Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
i
INNOVATIVE HEALTHCARE FINANCING AND EQUITY THROUGH
COMMUNITY BASED HEALTH INSURANCE SCHEMES (CBHIS) IN KENYA
BY
JANE WANGUI GITAHI
UNITED STATES INTERNATIONAL UNIVERSITY - AFRICA
SPRING 2017
ii
INNOVATIVE HEALTHCARE FINANCING AND EQUITY THROUGH
COMMUNITY BASED HEALTH INSURANCE SCHEMES (CBHIS) IN KENYA
BY
JANE WANGUI GITAHI
A Dissertation Report submitted to the Chandaria School of Business in Partial
Fulfillment of the Requirement for the Degree of Doctorate in Business Administration
(DBA)
UNITED STATES INTERNATIONAL UNIVERSITY - AFRICA
SPRING 2017
iii
STUDENT’S DECLARATION
I, the undersigned, declare that this is my original work and has not been submitted to any
other college, institution or university other than the United States International University
for academic credit.
Signed: ________________________ Date: _____________________
Jane Wangui Gitahi (ID 627753)
This project has been presented for examination with my approval as the appointed
supervisors
Signed: ________________________ Date: _____________________
Prof. Amos Njuguna
Signed: ________________________ Date: _____________________
Dr. Timothy C. Okech
Signed: ________________________ Date: _____________________
Dean Chandaria School of Business
Signed: ________________________ Date: _____________________
Deputy Vice, Chancellor Academic and Students Affairs
Signed: ________________________ Date: _____________________
iv
COPYRIGHT
All rights reserved. No part of this report may be photocopied, recorded or otherwise
reproduced, stored in retrieval system or transmitted in any electronic or mechanical means
without prior permission of USIU-A or the author.
Jane W. Gitahi © 2017
v
ABSTRACT
The purpose of this study was to examine innovative healthcare financing and equity through
CBHIs in Kenya. Particularly, the study explored the effect of the enrolment, mix of
contributions; risk pooling, strategic purchasing and the moderating effect of government
stewardship on equity in healthcare in Kenyan CBHIs. Out of the 115 CBHIs that are
registered with Kenya Community Based Health Financing Association, data was collected
from a sample of 79 CBHIs which had complete and consistent data. Responses were sought
from four members of each CBHIs management team. Data was collected by use of
questionnaires and data collection sheets for secondary data. 224 usable questionnaires were
collected, representing 71% percent response rate. Data was cleaned and analyzed using
Excel, SPSS version 20 and SmartPLS version 3. The study used descriptive statistics, factor
analysis, path analysis and multivariate regression analysis in structural modeling equation
(SEM) to investigate the relationship among variables and measure the strength and direction
of relationships between constructs.
Empirical results revealed that there exist a positive relationship between enrolment in
CBHIs and equity in healthcare in Kenya, a negative relationship between mix of
contributions in CBHIs and equity in healthcare in Kenya, positive relationship between risk
pooling in CBHIs and equity in healthcare in Kenya and positive relationship between
strategic purchasing in CBHIs and equity in healthcare in Kenya. Government stewardship
has a positive influence on the relationship between enrolment and strategic and equity in
healthcare.
The major conclusion drawn from the study is that CBHIs stimulates enrolment by focusing
on social capital and stimulating willingness to pay in excluded segments of the population.
The current mix of contributions in CBHIs does not offer an optimal mix of funds necessary
for increased access to care and financial risk protection for precluded groups. The study
recommends that for realization of equity goals in the healthcare in Kenya government and
sectoral partners should define the place of CBHIs with the national health policy by enacting
the requisite legal and regulatory framework to guide CBHIs administrative and fiscal
vi
structures. The government should also encourage the players in health financing to
consolidate the pooled funds to enhance risk equalization.
vii
ACKNOWLEDEGMENTS
The journey of completing this doctoral research would not have been possible without the
direction, encouragement and support of particular people. I am highly indebted to my two
supervisors Prof. Amos Njuguna and Dr. Timothy Okech for their great insight, valuable
guidance, encouragement and mentorship throughout the process of writing this thesis.
The realization of my dream of pursuing doctoral studies would not have been a reality were
it not for United States International University-Africa admissions office and the vetting
board under the leadership of Dr. George Achoki and Prof. Amos Njuguna. I cannot forget
my colleagues in DBA one for continuous encouragement and Ephantus Mutitu for his
invaluable support during data analysis.
