54

CHAPTER 3
METHODOLOGY

Studies to explore effect of PPPs face serious challenges. Health needs change with
time, demographic trends, technological advances and epidemiological contexts. The ability
to source data from all the stakeholders, reconcile their perspectives with the outcomes and
attribute the outcomes to the PPPs is a complex task. Health outcomes have multiple
determinants in addition to the physical availability and accessibility of health services.
Socio-economic background of the service users, the cultural and traditional context and the
perception of need as well as ability to exercise measures to address perceived needs also
have a role to play. Hence, a comprehensive assessment of PPPs and their impact on health
sector necessitates selection of PPPs with diverse characters. This would render the
applicability of findings and learning in other situations. There is a need to take into account
the views of all stakeholders associated with the PPPs in health, especially, the users of such
services. Their views have an important bearing on improving the reach and effect of the
health sector programmes.

The methodology adopted in this study is guided by the specific objectives and a
review of methodology followed in similar studies. The Chapter starts by specifying the
research objectives and research questions. It then presents conceptual framework that has
been used to explore responses to research questions and subsequently, outlines operational
definitions of various terms repeatedly used in the study. Then the study design is elaborated.
It starts with the sample of PPPs to be studied and the sources of information for studying
these PPPs. It then explains the data collection techniques adopted against the background of
methodologies adopted in other similar studies. It elaborates on the exact process of data
collection in the field, the tools used for data collection and the ethical considerations that
were kept in mind while collecting data. The data entry package and process followed are
included in the chapter followed by detailed description about the various variables and the
data analysis process followed. The chapter ends with statement about limitations of the
methodology adopted.

The main aim of the study was to explore effect of PPPs on reproductive health
services to women. The primary focus of the study was not to explore the effect in terms of
 55

conventional reproductive health outcomes, such as ante-natal care, institutional delivery,

postnatal care, safe abortion, management of gynaecological problems and contraceptive
acceptance but in terms of women‘s access, use and satisfaction with reproductive health
services.

3.2 Research Objectives

PPPs have been designed and implemented to improve the availability and access of
good quality reproductive health services, particularly for under-privileged, under-served
women. The objective was to explore the effect of PPPs on actual reproductive health service
delivery.

The overall goal of the study was to examine the theme of PPP as a way of addressing
the reproductive health of women and investigate whether the PPPs indeed measure up to the
inherent expectation that they would significantly improve reproductive health services.

The specific objectives of the study were

Objective 1: to examine the effect of select models of PPPs on availability, use, cost
and quality of health service delivery to women as perceived by them reproductive health
services to women as perceived by them.

Availability, access and use of reproductive health services to women, especially to

under-privileged served women through PPPs were studied. Additionally, an attempt was
made to understand their preferences, perceived quality of and satisfaction with the services
through PPPs.

Objective 2: to investigate the attributes of the three different models of PPPs that
account for this effect on availability, women‘s preference, access, use and satisfaction related
to reproductive health services.

The partnership attributes of interest identified for the study were the type of partners,
their motivation for seeking the partnership, their selection, roles and responsibilities in the
 56

PPP, the type of contract and the governance and administration structure of partnership.
Women‘s preferences, access, use and satisfaction were studied in the context of these
attributes. The relevant research questions were:

Questions 1: Who were the women who accessed and used these PPPs and what
guided their access and use?
Question 2: What reproductive health services did they seek and what were the ones
that they did not utilize or seek?
Question 3: What aspects of the RHS delivery were the women satisfied with?

Attributes of PPP are critical to successful and sustainable partnerships and these
attributes have an influence over the reproductive health service delivery to women,
particularly under-privileged, under-served women. The questions therefore are:

Question 4: What attributes of these PPPs had satisfied women or had a positive
impact on reproductive health of women?
Question 5: How had these attributes of PPPs contributed to improve accessibility,
efficiency, quality of RH services and use for under-privileged women?
Question 6: What was the contribution, role and function of each partner, the formal
management structure, system and processes for decision making, resource allocation and
oversight?

