOGY
The proposed study would be a field trial that would be completed in three phases: Phase
I - baseline survey; Phase II - concurrent observation/documentation of implementation;
and. Phase III - end line survey for impact assessment.
Phase I: Baseline Survey
Phase I of the study will make a baseline assessment of various parameters and
dimensions related to child survival and health systems in both the intervention and
control districts. These would involve derivation of estimates for childhood morbidity
due to the diseases covered under the IMNCI, the infant mortality rate and analyses of
causes of infant deaths.
Current practices and mechanisms involved in care of newborns and children at the
family/community, first facility and referral level will also be studied. Client perspective
on existing child health services in terms of their availability, affordability and
acceptability will be studied in detail.
The variables/dimensions for the .baseline study will be selected in such a manner that
their repeat assessment at the endline survey would enable us to capture the impact of
IMNCI at different levels in the study areas.
Phase II: Concurrent observation and documentation of the process of IMNCI
implementation in intervention districts.
The IMNCI envisages significant upgrade /inputs in the skill set of health providers at
different levels to improve management of childhood illnesses; improved logistics and
supplies; greater involvement of the community in child survival issues. It will be
important to capture the process of implementation and the different strategies adopted by
the health department in implementation of IMNCI.
Thus, this phase of the study will involve in-depth observation and documentation of the
various facets of the implementation process. It is important to emphasize here that the
Draft Proposal
August 5', 2006
9
study investigators will not be involved in the process of either implementation or
monitoring of the IMNCI so as to minimize bias.
Phase III: End line survey for impact evaluation
The aim of this phase will be to assess the impact of IMNCI on various parameters of
child .survival. It will also involve assessment of these parameters in the control districts.
The data will then be analyzed to assess the effects of implementation and non
implementation of IMNCI on child survival.
The variables used for assessment in the baseline survey will be replicated in the end line
survey as well to allow for comparability.
1. Study Design for Phase I (Baseline Survey)
It is a baseline assessment of existing child care services in the community and will be a
cross-sectional survey using qualitative and quantitative research method. For reasons of
cost and logistics, rapid population based cluster survey at household and health facility
level will be adopted. The sample communities will be selected using the ‘probability
proportionate to size’ (PPS) technique. Using this method, the likelihood of a community
being selected is in relation to the proportion of its population i.e. larger villages or towns
are more likely to be selected than smaller ones. This study will be conducted in 16
districts representing 8 states. The sampling universe will be the district for drawing
clusters.
Census data of 2001 (Registrar General of India) will be utilized for selecting clusters
after adjusting for population growth rate. The clusters will be drawn separately for urban
and rural areas due to inherent differences in the available health facilities, infrastructure
of urban and rural areas and differences in socio-cultural and demographic features of
urban and rural populations.
This rapid survey would be used to derive estimates for causes of childhood morbidity,
NMR and IMR.
Draft Proposal
August 5', 2006
10
2. Identification of Study States and Districts
Study districts have been identified in consultation with the Division of Child Health,
Ministry of Health and Family Welfare, Government of India, State Governments and
UNICEF. Two districts have been identified in each state for the study. The following
selection criteria were considered while choosing the study states and districts.
Selection of the study states
1
Four Empowered Action Group (EAG) states (Uttar Pradesh, Madhya Pradesh,
Rajasthan, and Orissa); three non-EAG states (Karnataka, Haryana, Maharashtra) and
one northeastern state (Meghalaya). The purpose is to sample a mix of states with
different level of health system performance, health statistics and geographic location.
2
Demographic profile of the states
2.1. Recently formed EAG states not included
2.2. Non-EAG states dispersed over all regions of the country
2.3. States with good performance and low IMR (<40/1000 LB) were excluded
because of potentially less visible impact of IMNCI implementation
2.4. Availability of functional IndiaCLEN Program Evaluation Network partner
institution in the state to participate in the study.
Selection of the study districts
In order to minimize influence of status of health systems and governance variables a pair
of districts is selected from each selected study states. While selecting districts, factors
such as density of population, SC, ST Population, minority population, illiteracy rate,
IMR etc. were considered, to have comparable population to the extent possible and later
draw meaningful inference to assess the impact of IMNCI. Based on the recent available
demographic and health service indicator data the district selection was based on
1. Poorly performing districts with IMR >50/1000 live birth and similar social and
demographic features.
2. Non-contiguous districts
3. Consent of State Health Departments (planned IMNCI implementation districts)
4. Availability of IPEN- partner institution for implementation and supervision
Draft Proposal
August 5', 2006
11
Figure 1: Selected Districts and States for IMNCI:
Baseline Study of Childhood Mortality and Morbidity
Draft Proposal
August 51,2006
12
Table 1 - Study and Control Districts for Baseline Survey
Study districts
Control districts
State status
State
EAG States
Uttar Pradesh
Kanauj
Mathura
Madhya Pradesh
Morena
Tikamgarh
Rajasthan
Baran
Chittorgarh
Orissa
Nayagarh
Sonpur
Maharashtra
Amravati
Parbhani
Karnataka
Gulbarga
Gadag
Haryana
Kaithal
Mewat
Meghalaya
East Garo
Jaintia Hills
Non-EAG States
North-eastern State
•
Study districts: The districts for immediate implementation of IMNCI program.