Special thanks goes to my husband Isaac and my sister Silvia for their prayers and unfailing
encouragement – you believed in me and endured long hours of absence in my pursuit of this
degree. Special thanks to my brother Jesse for encouragement and support. To my friends,
thank you for your prayers and support.
Above all, am thankful to God.
viii
DEDICATIONS
Dedicated to Isaac and Sylvia
ix
TABLE OF CONTENTS
STUDENT’S DECLARATION ............................................................................................ iii
COPYRIGHT ......................................................................................................................... iv
ABSTRACT ............................................................................................................................. v
ACKNOWLEDEGMENTS ................................................................................................. vii
DEDICATIONS ................................................................................................................... viii
TABLE OF CONTENTS ...................................................................................................... ix
LIST OF ABBREVIATIONS .............................................................................................. xv
CHAPTER ONE ..................................................................................................................... 1
1.0 INTRODUCTION............................................................................................................. 1
1.1 Background of the Study .................................................................................................... 1
1.2 Statement of the Problem .................................................................................................. 10
1.3 General Objective of the study ......................................................................................... 11
1.4 Specific Objectives ........................................................................................................... 11
1.5 Hypotheses: Null and alternative hypotheses ................................................................... 11
1.6 Justification of the Study .................................................................................................. 12
1.7 Scope of the Study ............................................................................................................ 13
1.8 Definition of Terms........................................................................................................... 13
CHAPTER TWO .................................................................................................................. 16
2.0 LITERATURE REVIEW .............................................................................................. 16
2.1 Introduction ....................................................................................................................... 16
2.2 Theoretical Review ........................................................................................................... 16
2.3 Conceptual Framework ..................................................................................................... 22
2.4 Empirical Review.............................................................................................................. 67
2.5 Chapter Summary ........................................................................................................... 109
CHAPTER THREE ............................................................................................................ 110
3.0 RESEARCH METHODOLOGY ................................................................................ 110
3.1 Introduction ..................................................................................................................... 110
3.2 Research Philosophy and Research Paradigm ................................................................ 110
3.3 Research Design.............................................................................................................. 112
3.4 Population ....................................................................................................................... 113
x
3.5 Sampling Design ............................................................................................................. 113
3.6 Data Collection Methods ................................................................................................ 114
3.7 Research Procedures ....................................................................................................... 117
3.8 Data Analysis Methods ................................................................................................... 123
3.9 Chapter Summary ........................................................................................................... 139
CHAPTER FOUR ............................................................................................................... 140
4.0 FINDINGS ..................................................................................................................... 140
4.1 Introduction ..................................................................................................................... 140
4.2 General Information ........................................................................................................ 140
4.3 Effect of Enrolment on Equity in Healthcare ................................................................. 155
4.4 Effect of Mix of Contributions on Equity in Healthcare ................................................ 166
4.5 Effect of Risk Pooling in CBHIs on Equity in Healthcare Indicators ............................ 171
4.6 Effect of Strategic Purchasing in CBHIs on Equity in Healthcare Indicators ................ 177
4.7 Moderating effect of Government Stewardship on Equity in Healthcare ...................... 183
4.8 Overall Model ................................................................................................................. 190