3.3 Conceptual Framework

The study looked at two aspects of the PPPs, the perceptions and experiences of
women in reproductive age group and the attributes of the PPPs that contribute to those
perceptions. The data was analysed to understand women who access PPP services for RH
reasons and their perceptions and experiences of using these. Perceptions and experiences
about availability, reach, acceptability, timeliness, effectiveness and quality of care of under-
privileged women formed the core of this analysis. The PPPs were also explored in detail to
understand the process of partnership, the motivation, role and contribution of each partner
and the governance, management and financial structures. A pictorial depiction of the
analysis framework is provided in Figure 3.1:
 57

Figure 3.1: Public Private Partnership in Health - Conceptual Framework

Positive RH outcomes
privileged
Under-privileged Non-under-privileged

Women in reproductive age group (15-45) years Stakeholders’ Perspective

Women’s’ Perspective  Efficiency
 Access  Quality attributes
 Use PPPs for Reproductive Health  Access
 Satisfaction Needs  Use
 Satisfaction

Operational management Attributes of PPP Organizational governance

 Health worker availability  Partners  Organisational capability
 Health worker competence  Contracts  Decision making
 Service package  Motivation  Capability support1
 Roles  Administrative/operational support2

Institutional mechanisms / environment

 National and state policies
 Legal & regulatory norms
 Health programmes

Political and Economic environment

 Political commitment
 Political leadership
 Budgetary allocations

Socio-cultural environment
 Norms
1: Leadership, motivation, skill development, oversight
 Traditional health seeking practices
2. Maintenance, cost control, risk management, conflict resolution
 58

3.4 Operational Definitions

PPP, access, use, quality of services and underprivileged are terms which are defined
in a multitude of ways. For instance in their study to understand paradigm of PPP in RCH,
Anand and Sinha (2010) used proximity to health facility, availability of doctor and
medicines, waiting time, cleanliness, treatment by staff and privacy as parameters that
reflected access and quality of services. Their study focused on use of any of the reproductive
health services for service utilization. Gill (2009) in her evaluation of service delivery under
NRHM used physical infrastructure, human resources, medicines, decentralized funding,
availability of services, distance travelled, waiting time, availability of medicines and patient
satisfaction as quality indicators. For this study, existing definitions of these terms were
referred to and used unaltered or adapted in the local context. The definitions and indicators
used for the purpose of this study are listed below.

Public Private Partnership: The definition of PPP as proposed by the Ministry of

Health and Family Welfare (MoHFW) was used as the partnerships studied were those
defined as PPP by them. PPPs are ‗collaborative efforts between public and private sectors,
with clearly identified partnership structures, shared objectives, and specified performance
indicators for delivery of a set of services in a stipulated time period‘ (MoHFW 2005).
―Public‖ means Government or organizations functioning under State budgets, ―Private‖
means the Profit/Non-profit/Voluntary sector and ―Partnership‖ means a collaborative effort
and reciprocal relationship between two parties with clear terms and conditions to achieve
mutually understood and agreed upon objectives following certain mechanisms(MoHFW,
2006).

Reproductive Health: WHO defines it as a state of complete physical, mental and

social well-being, and not merely the absence of reproductive disease or infirmity. For the
purpose of this study, however, reproductive health was largely limited (but not restricted) to
safe and healthy pregnancy and childbirth with right of access to appropriate health care
services that will enable women to go safely through these stages in her obstetric career.
 59

Access to services: Reachability, availability of personnel and supplies, acceptability

to women and affordability of services were considered to reflect the physical, financial or
psychological access to services.

Use of services: The extent to which women used reproductive health services i.e.
antenatal care, delivery, postnatal care, contraceptive and management of RTIs and
gynaecological problems in the reference period of one year prior to the survey was
considered as use of services.

Quality of services: Quality of services was defined in terms of whether women found
the services timely, effective, user friendly, needs oriented, and given with due explanations,
privacy and follow ups. Also if women preferred these services and the services were to their
satisfaction have been included to define quality of services as these reflect women‘s
perceptions about services.

Health care effectiveness: Health care effectiveness was defined in terms of

availability of staff and medicines to deliver services which are perceived as effective by
women.