•
Control districts: The districts with ongoing RCH I child health activities and
IMNCI program to be implemented after full operationalization of IMNCI in
study districts for 2 years.
The total population of eight selected states is 39.6 crore (census 2001). Total population
of districts under study is 2.14 crore; IMNCI will be implemented in 1.25 crore
population of phase-1 districts (study area) and 0.9 crore is the population in control
districts.
In case of a natural disaster (Flood, Famine/ earthquake/ riots) additional baseline survey
will be done in the affected area within a span of 6 months. This will be done to take care
of setback to health system due to natural disaster and this second baseline will serve as
the comparison parameter for end line survey. It is assumed that Industrial and economic
interventions are unlikely to affect major impact of the study within the study period.
NGO/ CBO and other civil society activities related to health in the area will be
documented and lists will be updated on yearly basis. Thus all major events, activities
that are likely to impact health systems particularly in relation to child survival will be
documented annually.
Draft Proposal
August 5", 2006
13
3. Study Procedure
3.1. Investigation Team
3.1.1. Investigation Team for Quantitative Component
Each survey team will consist of six members: two (SI) senior investigators (faculty from
partner medical colleges), two doctors or interns (MO) and two research associates (RA).
In every district eight survey teams will be constituted, each team will survey 10 clusters
thereby approximately eight teams will be covering 80 clusters (40 urban & 40 rural) in
the selected districts.
Table 2 - Survey Team
(2 Senior investigators, 2 medical officers, 2 research associates per team)
No. of persons (6 Per Team)
No. of Teams
Level
SI
MO
RA
Total
District
8
16
16
16
48
Total per state
16
32
32
32
96
128
256
256
256
768
(2 districts)
Total in 16 districts
spread over 8 states
Table 3 - Job responsibility of each team member
Members
First Senior
Investigator
Second Senior
Investigator
_________ Work load_________
1. 10 under 5 children per facility
(40 Govt and private health
facilities / district)
2. 40 Govt and private health
facilities / district
3. All instruments to be verified
for quality
4. Tracking of a neonate________
1. All death cases.
1. Conduct Verbal Autopsy
2. All deaths, 2 recovered non
2. Tracking of events (Deaths / recovered
hospitalized sick children, and 1
illness/ newborn).
recovered hospitalized sick child
3. Scrutinize and complete all instruments/
3. All instruments to be verified for
Quality Assurance
quality
___________ Responsibility__________
1. Observation of prescriber client
interaction (specific observation)
2. Interview of the doctor
3. Generic Observation of Health Facility
4. Scrutinize and complete all
instruments/ Quality Assurance
Draft Proposal
August 5', 2006
14
First Doctor +
First Research
Associate
Second Doctor
+
Second
Research
Associate
1. Screening of households
2. Observation of skills of
ANM/AWW/Depot holder
3. Screening of households
4. Inform 2nd PI about all deaths,
recovered illness; newborn, and
sickness U5____________________
1. Screening of households
2. Interview ANM/AWW/Depot holder
3. Inform 2nd PI about all deaths,
recovered illness, newborn, and
sickness
1. One per cluster
2. One per cluster
3. 7 households
1. 7 households
2. Three per cluster
3.1.2. Investigation Team for Qualitative Component
Qualitative methods would be used to assess health care seeking practices; client
perspectives in existing child health services; the role of civil society with regards to
prevention and management of childhood illnesses. A separate team consisting of one
Regional Coordinator, one Principal Investigator (PI) and two Research assistants (RAs)
will be responsible for the qualitative component. All the interviews and FGDs will be
recorded.
3.2. Sample size calculation for Quantitative Component
3.2.1. Mortality Survey
Sample size for cluster survey, was calculated on the basis available mortality indicators
(U5MR, IMR, NMR) related to child health. Mortality is the least frequently occurring
outcome (event) and hence we calculated the sample size according to mortality indices.
It is estimated that neonatal mortality rate (NMR) is 30-50/1000 live births (3-5% of all
live births) and infant mortality rate (IMR) is 60-80/1000 live births (6-8%), U5MR is
80-100/1000 live births.. To determine the NMR (4%) with an admissible error of +/- 1%
with 95% confidence, we would require to survey approximately 1500 live birth in
previous one year. This will allow detection of IMR (7%±1.3%), NMR (4%±1%) and
under 5 mortality (9%± 1.5%) with acceptable precision (Table 1).
Draft Proposal
August 5', 2006
15
Table 4-Sample size for estimating NMR and IMR
S.N
Indicator
Estimated
Admissible
Sample size
Estimated sample
error
(at 95% confidence
size
level)
1
Neonatal Mortality
4%
± 1%
1454
7%
± 1.3%
1478
9%
±1.5%
1498
Rate (NMR)
2
Infant
Mortality
^1500 live births
Rate (IMR)
3
Under 5 mortality
(U5MR)
We are proposing to take 80 clusters from a^district (40 each from urban and rural areas)
To obtain the desired sample size, we shall recruit 20Jive births taking place in previous
12 months in each cluster. Thus we will consider 1600 live births in each district.