4.9 Chapter Summary ........................................................................................................... 199
CHAPTER FIVE ................................................................................................................ 201
5.0 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS .............................. 201
5.1 Introduction ..................................................................................................................... 201
5.3 Discussion of Results ...................................................................................................... 203
5.4 Conclusions ..................................................................................................................... 213
5.5 Recommendations ........................................................................................................... 216
REFERENCES .................................................................................................................... 219
APPENDICES ..................................................................................................................... 246
APPENDIX 1: APPROVAL LETTER FROM USIU-A ...................................................... 246
APPENDIX 2: APPROVAL LETTER FROM NACOSTI .................................................. 247
APPENDIX 3: NACOSTI RESEARCH CLEARANCE PERMIT ...................................... 248
APPENDIX 4: QUESTIONNAIRE ..................................................................................... 249
APPENDIX 5: SECONDARY DATA SHEET .................................................................... 258
APPENDIX 6: LIST OF CBHIS AND THEIR RESPECTIVE NETWORKS .................... 262
APPENDIX 7: CROSS LOADINGS OF CONSTRUCTS .................................................. 263
xi
LIST OF TABLES
Table 3.1 Type of data and data collection tools .................................................................. 115
Table 3.2: Reliability Test of Constructs .............................................................................. 119
Table 3.3: Kaiser-Meyer-Olkin and Bartlett's test ................................................................ 120
Table 3.4: Measures to Fit PLS Model ................................................................................. 131
Table 3.5 Items of measure for study constructs .................................................................. 133
Table 3.6: Summary of Hypotheses Testing ......................................................................... 138
Table 4.1 Descriptive Analysis - Extent Effect of Equity in Healthcare- Healthcare Access
............................................................................................................................................... 148
Table 4.2 Descriptive Analysis - Extent Effect of Equity in Healthcare - Equity in
Contributions......................................................................................................................... 149
Table 4.3 Descriptive Analysis – Extent Effect of Equity in Healthcare - Quality of Care . 149
Table 4.4 Descriptive Analysis - Extent Effect of Equity in Healthcare - Sustainability
(Administrative and Managerial Capability) ........................................................................ 151
Table 4.5 Descriptive Analysis - Extent Effect of Equity in Healthcare – Sustainability
(Financial Sustainability) ...................................................................................................... 153
Table 4.6 Cronbach‘s Alpha Coefficients, AVE and KMO values for Equity in Healthcare
(Healthcare Access, Equity in Contributions, Quality of Care and Sustainability) .............. 155
Table 4.7 Extent Effect of Affordability on Equity in Healthcare ........................................ 156
Table 4.8 Extent Effect of Unit of Membership on Equity in Healthcare ............................ 157
Table 4.9 Extent Effect of Timing of Collections on Equity in Healthcare ......................... 158
Table 4.10 Extent Effect of Trust on Equity in Healthcare .................................................. 159
Table 4.11 Correlation between Affordability and Healthcare Access ................................ 160
Table 4.12 Correlation between Affordability and Equity in Contributions ........................ 160
Table 4.13 Correlation between Affordability and Quality of Care ..................................... 160
Table 4.14 Correlation between Affordability and Sustainability ........................................ 161
Table 4.15 Correlation between Timing of Collections and Healthcare Access .................. 161
Table 4.16 Correlation between Timing of Collections and Equity in Contributions .......... 161
Table 4.17 Correlation between Timing of Collections and Quality of Care ....................... 162
Table 4.18 Correlation between Timings of Collections and Sustainability ........................ 162
Table 4.19 Correlation between Trust and Healthcare Access ............................................. 162
xii
Table 4.20 Correlation between Trust and Equity in Contributions ..................................... 163
Table 4.21 Correlation between Trust and Quality of Care .................................................. 163
Table 4.22 Correlation between Trust and Sustainability ..................................................... 163
Table 4.23 Cronbach‘s Alpha Coefficients, AVE and KMO values for Enrolment ............ 164
Table 4.24 Path Coefficients (Mean, STDEV, t-value) ........................................................ 166
Table 4.25 Descriptive Analysis - Extent Effect of Mix of Contributions in CBHIs on Equity
in Healthcare ......................................................................................................................... 167
Table 4.26 Correlation between Mix of Contributions and Healthcare Access ................... 168
Table 4.27 Correlation between Mix of Contributions and Equity in Contributions ........... 168