Standard of living (SLI) index: This index defined in terms of ownership of

household goods used in NFHS II was adapted and scored as below:

1. House type: 4 for pucca, 2 for semi-pucca, 0 for kachha

2. Toilet facility: 4 for flush toilet, 2 for pit toilet, 0 for no facility
3. Electricity: 2 for electricity, 0 for no electricity
4. Source of drinking water: 2 for pipe & hand pump, 1 for public tap, 0 for other
source
5. Ownership of house: 2 for yes, 0 for no
6. Ownership of land: 2 for yes, 0 for no
7. Ownership of durable goods: 4 each for a car or tractor, 3 each for a
moped/scooter/motorcycle, telephone, refrigerator, or television, 2 each for a
bicycle, electric fan, radio/transistor and cart
Cumulative score ranged from 1 to 44.
Categories were: Low SLI: 1 to 14 score, Medium: 15 to 28 score, High SLI: 29 to 44
 60

Under-privileged / under-served women: Women in the reproductive age group (15

to 49 years of life or eligible women) from families that belonged to the category of ―Low‖
Standard of Living Index or Scheduled caste or tribe or possessed a BPL card were defined as
under-privileged women.

Attributes of PPP: The main attributes of PPP studied were its financial, governance
and operational management structures. The specific attributes considered under the study
were:

a. Partners and their selection: The type of partners and the process adopted for
their selection.
b. Motivation of partners: Motivation for seeking partnership in terms of
1. philanthropy,
2. fulfilment of social responsibility,
3. augmentation of resource inputs,
4. financial gains,
5. increase in scale of operations and cost of services
6. strengthening of own brand value, and
7. legitimacy was studied.
c. Roles and responsibilities of partners: Both partners‘ role and responsibilities
in
1. governance (strategic decisions),
2. management (provision of money, material, infrastructure or
manpower),
3. regulation (protect client‘s interest, control costs and ensure access and
equity in service delivery), and
4. the type of package, its scope; target population, quality and cost were
looked into.

Type of contract: The contract or Memorandum of Understanding between two

partners were scrutinized to explore whether these were classic or relational, occasional or
recurrent and complete or incomplete. The parameters used to assess the completeness were
inclusion in the contract of goals, intended outcomes, governance structure, organizational set
up, expected roles and responsibilities, timeline, payment mechanism, grievance redressal,
 61

conflict resolution, hiring/firing, performance assessment and contract renewal and

termination procedures

3.5 Study Design

PPPs are a complex development initiative. In their research to examine the

motivation for institutional PPPs and governance and managerial aspects that make these
effective, Cappelaro and Longo (2011) opted for a contextual and comparative case study
design. Such approaches are followed in many studies. The attempt in the study was to
explore and understand select PPP models and their contribution to address women‘s health
needs. Both quantitative and qualitative data was needed to understand the PPPs and their
outcomes through perspectives of the multiple stakeholders. The present study was planned
through mixed method case studies.

Mixed method case study: Mixed method case study offers the opportunity for an
empirical inquiry of emerging complex interventions in their social, political and cultural
context, uses multiple indicators and data sources for triangulation. It provides the
opportunity to collect empirical data using a combination of qualitative and quantitative data
collection techniques, apply both inductive and deductive analysis that involve descriptive
and numerical precision and provide results that are far more convincing. Data from multiple
sources in this approach enriches the analysis as it helps triangulation of results, illustrates
complementarity, and offers an opportunity to view the findings from new and different
perspectives. Therefore for a holistic, in-depth and contextual exploration of this complex
topic and issues around a mixed method case study appeared most appropriate. It offered the
platform to explore and describe the structure, processes and activities of the PPPs and the
way in which they affected women‘s reproductive health.

Sample of PPPs: Three PPPs implementing the reproductive and child programme at
the district, Community Health Centre (CHCs) and Primary Health Centre (PHCs) level 3
were selected for the study. Focus on reproductive health services for women, documented
successes, government‘s stated plans of upscaling, and feasibility of studying them guided the
selection of PPPs. Three PPPs that met this requirement were selected and finalized after
3
The PHC serves a population of 30 to 50,000 and provides primary health care. The CHC or the first referral
unit, serves a population of >1.2 lacs and provides secondary health care.
 62

consultation with officials of the Government of Gujarat about the availability of relevant
data and approvals to visit the facilities. These were the PHC at Dahej in Bharuch district run
by a corporate body, CHC Shamlaji in Sabarkantha district run by an NGO and facilities of
five private obstetricians empanelled under Chiranjeevi Yojana (CY, CY hospitals) in Surat
district. ―Control‖ facilities at same level of health service delivery in the same districts were
selected for comparative study. These were Tankari PHC in Bharuch, Prantij CHC in
Sabarkantha and facilities of five private obstetricians not empanelled under Chiranjeevi
Yojana (Non-CY hospitals) in Surat district. Picture 3.1 below indicates the location of the
selected PPPs in the State.