3.2.2. Morbidity Survey
One of the objectives of the study is to find out the morbidity, density in the study
population and pathway analysis of events prior to recovery. IMNCI considers 3
important morbidities - cough (ARI), fever (Malaria) and loose motion (diarrhea). These
three indicator conditions will be used for tracking morbidity impact of IMNCI
programme and hence for the baseline survey. According to NFHS II (1998-99^ the all
India prevalence of cough with fast breathing (ARI), fever and loose motions was 19.3%,
29.5% and 19.2% respectively. Among the 8 study states the least prevalence of these
three morbidities was 7.9% (ARI), 23.7% (fever) and 13.9% (diarrhea). UNICEF and
MOHFW-GOI conducted multi indicator cluster survey in 2000, covering all under 5
children. According to this, the least prevalence of these indicator conditions is 18.8%
(ARI), 21.3% (fever) and 14.5% (diarrhea) in the eight states chosen for the study
(Table 3).
Draft Proposal
August 5,f,t 2006
16
Table 5 - Prevalence of child morbidity in India
All India
NFHS II -1998-99
MICS- 2000
(0-3 years)
(0- 5 years)
Cough
Fever
Diarrhea
Cough
Fever
Diarrhea
19.3
29.5
19.2
29.1
29.8
23.1
7.9
23.7
13.9
18.8
21.3
14.5
«8
«24
«14
«19
«21
«15
( % prevalence)
Least in 8 states
( % prevalence)
( % prevalence)
Table 6-Sample size for estimating morbidity density
S.
Indicator
Source
Prevalence
No
Admissible
Sample size at
Sample
Design
Estimated
error (20%)
95% Confidence
size
effect
sample
size
level
1
2
3
Diarrhea
ARI
Fever
NFHS II
14
±2.8
590
MICS
15
±3
544
NFHS II
8
± 1.6
MICS
19
±3.8
615
NFHS II
24
±4.8
304
MICS
21
±4.2
361
615
923
1.5
«l,120
Since there is a wide variation in the prevalence of ARI reported in both studies (NFHS II
- 8 %, MICS - 19%), we considered the mean prevalence to be 13.5 % and calculated the
sample size with 20 % admissible error. Assuming a design effect of 1.5, we need to
study approximately 923 under 5 children to estimate morbidity density. We will be
taking 80 clusters from a district (40 each from urban and rural areas). To obtain the
desired sample size, we shall recruit 14 under 5 children in each cluster. So, we propose
to study a total of 1120 under 5 children to find out the morbidity density.
Draft Proposal
August 5, 2006
17
Table 7-Sample size for quantitative survey
Cluster Live
births*
Expected
No. of under 5 children to be
Mortality*
screened for morbidity density
(@ 20 per (@lper cluster)
(@14 per cluster)
cluster)
District
80
1,600
80
1,120
State(2 districts)
160
3,200
160
2,240
National (8 states)
1,280
25,600
1,280
17,920
* Live births and mortality will be considered in previous 12 months.
3.3 Sampling Technique for Quantitative Component
3.3.1. Methods of Drawing Clusters
Sampling universe will be one district from each state, which will be stratified into urban
and rural areas. The data from census 2001 will be arranged in three columns: the first
column lists the names of the villages or towns and cities [i.e. the community], second
column indicates the total population of the communities and third column contains the
cumulative population which is obtained by adding the population of all the communities
preceding it on the list. The list is arranged in order given in the national census data. The
sampling interval is obtained by dividing the total population of the district by the
number of clusters desired. A random number between 1 and the sampling interval is
chosen as the starting point and the sampling interval is added sequentially to the random
number until the desired number of clusters is chosen. The villages / towns whose
cumulative population includes these numbers are chosen for the cluster survey. The
selected clusters are plotted on a map of the respective zone, and a logical sequence
(route map) for the fieldwork is developed for each of the survey teams.
3.3.2 Selection of study participants
After reaching each selected cluster, the cluster (village) is mapped and the following
steps are to be followed.
STEP J:
Draft Proposal
August 51,2006
Division of cluster
18
Cluster will be divided into two equal halves. Teams will also be divided
into two groups and each group will have one doctor with one research
associate. Two groups will move in opposite direction and will follow
each step given below separately.
STEP 2:
Selection of Mohallas/ colonies
A list of all mohallas /colonies/ tolas/ lanes in each half of the cluster will
be prepared. A mohalla will be selected randomly using currency note
available with the investigator 1.
STEP 3:
Selection of Household
Following the selection of mohallas /colonies/ tolas/ lanes, first household
will be selected randomly by using currency note available with the
investigator 2.
STEP 4. a:
Screening households for live births in last one year
All pregnancy culminating in last one year in the cluster, will be screened
by using screening instruments. The screening will continue till 10 live
births are included in the survey in each half of the cluster.