Table 4.28 Correlation between Mix of Contributions and Quality of Care ........................ 169
Table 4.29 Correlation between Mix of Contributions and Sustainability ........................... 169
Table 4.30 Cronbach‘s Alpha Coefficients, AVE and KMO values for Mix of Contributions
............................................................................................................................................... 169
Table 4.31 Path Coefficients (Mean, STDEV, t-values) ...................................................... 171
Table 4.32 Extent Effect of Risk Pooling in CBHIs on Equity in Healthcare ...................... 172
Table 4.33 Correlation between Risk pooling in CBHIs and Healthcare Access ................. 174
Table 4.34 Correlation between Risk Pooling and Equity in Contributions......................... 174
Table 4.35 Correlation between Risk Pooling and Quality of Care ..................................... 175
Table 4.36 Correlation between Risk Pooling and Sustainability ........................................ 175
Table 4.37 Cronbach‘s Alpha Coefficients, AVE and KMO values for Risk Pooling ......... 175
Table 4.38 Path Coefficients (Mean, STDEV, t-values) ...................................................... 177
Table 4.39 Descriptive Analysis - Extent Effect of Strategic Purchasing on Equity in
Healthcare ............................................................................................................................. 178
Table 4.40 Correlation between Strategic Purchasing and Healthcare Access .................... 179
Table 4.41 Correlation between Strategic Purchasing and Equity in Contributions ............ 179
Table 4.43 Correlation between Strategic Purchasing and Sustainability ............................ 180
Table 4.44 Table 4.44 Cronbach‘s Alpha Coefficients, AVE and KMO values for Strategic
Purchasing ............................................................................................................................. 182
Table 4.45 Path Coefficients (Mean, STDEV, t-values) ...................................................... 182
Table 4.46 Descriptive Analysis - Extent Effect of Government Stewardship on Equity in
Healthcare - Advisory Role on Design ................................................................................. 184
xiii
Table 4.47 Descriptive Analysis - Extent Effect of Government Stewardship on Equity in
Healthcare- Monitoring ......................................................................................................... 184
Table 4.48 Descriptive Analysis - Extent Effect of Government Stewardship on Equity in
Healthcare- Training ............................................................................................................. 184
Table 4.49 Descriptive Analysis - Extent Effect of Government Stewardship on Equity in
Healthcare- Co-financing ...................................................................................................... 186
Table 4.50 Cronbach‘s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Design ............................................................................................................ 186
Table 4.51 Cronbach‘s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Monitoring ..................................................................................................... 188
Table 4.52 Cronbach‘s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Training .......................................................................................................... 188
Table 4.53 Cronbach‘s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Co-financing .................................................................................................. 189
Table 4.54 Multicollinearity Test ......................................................................................... 190
Table 4.55 Construct reliability ............................................................................................ 190
Table 4.55 Convergent Validity of outer model ................................................................... 190
Table 4.57 Measures of Discriminant Validity ..................................................................... 192
Table 4.58 Latent Variable Correlations / Correlation matrix of constructs ........................ 192
Table 4.59 Path Coefficients (Mean, STDEV, t-Values)...................................................... 196
Table 4.60 Path Coefficients (Mean, STDEV, t-values) ...................................................... 198
xiv
LIST OF FIGURES
Figure 2.1 Conceptual Framework ......................................................................................... 23
Figure 3.1 Path model for Equity in Healthcare ................................................................... 132
Figure 4.1 Figure 4.1 Response Rate .................................................................................... 132
Figure 4.2 Responses based on Years of Operation ............................................................. 141
Figure 4.3 Households targeted by CBHIs ........................................................................... 142
Figure 4.5 Relationships between the Number of Households Targeted and Those Covered
by CBHIs .............................................................................................................................. 143
Figure 4.6 Benefits Package, Premiums and Products Uptake in CBHIs ............................. 144
Figure 4.7 Methods of Payment: Inpatient and Outpatient services ..................................... 145
Figure 4.8 Distance to the Nearest Contracted Service Provider.......................................... 146
Figure 4.9 Average mix of contributions in CBHIs .............................................................. 146
Figure 4.10 Trends of Total Premiums collected, healthcare cost reimbursements,
administration cost and deficit/surplus in CBHIs between 2010-2015 ................................ 147