Picture 3.1: Location of Selected PPPs

Note: Districts where the selected PPPs are located

Sources of information: Sources of information or data were both primary and

secondary. Primary data sources were the facilities, the Health Service Providers (HSPs),
women who had used reproductive health services from the selected facilities and key
informants conversant with PPPs. Secondary data sources were the state government‘s Health
Management Information System (HMIS) and small studies conducted on the selected
facilities. Relevant government/ PPP documents and reports were also sources of
information.
 63

3.6 Data Collection Techniques

Review of data collection techniques: Before finalizing the data collection methods, a
review of various studies which have looked at the performance of reproductive health
service delivery, specifically through a PPP mechanism was undertaken to understand the
methods used by these researchers. The review was done with a view to consider the various
method used by other researchers in the context of their success or limitations in assessing
PPPs.

The review included ten studies conducted in the last five years (Acharya 2009; Bhat
et al. 2009; Bhimani et al. 2013; Blanco 2010; Kastia et al. 2010; Singh 2009; Deoki Nandan
et al. 2010; Ganguly et al. 2014; Jega 2007 and Sewa Rural 2003). All the studies were
conducted with the aim of exploring the respective Public Private Partnership model and its
effect on the partners and / or the beneficiaries. Eight of the studies focused on Chiranjeevi
Yojana. Two were on initiatives carried out by government in collaboration with NGOs and
the one was a study of social franchising mechanism across the State of Uttar Pradesh. The
sources for data were both primary and secondary. Primary data was collected using both
qualitative and quantitative methods. All the studies had interviewed beneficiaries of the PPP
models. Barring three studies, all the other studies had interviewed a purposive sample of
beneficiaries. In five studies, even the non-beneficiaries were interviewed to understand their
preference for other models of service delivery, mainly the government centres. The
qualitative methods of data collection also included interviews of service providers,
government officials and key informants, informal discussions with stakeholders and key
informants and Focus Group Discussions with service providers. Government Health
Management Information System (HMIS) was used as a source of secondary data for
assessing performance of Chiranjeevi Yojana. Six of the studies reviewed government policy
and programme documents for the assessment.

In summary, most of these researchers used a mix of quantitative and qualitative

methods to collect data and used both primary and secondary data sources. Also, half of these
studies had a comparison group of non-beneficiaries. For the purpose of this study, the
learning from the review was used to finalise the methods of data collection.
 64

Methods used: Primary data was collected through facility survey, in-depth
interviews of HSPs, women beneficiaries and key informants. Review was conducted of
unpublished secondary sources of secondary information such as relevant documents, reports
and registers. Secondary data was sourced from the state MIS and national, state and district
level surveys.

Data collected from facility survey and interviews of women was quantitative.
Qualitative methods i.e. in-depth interviews were used primarily to garner information about
PPP models and intangible or subjective aspects of service delivery and quality.

Since a pre and post PPP study to explore the effect of PPP on reproductive health
service delivery to women was not feasible in a one-time data collection exercise, a design
akin to case-control was adopted. Reproductive health service delivery through PPPs was
compared to that provided through conventional government facilities i.e. ―Control‖ facilities
operating at the same level of primary health care as the PPP.

a. Primary data collection

1. Facility survey: A facility survey was necessary to understand the available health
infrastructure and its functioning. Two facilities, Dahej PHC and Shamlaji CHC were visited
for the survey.

2. In-depth interviews of health service providers: In-depth interviews were

conducted with health facility managers, medical officers and nursing staff of Dahej PHC and
Shamlaji CHC to seek their perspective on the partnership and its effect on functioning of the
facility.

3. Survey of women: A survey was carried out among women in the reproductive age
group i.e. currently married women 15 to 49 years of age who had used the reproductive
health services in the year prior to the survey. The purpose of the survey was to understand
women‘s preferences, access and use of services, their perceptions about efficiency and
quality and their satisfaction with the reproductive health services at each of these facilities.
To serve this purpose and to make statement about their preferences, perceptions and
experiences at the PPP models with some degree of confidence, government operated
 65

facilities in the same taluka were chosen for comparison. Since the purpose of the study was
very specific exploration of these PPP models, women who did not use these were not part of
the study sample. However, since one time data collection was unlikely to bring out the Pre
and Post PPP change in experiences, comparable non-PPP area were to be used as control for
all three models.