(For example if the date of survey is 25th October 2006, all the
pregnancies culminated from 25 th Oct 2005 till 24th Oct 2006 will be
included in the study).
STEP 4.b:
Screening households for child death
During screening each team will inquire about any death of child under 5
years in previous one year.
STEP 4.c:
Screening for neonate
During screening for live births, any neonate between age of 10 -28 days
as on date of survey will be identified and accordingly inform principal
investigator will be informed for tracking.
STEP 4.d:
Screening for morbidity in children under 5 years
First seven male and seven female children under 5 years will be identified
by each team in their respective direction of household survey.
STEP 4. e:
Screening for recovered morbidity in children under 5 years who
were non-hospitalized
Draft Proposal
August 5', 2006
19
In addition to this, each team will identify one recovered case of morbidity
in their respective area (i.e. 1 male/1 female in each half of the cluster)
Children under 5 year with illness in previous two weeks before survey,
who recovered without hospitalization and are asymptomatic for previous
3 days from any illness will be included in survey.
STEP 5:
Screening for recovered morbidity in children under 5 years who
were hospitalized
In addition to this, each team will identify one recovered case of morbidity
in their respective area, which was hospitalized for any illness in previous
3 months. Only one such case (preferably female) will be included in the
survey in each cluster (as per the discretion of principal investigator of the
team).
3.4. Data Collection
3.4.1. Data Collection and Transmission for Quantitative Study
Data will be collected from forty rural and forty urban clusters each from sixteen
districts. Collected data would be checked for completeness and quality at the study
centers as well as regional centers. The regional centers would then send the data every
fortnight to the CCO, Delhi. The Questionnaires and Observation Checklists are assigned
unique numbers facilitating their identification such as the center, cluster and category
they belong to. Within 72 hours of data collection; the team will make one set of
photocopy of the completed instruments. While the original set is dispatched to the CCO,
the other will be retained by the survey team and submitted to the Regional Coordinator
at the end of the survey after completing all ten clusters. This is done to safe guard
against accidental data loss during transmission.
3.4.2. Data Collection for Qualitative Study
District: At all 16 districts, all categories of stakeholders will be interviewed.
Stakeholders at the district level from both rural and urban settings will be interviewed at
all the study sites. In rural areas, population and officials of a PHC will be the sampling
frame. For urban areas reference point will be the government hospital.
Draft Proposal
August 5", 2006
20
The various providers and clients will be identified in reference to these dispensary or
PHC areas respectively.
A prior written consent will be obtained from all the study participants (interviewees) as
per a format provided with questionnaires. Signatures of those refusing consent will also
be obtained along with basic information about them
Efforts will be made to record all the interviews on the audiotape. In all situations,
responses will be recorded in interview schedules verbatim (local language) as the
interview is in progress, supplemented later by the tape recorded versions, and there after
translated into English. An ‘insider's view’ will be obtained by ensuring that investigators
are 'local'.
3.5. Study Instruments
A team of program evaluation experts, social scientists, epidemiologists and
anthropologists at the central coordinating office conducted several brain storming
sessions to identify key issues to be evaluated, keeping in mind the interest of policy
makers, program managers, implementers and clients. Interview schedules would
comprise of open-ended questions. The guiding principle will be that the responses
should be able to achieve the stated objectives of the study. The draft instrument will be
developed by the Team Leader, Principal Investigator, Project Coordinator and other
CCT members in close partnership with the program managers, and the investigators
from all participating institutions. The instruments will be translated into local languages
and pilot tested at 4-5 centres and finalized during the National Protocol Finalization
Workshop.
The interview schedules and observation checklists are developed keeping in mind the
objectives of the study. The interview and observation schedules will have close-ended
questions.
Focus group discussion topic guides
Guidelines for focus group discussions will be prepared in tune with the objectives of the
study.
Draft Proposal
August 5'\ 2006
21
Unique serial number: Every interview schedule will be given a unique serial number
and every participating institute will have a set of unique numbers assigned to it. Thus
from unique serial number- the institutions, rural or urban and category of stakeholder
could be tracked.
3.5.1. Verbal Autopsy and Tracking of Events for death:
In order to establish the cause of death, those households where neonate/ infant/ child
death had occurred within the last one year from the date survey would be visited by one
of the senior investigator. Senior Investigator would visit these households and establish
the cause(s) of death by using standard pre-tested verbal autopsy questionnaire. Pathway
analysis will be undertaken to dissect the sickness seeking behavior of the household. If
no caretaker is available repeat attempt will be made in next 48 hrs as per the convenient
time for the family.
Selection of clients for verbal autopsy:
■
All deaths in the cluster will be included indhe study sample for verbal autopsy and
tracking of events.
3.5.2. Tracking of events
•
For Recovered Morbidity
Two households with children with illness in previous two weeks before survey and who
have recovered (symptomatic for at least 3 days prior to survey) will be included in the
study for comparison of sickness behavior related to illness resulting in recovery or death.