Figure 4.11 Path coefficients for effect of enrolment in CBHIs on equity in health care .... 165
Figure 4.12 t-values for effect of enrolment in CBHIs on equity in healthcare ................... 165
Figure 4.13 Path coefficients for effect of mix of contributions on equity in healthcare in
CBHIs ................................................................................................................................. 1713
Figure 4.14 t-values for effect of mix of contributions on equity in healthcare in CBHIs ... 171
Figure 4.15 Path coefficients for effect of Risk pooling in CBHIs on equity in healthcare . 176
Figure 4.16 t-values for effect of Risk pooling in CBHIs on equity in healthcare ............... 176
Figure 4.17 Path coefficients for effect of Strategic purchasing in CBHIs on equity in
healthcare ............................................................................................................................ 1824
Figure 4.18 t-values for effect of Strategic purchasing in CBHIs on equity in health care .. 182
Figure 4.19 Path coefficients for the optimum model without moderation .......................... 194
Figure 4.20 t- values for the optimum model without moderation ....................................... 195
Figure 4.21 Path coefficients for the optimum moderated model ........................................ 198
Figure 4.22 t-values for the optimum moderated model....................................................... 197
xv
LIST OF ABBREVIATIONS
CBHI Community Based Health Insurance
CHAT Choosing Healthplan All Together
DFID Department for International Development
GIZ Deutsche Gesellschaft für Internationale Zusammenarbeit
GTZ Deutsche Gesellschaft für Technische Zusammenarbeit
GDP Gross Domestic Product
GNI Gross National Income
HLTIIFHS High- level Taskforce on Innovative International Financing for
Health Systems
IMF International Monetary Fund
KIHBS Kenya Integrated Household Budget Survey
KHHUES Kenya Household Health Expenditure and Utilization Survey
KNBS Kenya National Bureau of Statistics
LMICs Low and Middle Income Countries
MDG Millennium Development Goal
MLSSS Ministry of Labour Social Security and Services
MoH Ministry of Health
MOMS Ministry for Medical Services
NHA National Health Accounts
NHIF National Hospital Insurance Fund
NPISH Nonprofit Institutions Serving Households
NSHIF National Social Health Insurance Fund
ODA Official Development Assistance
OECD Organization for Economic Co-operation and development
OECD-DAC Organization for Economic Co-operation and development -
Development Co-operation Directorate
OOP Out-of-Pocket
SDGs Sustainable Development Goals
SEAR South East Asia Regions
SWAp Sector-Wide Approach
xvi
THE Total health expenditure
UHC Universal Health Coverage
UN United Nations
VAT Value Added Taxes
WHO World Health Organization
WHA World Health Assembly
1
CHAPTER ONE
1.0 INTRODUCTION
1.1 Background of the Study
The subject of disparities and inequities in healthcare is increasingly being recognized as a
central issue in current policy debates on healthcare. Huge differences in healthcare access
and high levels of financial risks associated with healthcare payments have been
documented; most of the cases are widespread in low and middle income countries (LMICs).
Globally an estimated 400 million people lack access to essential health services, 17% of the
people are impoverished or pushed deeper into poverty by healthcare costs (Starfield, 2011;
WHO, 2015a; Asante, Price, Hayen, Jan & Wiseman, 2016) whilst almost a third of
households in Africa and South East Asia regions (SEAR) of WHO are forced to borrow
money or sell assets to pay for healthcare at the point of use (Kruk, Godmann & Galea,
2009).
Enormous discrepancies in healthcare expenditure are evident among countries with LIMCs
relying heavily on out of pocket (OOP) expenditure to finance healthcare. In 2013,
households in LIMCs contributed 42.3% and 40.6% respectively of Total Health Expenditure
(THE) compared to 21.2% in high income countries (WHO, 2016). Additionally, while
poorer countries in African and SEAR of WHO account for over half of global burden of
disease and 39% of world‘s population they spent only 3% of world health resources in 2012
(WHO, 2015b). Similarly, the WHO African region and SEAR are deprived of access to
quality healthcare due to large deficits of skilled health workers (4.2 million and 6.9 million
respectively). Within countries, the disparities in access are driven by differences in
socioeconomic status with Sub-Saharan Africa and SEAR having the highest child mortality
rates (WHO, 2016). The ensuing disparities in access to quality healthcare and financial
protection places equity at the heart of current policy debates of Universal Health Coverage
(UHC) and in the post 2015 Sustainable Development Goals (SDGs) agenda (WHO, 2015c).
UHC was founded on the principle of access to healthcare for all with financial risk
protection while SDGs are founded on the theme of inclusiveness.
2
The call for health for all dates back to 1948 when the World Health Organization (WHO)
constitution declared health a fundamental human right and the later reaffirmation in the
Alma Ata Declaration of 1978 (WHO, 1948; 1978). Since the fifty- eighth World Health
Assembly (WHA) endorsed the resolution of 2005 that called ‗for sustainable health
financing, universal health coverage and social health insurance‘ (WHO, 2005), the visibility
and importance of UHC has been increasing steadily. The entire World Health Report of
2010 dwells on UHC and acts as a guide on innovative and sustainable health financing
methods for countries at all levels of development (WHO, 2010a). More recently the United
Nation (UN) resolution of 2012 urged governments to ‗accelerate the transition towards
universal access to affordable and quality health care services‘. This not only proves the wide
consensus on the urgency to act but also the increasing concern about the large differences in
healthcare access and financial risk protection particularly in LIMCs (WHO, 2015c).