Proposed sample: The area of operation of the select PHC and CHC was considered
for selection of sample of married women in the 15-49 years age group. To begin with,
percentage of institutional deliveries at these PPP models, one of the most appropriate,
relevant and measurable indicator for RH service use and one of the objectives all the three
PPPs are expected to meet was used for sampling. As per the State MIS, Dahej PHC reported
169 and Shamlaji CHC reported 351 deliveries in the year 2010-11. Based on these, and
p=0.5, margin of error 7 per cent and confidence level of 90 per cent the required sample size
for Dahej PHC was estimated at 80 and for Shamlaji CHC at 100 women. Accounting for 10
per cent non-response the sample was finalised as 88 for Dahej and 110 Shamlaji.

For Chiranjeevi Yojana the sample was proposed from Surat district. Surat at the time
of finalising the sample had reported 21 private obstetricians empanelled under the Yojana.
In the district on an average about 2,000 deliveries a year were reported under Chiranjeevi
Yojana. Surat was considered as an average district as far as proportion of deliveries at
Chiranjeevi hospitals were concerned and was therefore selected for study. It was proposed
that a random sample of 50 per cent of the Chiranjeevi hospitals (i.e. 10 Chiranjeevi Yojana
hospitals) would be selected randomly and approximately 50 eligible women who had
delivery at these hospitals (5/hospital) in last one year would be interviewed. A similar
proportion of eligible women who had institutional deliveries at private obstetricians who are
not empanelled under Chiranjeevi Yojana was also proposed.

Revised sample: However during the preparatory field visits and pre-testing visits the
institutional records as well as interviews with health staff indicated that the number of
institutional deliveries at Dahej and Shamlaji were significantly less than what was reported
for these facilities. It was clear that it was difficult to cover the sample as planned. The
sampling procedure was therefore modified. Since there was no data from previous service
use surveys for these three PPP facilities, a conservative estimate of p=0.5 that guarantees the
 66

largest sample size calculation was used. The margin of error was increased and so was the
number of villages across which the sample as to be spread. Based on p= 0.5, margin of error
10 per cent, with a confidence level of 90 per cent and applying FPC i.e. finite population
correction the required sample size for Dahej PHC was now estimated to be 50 and for
Shamlaji CHC it was 60 women. The plan was to interview an equivalent sample from a
control PHC and CHC run by the government in the same districts. To control non sampling
error it was decided to keep sample size minimum and manageable.

Similarly, in Surat, by the time of the field work in started, only six obstetricians
remained empanelled under Chiranjeevi Yojana. Of these five were available and agreeable to
participate in the study. It was decided to interview 10 eligible women who had delivery at
these facilities in last one year and a similar sample from five private obstetricians who were
not empanelled under Chiranjeevi Yojana was proposed.

Thus a total of 320 women beneficiaries were to be interviewed for this study. Since
aim of the study was not to look at prevalence of health needs or vital rates, but to examine
women beneficiaries‘ perceptions about quality of care and satisfaction with services and
there were attendant limitations of budget and time, the sample size was considered to be
practical.

Self-reported data: The data collected from women‘s interviews regarding their
reproductive health needs and health seeking was self-reported. There are inherent
advantages and disadvantages in self-reported data. It provides the respondent‘s own
perspective and information about behaviour which may or may not be observable. However,
there are potential validity problems as the data may not reflect ‗reality‘, respondents may not
reveal socially undesirable behaviour or may not dwell on details or concepts of interest to
the study. Keeping in mind these limitations of self-reported data, the findings were
augmented by data from interviews of key informants and secondary data.