In this, children who have been hospitalized within last six months (severe
morbidity) and recovered and the children who have not been hospitalized (mild
recovery) and recovered will be included to track for the severity of illness. The
pattern of sickness behavior related to illness will be compared with that related to child
death.
Table 8-Criteria of Selection of under 5 children in household survey
S.N Age group
Draft Proposal
August 5', 2006
No. of children
1
2-12 months
1
2
13-36 months
4
3
37-59 months
2
22
•
For New Borns
During screening one neonate per cluster in the age group of 10-28 days will be evaluated
for their care seeking behavior. It is assumed that at least one neonate will be present in a
cluster at the time of.the survey.
3.5.3. Household Survey
Fourteen households in each cluster will be studied for childhood morbidity related health
behavior and sickness behavior of the households. 7 households with male index child
and 7 households with female index child will be taken for survey.
3.5.4. Health facility observation
In a district, total of 20 urban and 20 rural health facilities will be selected for generic
observation. Two health facilities (1-government and 1- private health facility) will be
selected for the survey for every alternate cluster. Observation of health facility will also
be made in terms of available facility infrastructure/ available manpower, logistics, and
referral services.
Selecting Health Facilities for Observation
Government'.
Government Health Facility (PHC / CHC / district hospital which
are not attached /linked to medical college) located in (or nearest
to) the selected cluster. With OPD facility will be selected for
direct observation and prescriber interview.
Private:
Private Health Facility, which is located in (or nearest to) the
selected (formal / Informal Practitioner) cluster where the newborn
care or childcare OPD is available will be selected. (If more than
one private health facility exists in the cluster, select one, who
• treats more number of children.
3.5.5. Assessment of skills of health care provider
Draft Proposal
August 5', 2006
23
One doctor, delivering child health services, will be selected for assessment of skills from
each of the selected health facility (government /private). He will be directly observed by
one of the senior investigator for appropriateness of treatment for management of 5 cases.
Out of 80 clusters in a district, we propose to observe the skills of health care provider
from 40 health facilities (20 urban, 20 rural). Out of these, 50% health care providers will
be from government and rest will be from the private sector.
3.5.6. Assessment of skills of paramedical workers
Skills of ANM/AWW will be assessed for management of common childhood illness by
the doctor of the team. Management of one newborn and one child under five will be
observed for each worker (ANM/AWW). During the time of survey the district
authorities will be requested to coordinate availability of ANM/AWW in their respective
villages. The cases for this purpose will be made available from the same cluster. After
observing the skills of ANM/AWW the doctor in the survey team will advice regarding
management. If on first day the team is unable to find the cases for observation, the team
will complete the activity on the next day.
Out of 80 clusters in a district, we propose to observe the skills of one ANM and AWW
in alternate cluster (40 clusters for ANM and 40 clusters AWW). In urban areas, instead
of ANMs lady health visitor (LHV)/ public health nurse (PHN) will be considered.
3.5.7. In-depth interviews
Selection of stakeholders for In-depth Interview
District would be the sampling unit. The purposive sampling will be done keeping in
view the convenience of data collection. It is important to interview all stakeholders i.e.
the health care provider, community leaders and clients to understand and assess the
situation. Stakeholder will be selected from both urban and rural areas.
Providers: Health and non-health officials at the district level and block level i.e. Chief
medical Officer, RCH officer, District MS/ Pediatrician and District magistrate at district
level.
Facilitators: Representatives from local NGO/ CBO and Community Leader.
Draft Proposal
August 5', 2006
24
3.5.8. Focus group discussions (FGDs)
FGD’s will be conducted among the community health worker and health care provider
to assess the health promotional and sickness behavior, client and health provider
perspective about existing child health activities and problems related to them. In each
district 4 FGDs with mothers of target children and 2 FGD each with health workers,
ANM and AWW will be conducted.
3.6. District Level Workshops
Sixteen district level workshops would be organized, where research teams and
investigators from each district would participate. The aim of these workshops is to
familiarize the research teams about the study instruments and other operational details.
Two CCT members also participate in each workshop. These workshops act as a quality
assurance mechanism, thereby ensuring that the entire research staff for the study
operates out of the same principles.
3.7.Data analysis and reporting
Data would be collected, collated and organized at the CCO, Delhi. A multidisciplinary
team guided by the CCT and followed by report writing would do the data analysis.
Draft Proposal
August 5', 2006
25
MORTALITY STUDY
S.N.
1.
Verbal Autopsy
2.
Tracking of events for
death_________________
3.
Tracking of newborn
4.
Tracking of events for
sickness (2 per cluster- one
hospitalized and one non
hospitalized)___________
5.
Household survey (14 per
cluster - 50% males and
females.each)
OBSERVATION
1.
Health facilities
(Both govt./private)
(One facility every
alternate cluster)________
2.
Prescriber client
interaction. (5 cases per
prescriber)____________
3.
Doctors Interview at the •
health facility (one every
alternate cluster)________
4.
ANM
(2 cases observed per
ANM/AWW per cluster)
5.
AWW/AWW Interview
(one per cluster)
A
Depot holder
INTERVIEW
1.