UHC is an aspiration that all people have access to quality and effective promotive,
preventive, rehabilitative, palliative and curative health services at the time of need without
suffering financial ruin (WHO, 2010a; 2015a). Realization of UHC and SDGs will require
commitment of more resources to healthcare (WHO, 2010b; Tangcharoensathien, Mills &
Palu, 2015). The desire to enhance financial risk protection and improve access to quality
healthcare services therefore lies at the core health financing. An important challenge
therefore is developing health financing mechanisms that guarantee access to quality
healthcare and offer financial protection for all. Health financing encompasses three
functions; revenue collection (includes enrolment rates and mix of contributions), risk
pooling and strategic purchasing functions. Revenue collection involves raising of funds; risk
pooling encompasses accumulation and use of the funds in equalization of financial risks
associated with ill health which strategic purchasing involves sourcing for cost effective and
quality health services from healthcare providers and paying for them (WHO, 2010a).
Government stewardship is critical for steering the implementation of these functions since
the government bears the ultimate responsibility for the health of its people (WHO, 2000).
Availability of sufficient financial resources for health remains a fundamental question for all
countries. Many high income countries experience limited fiscal space as their health
3
demands evolve driven by a large ageing population and declining payroll taxes as a result of
shrinking workforce (WHO, 2010a). On the other hand, many of the poorest countries
struggle to ensure availability of basic health services (WHO, 2010a; 2014a). In 2012, low
income countries spent an average of US$ 32 per capita which is only a small fraction of
US$ 4625 spent by Organization for Economic Co-operation and Development (OECD)
countries (WHO, 2015b). This is way below the minimum recommended per capita
expenditure of US$44 in 2009 to more than US$60 in 2015 for provision of essential health
services and attainment of health related Millennium Development Goals (MDGs) (WHO
High-level Taskforce on Innovative International Financing for Health Systems (HLTIIFHS),
2009). This highlights the absolute need for more resources in low income countries for
them to achieve equity as envisioned by UHC and SDGs.
The financing gap is exacerbated by the rising costs of health systems. In 2013 global health
expenditure is estimated to have increased by 2.8% from US$ 7.2 trillion in 2012. Further,
the spending is projected to increase sharply by 5.2% annually between 2014-2018 to US$
9.3 trillion (WHO, 2013; 2012; 2010). This escalation is driven by aging and growing
populations, rising prevalence of chronic diseases and expensive medicines, procedures and
technologies (WHO, 2015d). High income countries continually seek more financial
resources to pay for rapidly expanding technologies and options for improving health while
low and middle income countries bear a staggering double burden of stubbornly high
communicable diseases and an escalating prevalence of non-communicable diseases (WHO,
2010a). This accentuates the need to prioritize health in government budgets while
diversifying sources of domestic funding through innovative financing mechanisms.
A significant number of countries at various levels of development have embraced UHC.
Most of these countries are in the WHO regions of America, Europe and Western Pacific. In
these countries healthcare is largely financed through general taxation and health insurance
(WHO, 2010a). As result their total health expenditure (THE) as percentage of Gross
Domestic Product (GDP) is well above 5–6%, a level recommended for realization of UHC
(WHO, 2010a; 2015e). In contrast, OOP still accounts for almost half of healthcare funds
(WHO 2016). In 2012, OOP as a percent of THE was more than 20% in 37 out of 45
4
countries in African region (WHO, 2014a). In addition, key health financing indicators are
below the target recommended. In 17 out of 45 countries the THE per capita is less than
US$44; health expenditure as a % of government expenditure is less than 15% and THE as a
percent of GDP is less than 5%. General taxation (government financed system) and health
insurance (market based) approaches have been recognized as progressive methods of
achieving UHC (WHO, 2010a; 2015a).
Government spending is seen as a stable and sustainable source of healthcare funds, making
it an important source of financing healthcare for UHC. Government spending is critical for
ensuring access to essential health services and cautions the poor and vulnerable against
financial risk associated with healthcare costs (Durairaj & Evans, 2010). This fact was
recognized by African heads of states when they committed to spend at least 15% of
government budgets on health in 2001 (Organization of African Unity (OAU, 2001). Thirty
nine out of 45 countries, Kenya being one of them have not achieved the target (WHO,
2014a). The percentage of government spending on health in Kenya have fluctuated from a
base of 7% in 2001/02, rising to 8.6 % in 2002/03, then falling to 4.6% in 2009/10 and then
rising to 6.1% in 2012/13 (MoH, 2015). There is a general consensus that the levels of
government expenditure are determined by the revenue generated and the macro-economic
policy (McIntyre & Meheus, 2014; Doherty, Kirigia, Ichoku, McIntyre, Hanson & Chuma,
2014). Like in many African countries, increased government revenue and economic growth
in Kenya have not translated into expanded fiscal space for health. This highlights the need
for tradeoffs between improving health particularly for the poor and vulnerable and growing
the economy. In essence, UHC and long-term economic development are inextricably linked
(WHO, 2001).