4. In-depth interviews of key informants: In-depth interviews were conducted with

eight key informants representing policy makers, researchers, programme managers and
private service providers to seek their views on and experiences of PPPs, specifically models
of selected PPPs. Barring one, all were interviewed in person. One key respondent was based
in another city and responded to the set of questions by email.
 67

5. Review of literature, documents and reports: Existing relevant literature,

documents, reports and studies were reviewed to understand the concept, rationale and
genesis of these PPPs. Available policy statements, contract documents, government
resolutions, monitoring and performance reports were studied. These were expected to
provide information on the aim of each partner, their motivation to get into these partnerships
and their attributes that work as facilitators or deterrents for women‘s RH and services
women use and satisfaction.

b. Secondary data collection

Relevant secondary data, published as well as unpublished, on conventional

reproductive health outcomes such as ―Full‖ antenatal care, institutional delivery, receipt of
recommended postnatal care, contraceptive prevalence rate, care seeking for reproductive
tract infections and prevalence of malnutrition and anaemia for the selected PPPs was
reviewed. Information on the physical infrastructure, manpower, medicine supply, and
services provided was sourced from the state HMIS and the State and district level surveys.

Process of data collection: As the first step, a formal approval was obtained from
Commissioner of Health, Medical Services & Medical Education, Government of Gujarat to
conduct the study. The office of the Commissioner of Health issued letters to District Health
Officers of Bharuch, Sabarkantha and Surat to co-operate in conduction of the study.

Thereafter with a copy of the letter of approval and after seeking appointment with the
facility manager, the facilities were visited for assessment. Informal discussions were held
with the available staff. Interviews of the select health service providers were conducted after
obtaining their informed consent. Two research assistants, graduates in Social Work and with
5 years of experience assisted with of data collection in the field. Help in data collection was
sought from these research assistants as they were familiar with the local context and
language.

Proposed plan: As per the initial plan, the process to be followed in the field was to
proportionately allot women in the sample to 4 randomly selected villages in the jurisdiction
of Dahej, Shamlaji and their comparative control facilities. Within each of these villages
 68

selection of number of women proportionate to the population of women in the age group 15
to 49 years and eligible for the survey was planned. For this social mapping and list of
women with the health workers were to be used. The plan was to be prepare a rough map of
the area showing different localities. If the allotted number of eligible women was 10 or less,
then one locality was to be selected randomly, household list with the health worker sought
and required number of women selected randomly for visit. If the number of eligible women
was 10 to 14, 15 to 19, 20 to 24 or more than 25, then 2, 3, 4 and 5 localities were to be
randomly selected and equal number of households selected randomly for visit from each
selected locality. If no woman was found in the household, then next household was visited.
If more than one woman was found in the household, younger one was to be interviewed.

Revised plan: Four villages were selected randomly from the villages in the
jurisdiction of the PPP and comparative control facilities. As mentioned earlier, during the
actual field work it was found that number of women seeking reproductive health care in
general and of deliveries at Dahej PHC and Shamlaji CHC was much lower than what was
assumed based on the reported deliveries in a year at each of the facilities in the State MIS. It
was not clear if there was a decline in deliveries at these centres or the reported showed
cumulative deliveries even in the Subcentres under the PHC and CHC. Further, the list of
service users from the ANMs showed that the number of women seeking deliveries or
reproductive health at these facilities was not proportionate to the eligible women population
in the selected four villages. Three of the four facilities had similar number of women seeking
reproductive care in the last one year. As a consequence of these findings, the sampling
procedure was revised to try and get equal number of women from each of the four randomly
selected villages (i.e. about 12 per village in the four selected villages under both Dahej and
Tankari PHC and 15 per village in the four selected village under both Shamlaji and Prantij
CHC). The process of mapping of area, interviewing randomly selected eligible women in the
ASHA/AWW/ANM list was followed in these villages. Wherever the number of women in
the village eligible for the survey was less than the specified number, the neighbouring
villages were visited to interview eligible women and meet the sample. In Dahej for
Ambheta, Jageshwar and Kaladara, in Tankari for Devala and Dabha, in Shamlaji for Napda
Khalsa and Shamalpur and in Prantij for Dalpur villages neighbouring the selected villages
had to be visited to meet the sample. The maps of Districts /Blocks / Talukas with selected
villages highlighted are presented in Appendix 3.1, 3.2, 3.3 and 3.4 (see Appendix 3.1: Map
of Selected Villages in Dahej PHC area, Appendix 3.2: Map of Selected Villages in Tankari
 69

PHC Area; Appendix 3.3: Map of Selected Villages in Shamlaji CHC area; and Appendix
3.4: Map of Selected Villages in Prantij CHC Area).