Prescriber
(both govt./private)
(one facility every
alternate cluster)
2.
ANM/AWW
DM____________
3.
4.
DHO/CMQ/RCHO
5^
NGO/CBO______
6.
Community leader
FOCUS GROUP DISCUSSION
_L
Mothers
2_
ANM
3
AWW
TABLE 9 - Summary of Activities
TOTAL/
DISTRICT
(80 clusters)
STATE
(2 districts)
TOTAL
(8 states)
Actual
Actual
Actual
160
320
2560
40
80
80
160
160
320
1280
2560
560
560
1120
2240
17920
20
(10 govt.
1 Oprivate)
20
(10 govt.
1 Oprivate)
40
80
640
20
(10 govt.
1 Oprivate)
20
20
(10 govt.
1 Oprivate)
20
40
80
640
40
80
640
20
20
40
80
640
20
20
40
80
640
5
10
80
ACITIVITY
CLUSTERS
URBAN
All deaths <5
yrs
80
RURAL
All deaths
<5 yrs
80
40
80
20
(10 govt.
1 Oprivate)
20
(10 govt.
1 Oprivate)
20
20
40
80
40
1
I
4
1
640
80
2
2
8
2
640
4
4
16
4
4
64
32
32
£
2
V Network Dynamics
Draft Proposal
August 5,f\ 2006
26
1. Network Dynamics: The objective of network dynamics will be to:
•
Maintain high level of motivation and commitment to the project activities
•
Maintain highest level of quality assurance and
•
Foster camaraderie among all partners
2. Network Structure and Dynamics
2.1. Constitution of the Central Coordination Team
The central coordinating team (CCT) is a multidisciplinary team comprising of 10-12
investigators co-opted from various medical colleges and professional institute. The
Central Coordinating Team has guided in all the previous IPEN studies. The CCT
operates from the Central Coordinating Office (CCO). The CCT holds several meetings
to arrive at a draft protocol for the study and to discuss other operational details. The
CCT will facilitate development of operational manual, study instruments, undertake
quality assurance visits to the study sites, supervise and conduct a focus group discussion
and provide inputs at the time of project report writing.
2.2 Central Coordinating Office - CCO, Delhi
Central Coordinating Office (CCO) of the study is located at International Clinical
Epidemiology Network office, New Delhi. The CCT and IPEN office will provide
technical and quality assurance support for the study. CCO will manage the network
activities and will be responsible for coordinating with all partner medical colleges,
monitoring progress of the network, screening the data to ensure quality, processing and
analyzing the data. Queries from partner medical colleges will be entertained and
responded by CCO.
2.3 Regional Centers
The study will be conducted in 16 districts representing eight states. Eight institution
centers will be nominated as a regional center in each state. The regional coordinator
along with the investigator from other medical college in vicinity to eight districts and
CCT member will implement the study. The regional center will coordinate data
collection and quality assurance activities of both the centers.
Draft Proposal
August 5', 2006
27
2.4 Partner Institutions
There will be 5 to 6 partner medical college involved in each state. Each partner medical
college will be responsible for formation of 2-3 survey teams. These network partners are
medical colleges or NGO’s or institutions working in the field of health.
2.5 Field teams
Survey Teams at Partner Institutions: In every district eight survey teams will be
constituted, each team will survey 10 clusters thereby approximately eight teams will be
covering 80 clusters (40 urban & 40 rural) in the districts. Each survey team will consist
of six members: two senior investigators, two doctors or inters and two research
associates (if possible one male and one female). All centers are encouraged to co-opt the
social scientist / anthropologist of the institution into the team. Research associates with
social science or social work background will be preferred. Thus it will be a unique
opportunity when senior faculty members along with the whole research teams will be in
the field to undertake data collection of highest quality.
2.6 Teams at Regional Center: In addition to field teams, one Principle Investigator and
two additional research associates will be provided to Regional Coordinator at each
regional center for a quality assurance and organize a interview at district level.
Fastest mode of communication will be maintained between all partner medical colleges
and CCO to clarify any doubts regarding various components related to baseline survey.
In addition, the study centers will work according to a fixed time line with close and
regular monitoring of all activities at partner medical colleges by regional centers and
CCO. Deadline will be strictly adhered to.
Movement of survey teams in the field
Draft Proposal
August 5', 2006
28
Each team will cover 10 clusters. This would include data collection in both urban and
rural clusters. It is estimated that each team would be able to cover one cluster in two
days and 10 clusters in about 20 days time.
3. Network Monitoring
3.1 Regional coordinator: The regional coordinator will be in constant touch with the
CCO, Delhi. Soon after the regional workshop, the plan of field activities will be faxed to
CCO. Regional coordinator will speak to CCO, Delhi every alternate day to update the
progress till that day. They will also fax the details of network progress of their Partner
Medical College’s (PMC) to CCO on every Monday, Wednesday and Friday and
communicate the problems faced by any team.
To coordinate the activities in each state, each regional centre will have a regional
coordinator. The tasks of regional coordinator will be to
1. To conduct and Organize a three-day orientation workshop in two-selected district
separately in coordination with the PMC’s to understand the objectives and methodology
including hands on training for all team members to conduct the interviews and health
facility observation. In hands on training selected clusters are not included.