The World Health Report of 2010 emphasizes on the absolute need for donor support in
LIMCs in the short term for realization of UHC (WHO, 2010a). Efforts to rally global
solidarity have however remained unsuccessful. Donor funding is still unpredictable and fails
to meet the set targets in most cases. For instance, only 5 out of 22 donors met the set
requirements in 2009 while the targets set for 2015 were not achieved (WHO, 2013; 2016).
Progress in coordination and harmonization of donor aid through sector wide approach
5
(SWAp) is still slow in many recipients‘ countries (WHO, 2013). This approach is aimed at
giving recipient countries control in allocating the aid to priority interventions that are critical
for reducing the systemic access discrepancies and financial risks.
At independence Kenya‘s health system was predominantly tax- funded. As part of broader
structural reforms the government liberalized its economy in the late 1980s. As a result, a
user fee was introduced in the health facilities to cover part of services costs in government
facilities. As a consequence there was challenge of affordability and decline in utilization of
healthcare services especially among the poorest population (Chuma and Okungu, 2011).
The user fee was modified following the introduction of free health services in dispensaries
and health centers in 2004 (Carrin et al, 2007) and free maternity services in all public
hospitals in 2013. In 2010, the Health Sector Services Fund (HSSF) was established. The
fund represents a systematic way of channeling pooled government and donor funds to level
2 and 3 facilities with an aim of cautioning them against decline in revenue associated with
abolishment of user fees (Chuma and Okungu, 2011). Today Kenya operates a pluralistic
health financing system with major contributors being government, households and donors.
Recent analysis of health expenditure and financing flows in the health sector in 2012/13
National Health Accounts (NHA) shows that the private sector is major financier
contributing 40% of THE, while government and donors contributed 34% and 26% of THE
respectively. Donor funding declined from 35% in 2009/2010 to 26%. The percentage of
THE mobilized through OOP (excluding cost sharing) was 27% up from 19% in 2009/10
whereas Non-Profit Institutions Serving Households (NPISH) financing schemes declined by
45% compared to 2009/10 estimates (MoH, 2015). Estimates based on Ministry of Health
(2014) indicate that OOP spending on outpatient and inpatient accounted for 78% (48.4
billion) and 22% (13.7 billion) of total household health expenditure respectively (MoH,
2014). Various studies estimate that OOP push about 1.48 million Kenyans below the
poverty line (Xu, Evans, Kawabata, Zeremdini, Klavus & Murray, 2003; Chuma and Maina,
2012). Although this is a strong predictor of catastrophic health expenditure especially
among the poor (WHO, 2010a), it implies that equity in healthcare for poor and vulnerable
6
groups can be realized by channeling a fraction of OOP payments to CBHIs that are
incentivized by government and donor funds (WHO, 2001; 2010).
Risk pooling mechanisms are poor with only 4% of all health funds are pooled through
insurance (Chuma & Okungu, 2011). The estimated health insurance coverage in 2014 was
about 17.1% (44% of the population). Out of this coverage, the National Insurance Hospital
Fund (NHIF) covered 88.4%, while private insurance, Community Based Health Insurance
Schemes (CBHI) and others forms of insurance covered 9.4%, 1.3 and 1.0% respectively. On
the basis on wealth status, only 2.9% of the poorest quintile is covered (MoH, 2014). Out of
the 18 million poor, 10 million people are extremely poor can only be covered through
dedicated taxes and donor support while 8 million can access care through a government
subsidized cover (Ministry for Medical Services (MOMS), 2011). This evidences a potential
of increasing access to healthcare with adequate financial protection for all through
innovative health financing mechanisms that guarantees rapid coverage of the informal sector
and inclusion the poor.