Though, the plan was to interview women service users from 10 randomly selected
Chiranjeevi ‗empanelled‘ and ‗non-empanelled‘ private hospitals in Surat district, by the time
the study started only six private gynaecologists‘ facilities in rural areas of the district were
still empanelled under the Chiranjeevi Yojana of which only five were willing to participate
in the study process. These practitioners did have fairly successful practice in terms of case
load for deliveries and a sample of 50 women was not difficult to cover. Therefore 10 eligible
women who had delivery at these facilities in last one year were purposively selected and
interviewed after ensuring that the provider had sought their prior informed consent. On the
other hand the Non-Chiranjeevi private gynaecologists in the rural areas surrounding Surat
did not seem to have enough caseload. So, to get requisite number of interviews women
service users, a larger number of Non-Chiranjeevi facilities (14) had to be approached. As in
case of Chiranjeevi empanelled facilities, 10 eligible women who had institutional deliveries
at these facilities in last one year were purposively were interviewed.

None of the interviews were taped in view of reluctance of the respondents to speak
on tape. Data was checked for internal consistency. About two to three women were re-visited
me and informal interviews were conducted again to check if their responses were consistent
with the information collected earlier.

Key informants were selected based on their work, availability and willingness to
respond. Documentary information was collected from Government of Gujarat, private
partners and from the public domain after apprising them of the purpose of the study.

Tools used: Structured tools were used for collecting data from the facilities and the
interviewed women. The tools were developed in English and translated in the local Gujarati
language. They were finalised after pre-testing in the field. Interview Field Guides (IFGs)
were used for in-depth interviews of key informants and health service providers. The IFGs
were in English. The four tools used were as follows:

Tool 1- Tool for assessment of Primary Health Centres / Community Health Centres:
This tool (see Appendix 1) was used to collect data on physical infrastructure, manpower,
 70

furniture and equipment, drugs and supplies, services provided and the reporting process in
the selected facility.

Tool 2 - Women‘s questionnaire: This tool (see Appendix 2) was used to collect socio-
economic information about the household. It also collected data about the demographic and
obstetric profile of the woman and her service use, experiences and satisfaction.

Tool 3 - Health providers‘ interview field guide: This guide (see Appendix 3) included
questions on the actual responsibilities of the staff, service delivery and its monitoring at the
facility, their views on the existing partnership model as well as the utility of such models.

Tool 4 - Key informants‘ interview field guide: This guide (see Appendix 4) included
questions on PPPs, the process of forming partnerships, the partner motivation, selection and
the utility, replicability and sustainability of PPPs.

3.7 Ethical Considerations

Ethics governing research in sensitive subjects like reproductive health demand that
the consent of respondents should be taken and also a guarantee of confidentiality be
expressed by the interviewer at the start of the interview. During data collection the
interviewer introduced herself to each respondent, briefed her about the purpose of the study
and sought 45 minutes of her time to conduct the interview. Respondent was assured that her
participation or refusal to participate in the study will in no way affect the health services she
and her family receive or will receive and that the health services are meant for everyone,
regardless of participation in the study. Respondent was also informed upfront that her
participation was completely voluntary and that she will not be paid for responding to the
queries. She was told that if for any reason some of the topics under discussion made her
uncomfortable, she had the right to refuse to answer these. This would have no consequences
to the health services she received now or in the future. Similar care was exercised during the
in-depth interviews of health service providers and key informants. Throughout the study,
confidentiality of data collected and anonymity of respondents was ensured and data was
collected only with the informed consent of respondents. After receiving the consent the
interviews were conducted in a private setting. The computerised data did not use or record
 71

the respondent‘s name. A study identification number was used in place of respondent‘s
name.

3.8 Data Entry and Analysis

Data from in-depth interviews was transcribed in English and analysed manually.
SPSS was used to develop a data entry structure for the quantitative data from health facility
assessment and women‘s survey. Double entry was done for ten per cent of the data to check
for data entry errors. This data was analysed using the same package.