2. Help / guide the PMC to prepare a route map for every team to cover ten clusters per
team.
3. Funds will be transferred to regional coordinator office from CCO. It is the
responsibility of the regional coordinator to disburse funds for the field travel.
4. Coordinate movements of all teams, solving problems and facilitating local
arrangements.
5. Identify triggers of a problem and taking pre-emptive actions.
6. Monitor movements of teams on the field and facilitate quality assurance visit by CCT
member.
7. Facilitate smooth data collection and data transmission.
8. To conduct the district level interviews (DHO, MO, RCHO) and conduct FGD’s.
Draft Proposal
August 5', 2006
29
3.3 Monitoring network progress: All partner medical colleges will prepare an activity
planner to accomplish the targeted number of interviews within the stipulated time frame.
This will be closely monitored through weekly interview log sheets. The network
progress regarding interview accomplished during the week will be monitored at CCO.
Whenever a partner medical college defaults in sending the weekly interview log sheets,
contact will be made through telephone/fax/email to remind them or to know if there was
a problem. Such phone calls will be entered in monitoring sheets. In addition, the medical
college will dispatch the completed interview schedule to their regional centers after
initial quality checks. After the regional center finishes with the quality checks, these will
be forwarded to CCO.
3.4 The CCT member who is out in the field to ensure quality will contact the CCO
Delhi, every evening regarding
■
Details of cluster survey
■
Any problem in methodology
■
Corrective measure taken
■
Plan for the next day
4. Network Communication
This would be critical for operating the network. Its significance would be emphasized
during all contacts and correspondence with regional centers and partner institutions. A
network directory containing office and residential addresses along with the telephone
and fax numbers of senior investigators will be compiled and circulated to every member
of the network with the explicit request of establishing contact with their regional
coordinating centre or central coordinating office in case of need at any time. Motto of
the network communication is to use “the fastest mode of communication”. Decisions
will be made quickly and no matter would be kept pending for more than 24 hours at the
coordinating center. All network letters and completed study instruments will be mailed
through speed post / courier, ensuring quick and safe delivery.
5. Quality Assurance
Draft Proposal
August 5', 2006
30
Understanding the limitations of multi-centric project, quality assurance measures will be
incorporated at every stage of the study. This will be done at four levels:
Level 1: Program Managers Meeting: one day program manager meeting will be
organized for seeking partnership from the respective state government and to facilitate
the activities of the project.
Level 2: One day International Advisory board meeting will be held after the National
Workshop to share experiences and get technical inputs for the present study.
Level 3: National Protocol Finalization Workshop at Delhi: A 3-day workshop will be
held for all the investigators including CCT members, and partners from Ministry of
Health & Family Welfare, Government of India, donor agencies and other stakeholders.
Objectives of the national workshop will be to finalize the study protocol including the
study instruments. Modalities of data collection and transmission would also be finalized.
Level 4: Regional Orientation Workshops: Three day regional workshops will be
conducted by CCT members and regional coordinators for principal investigators and
research teams from partner institutions in the region. Data collection techniques and all
the study instruments are discussed in detail at these workshops. Participation of CCT
members ensures that there are no distortions or dilutions in communicating the study
protocol to the research teams across the network. All participants will have hands on
experience to conduct interviews and observation of health facilities in the near by village
under the close supervision of regional coordinator and CCT member.
Level 5: Quality Assurance visits: These will be undertaken by the central coordinating
team members during the data collection phase to supervise the process. Each CCT
member will visit at least four clusters in every district to assess two parameters
Authenticity of data that has actually been collected
■
Quality of interviews being conducted by the researchers though direct
observation
Level 6: After completing a cluster, the investigator sits with the RA and scrutinizes the
filled schedules for completeness and appropriate marking of the answer. If satisfied they
will counter sign the schedule or else the researcher will be asked to do an extra interview
to replace it.
Draft Proposal
August 5', 2006
31
Level 7: Quality check at regional centers: At each regional centre, all the study
instruments sent by the partner medical colleges would be checked for correctness and
completeness before transmission to CCO, Delhi.
Draft Proposal
August 5', 2006
32
VI. Expected outcomes of the study
Following are the expected outcomes of the study:
1. Baseline estimates for burden of childhood Illnesses covered under IMNCI (Neonatal
Sepsis, Cough, Fever, Diarrhea)
2. Baseline estimates for infant mortality in the study districts and causes of infant
mortality.
3. Existing sickness care seeking and referral pathways.
4. Determinants of health seeking behavior of community regarding preventing
childhood illnesses.
5. Availability of facilities, quality of care in public and private health care facilities in
the study districts.
Draft Proposal
August 5‘, 2006
33
VII. Limitations and challenges
The potential limitations of Rapid Appraisal Procedures (RAP) are both intrinsic to the
methodology and extrinsic for the purpose of generalization. The problems are primarily
related to:
(i) Accuracy of the information
(ii) Representative nature of the respondents and their responses.
(iii) Cultural appropriateness
(iv) Subjectivity of the investigators leading to a bias in the interviews and subsequent
interpretation of data.
Draft Proposal
August 5', 2006
34
VIII. Timeline of Baseline Survey
IMNCI Baseline surveyPhase I
TIME LINE
Activity
1 .Development of study
instruments and operational
manual
2. Field testing_________
3. Finalisation of Study
Instruments___________
4. Meeting of State Health
Secretaries and Program
Officers
^U<LY
AUGUST
2 3 4
1^.3^ MM 1
I
SEPT
1 2 3
4
OCT
1 2
NOV
1 2
DEC
3
4
liiiiiii
—r—
_________
3
__
_____
5.Identification of Regional
coordinator and PMC
6.Preparation of national level
worhshop
7.international advisory borad
meeting
8. National workshop______
9. Preparation of district level
workshop
10.District level workshop (16)
11 .Data collection__________
L
..
L
___
__
12. Data analysis___________
13. Draft report
Draft Proposal
August 5”, 2006
__ I :. ...
35
I
Annexure 1: Field Activity
STEP1
FIELD ACITIVITY
Quantitative survey
Qualitative survey
Visit the cluster, prepare
a rough map and divide
it into two halves
A team of
1 RC + 1 PI + 2 RAs
In each half of the cluster, list of all the
mohallas and identify the starting point
using currency note from PI -1
STEP 2
Identification of the first house
within the mohalla by using currency
note from PI-2
STEPS
____________ O______________ g
In each half of the cluster, the following
steps will be followed
STEP 4
u
n
V
Interviews of
DM, CDMO, Prescriber (Govt,
Pvt), NGO/ CBO, Community
Leader
FGDs
Mothers, community leaders,
CBOs, NGOs, SHGs, HWs
(ANM, AWW, MPW, ASHA)
__________
a) Screen households for live birth. (Each half continues screenings till 10 live
births in previous 1 year)
b) Screen for death in under 5 children in previous 1 year within the sampling frame
of households i.e. 20 live births (e.g. if survey on 25th Oct-06. So death between
26th Oct -2005 to 25th Oct -2006 will be included). The identified case will be
taken for verbal autopsy and tracking events of death by principal investigator 2.
c) Screen households for Neonate aged 10- 28 days. The identified neonate will be
taken for tracking events of early infancy by principal investigator 1
d) Screen for childhood morbidity in under 5 children. (First 7 male children, to be
taken in one direction and first 7 female children in the other direction). Conduct
household survey in each of them. Conducted by doctor and research associates.
e) After the seven cases (male/female) are completed then screen for recovered
morbidity (non -hospitalized) one case- in last two weeks and three days of
asymptomatic recovery.
STEPS
_____________ ___________________________________________
Recovered morbidity (hospitalized) one case - in last three months preference given to female
Observations
• Generic observation of health facility (govt/private)in every alternate cluster will
be don by principal investigator -1
• Specific observation of prescriber client interaction.-5 cases (under 5 children
suffering from any illness)- done by principal investigator -1
• Skill assessment of ANM AWW/ASHA — 1 case (any morbidity) and Neonate
(ill/ not ill )per cluster alternatively ( identified during survey in the cluster) done by doctor
36
Annexure 2: Study Instruments to be Utilized for Each Cluster
At Community Level
Component
Description
Tools
1
Mortality
All the deaths in the cluster will be
considered to establish the cause of
death
Screening Format,
Verbal Autopsy and
Tracking of Events.
2
Morbidity
Screening Format and
Household Survey
3
14 households per cluster will be
considered to assess the morbidity
density
Tracking of Events
3.1
Mortality (U5)
All deaths in the last one year
Tracking of Events for
Morality
3.2
Morbidity (U5)
3.3
Neonate
4
Assessment of
ANM/AWW
Hospitalized Child in the last
three months (1)
Non-hospitalized Child
(Recovered) (1)__________
Neonate (10-28 days) for
studying practices related to
neonatal care (1)
4.1
Skill
A sick child having symptoms such
as cough, fever and diarrhea will be
managed by ANM in presence of
senior investigator (1)
4.2
Knowledge, Practices
and Barriers
ANM/AWW will be interviewed (1)
______ ,_____________________________
Tracking of Events for
recovered Morbidity
Tracking of Events for
Neonate
Specific Observation
ANM/AWW interview
schedule
At Health Facility Level
5 Assessment of
5.1
Doctor’s Skills
5.2
Doctor’s Knowledge,
Practices and Barriers
5.3
Health Care Facility
Draft Proposal
August 5', 2006
For every alternate cluster, 1
government and 1 private health care
provider will be selected. (1)_______
Conducted by the doctor of the team
(D__________________
For every alternate cluster, 1
government and 1 private health
facility will be selected for generic
observation to assess the logistics
and equipment available for child
health (1)
Specific Observation
Doctor’s Interview
Schedule
Generic Observation
37
- Media
SDA-RF-CH-1A.20.pdf
Position: 6098 (1 views)