Variables: The variables used for analysis were as follows:

a. Independent variables
1. Type of partner: There were 3 types of partner corporate, NGO and private
providers.
2. Motivation: The motivation of partners was categorized into three categories:
social responsibility, philanthropy and commercial or profit making.
3. Location: The location of the facility was categorized into accessible and
inaccessible.
4. Timings of the facility: The timings or working hours of the facility were
categorized into convenient and not convenient.
5. Cost of care: Cost was in two categories, affordable and unaffordable.
6. Staff response: Staff‘s response to women‘s health complaints and issues was
categorized into responsive and not responsive.
7. Staff capability: Capability of staff was in two categories capable and not
capable or incapable.
8. Maintenance of privacy: Privacy during women provider interaction was in
two categories maintained and not maintained.
9. Explanations of management: Explanations regarding management of the
health complaints and issues was categorised into provided and not provided.
10. Staff behaviour: Behaviour of the staff was categorized into friendly and not
friendly.
 72

11. Availability of health staff: The availability of health staff was categorized
into available and not available.
12. Availability of medicines: This was in two categories available and not
available.
13. Efficacy of treatment: This was in two categories effective and not effective.

b. Control variables
1. Caste: Women‘s caste was categorized into a binary variable with 2 categories
scheduled caste/ tribe and others.
2. BPL cards: This again was a binary variable with 2 categories possessed BPL
card and did not possess BPL card.
3. Standard of Living Index: Women‘s SLI status was categorised into a binary
variable with two categories ―Low/Medium‖ SLI and ―High‖ SLI .

c. Dependent variables
1. Preference for health facility: Women‘s preference for the facility
(irrespective of reason) for reproductive health care was a binary variable with
2 categories ―Preferred‖ and ―Not preferred‖.
2. Satisfaction with the health services: Women‘s satisfaction with the health
services (irrespective of reasons) for reproductive health care was a binary
variable with 2 categories ―Satisfied‖ and ―Not satisfied‖.

Data was analysed to explore the demographic profile and reproductive health
services used by women. Simple frequencies and bivariate analysis was done to understand
the distribution of variables in and across both the PPPs and control areas. Chi square test of
independence was used to determine significant association between the variables. Logistic
regression was done to establish causal relationship between independent and dependent
variable. In-depth interview data was analysed to explore the emerging themes, patterns and
trends. Multiple logistic regression was attempted to determine the extent to which the
independent variables were predictors of preference for and satisfaction with services at
PPPs.
 73

Indices: For the purpose of in-depth analysis of the causal relationship between
facility attributes and preference and satisfaction the following indices were used:

Facility accessibility: Three indicators location (accessible/inaccessible), working

hours or timings (convenient/not convenient) and perceived cost of care
(affordable/unaffordable) were given equal weightage and based on their cumulative score an
index called ―Accessibility‖ was created. Based on the index score the accessibility of facility
was categorized into 3 categories, ―High accessibility‖ (Score 3), ―Medium accessibility‖
(Score 2) and ―Low accessibility‖ (Score 1 and 0).

Staff competence: Two indicators responsiveness (responsive/not responsive) and

perceived capability of staff (capable/incapable) were given equal weightage and based on
their cumulative score an index called ―Competent‖ staff was created. Based on the index
score the competency of facility staff was categorized into 3 categories, ―High competency‖
(Score 2), ―Medium competency‖ (Score 1) and ―Low competency‖ (Score 0).

Client provider interaction: Three indicators maintenance of privacy (maintained/not

maintained), provision of explanations (provided/not provided) and friendly behaviour of
staff (friendly/not friendly)were given equal weightage and based on their cumulative score
an index called ―Interface‖ staff was created. Based on the index score the client provider
interaction was categorized into 4 categories, ―Good interface‖ (Score 3), ―Poor interface‖
(Score 2) and ―No interface‖ (Score 1 or 0)

Availability and perceived utility of health care: Three indicators availability of staff
(available/not available), availability of medicines (available/not available) and efficacy of
treatment (effective/not effective) were given equal weightage and based on their cumulative
score an index called ―Health care‖ was created. Based on the index score the health care was
categorized into 3 categories, ―High utility‖ (Score 3), ―Medium utility‖ (Score 2), and ―Low
utility‖ (Score 1 or 0).

The findings were triangulated with relevant qualitative data and co-related with
existing literature. In-depth analysis of each PPP was done in the context of the research
questions. The cases were studied independently. Cross case analysis was not the intention of
this study as the three PPPs are different in multiple ways.
 74

Appendix 3.1: Map of Selected Villages in Dahej PHC area

 75

Appendix 3.2: Map of Selected Villages in Tankari PHC Area

 76

Appendix 3.3: Map of Selected Villages in Shamlaji CHC area

 77

Appendix 3.4: Map of Selected Villages in Prantij CHC Area