Despite the general agreement that government spending is critical for realization of equity
given its stability and sustainability as observed by Durairaj & Evans (2010), the country
faces immense hurdles in finding adequate fiscal space for health. Like many African
countries, Kenya is composed of a large informal sector and a comparatively small and
stagnant formal sector (Kenya National Bureau of Statistics (KNBS), 2016). This presents
practical difficulties in collecting tax and health insurance contributions particularly from the
informal sector due to lack of institutional capacity to collect taxes (WHO, 2010a). For
instance, the small and stagnant formal sector account for 61.1% of NHIF membership
(KNBS, 2016). This implies that the contributions from the salaried sector are not adequate
to cross-subsidize the poor and vulnerable group (Carrin et al., 2007). According to Kenya
Integrated Household Budget Survey (KIHBS) 2005/06) almost half of the Kenyan
population lives below the poverty line (KNBS, 2007).
Based on the 2009 population estimates, this translates to 18 million, 10 million of which are
estimated to be extremely poor (MOMS, 2011). Inclusion of this category is urgent and
7
vexing task that is dependent on expansion of fiscal space for health coupled with
diversifying domestic sources through innovative health financing mechanisms. Government
contribution through trusted institutions is imperative for subsidization of this category
(WHO, 2010a). The phased expansion of NHIF has been disputed due to concerns of poor
governance and lack of capacity among other reasons (Munguti, 2010). Furthermore, current
NHIF contribution rates for the informal sector are flat. The flat rates are regressive mainly
for the poor and vulnerable groups as they do not match payments with ability to pay. In the
meantime, the poor and vulnerable are expected to receive financial protection through the
existing waiver system. Research findings by Collins, Quick, Musau, Kraushaar & Hussein
(1996) & Bitrán & Giedion (2003) show that the system is fraught with weakness. Deciding
on who is eligible is difficult; lack of awareness on the waiving mechanisms and the waiving
process is complicated and time consuming are some of its shortcomings. This system has
therefore failed to offer financial protection to the poor and vulnerable.
Given the limited fiscal space for health and the urgency of achieving equity goals, it would
therefore be imperative to explore other financing alternatives that can circumvent the current
political and organizational challenges being experienced at the national level. In this context
community involvement in healthcare financing presents an option for improved access to
healthcare by the poor and financial protection against catastrophic health expenditures
(Jakab & Chrishnan, 2001; Carrin, 2003). Community financing have emerged in the
backdrop of economic constraints, lack of good governance and lack of government
stewardship in the informal health sector (Preker, 2002). Community financing schemes
continues to attract to widespread attention due to its potential to achieve a large degree of
penetration through community involvements compared government and market based health
insurers (Jakab & Chrishnan, 2001; Preker, 2002).
Various forms of community financing exist. First, community managed user fees mainly
involves payment of user fee at the point of use. An example is the Bamako Initiative, that is
involves the community in setting the user fee levels, apportioning funds, developing and
management of waiver criteria and carrying out general administration and oversight (Gilson,
Kalyalya, Kuchler, Lake, Oranga & Ouendo, 2001; Preker, 2002; Carrin, Waelkens & Criel,
8
2005). Second, is the community prepayment or mutual organizations which are typified by
voluntary membership, single annual prepayment and community involvement in designing
and management of the scheme. Third, the government or social insurance supported
community driven schemes enlist the community in reaching the rural and excluded groups
on behalf of formal government or a social health insurer (Preker, 2002). Fourth, the
provider based health insurance which is localized around a provider unit usually a town, city
or regional hospital. The schemes membership is voluntary and is designed by local
healthcare providers to promote utilization of health services. The schemes are managed by
the healthcare providers who collect premiums. They often cater for expensive inpatient
services (Melbratie, Sparrow, Alemu & Bedi, 2013).
This study focuses on community prepayments and mutual organizations; a form of
Community Based Health Insurance (CBHIs) that emerged in late 1980s. In Anglophone
literature, the schemes are mostly referred to as Community Health Insurance or Community
Based Health Insurance while in Francophone countries they are labeled as Mutuelle de
Santé (Soors et al., 2010) depicting the underlying social drive and solidarity as one their
major principle. The schemes are characterized by voluntary membership, nonprofit health
financing mechanism whereby the households in the community finance or co-finance a set
of health interventions as well other needs outside healthcare (McPake, 1993; Carrin et al.,
2005, Soors, Devadasan, Durairaj & Criel, 2010).
The governance of the scheme is entrusted on a committee which is selected by the
community members. The committee is responsible for revenue collection, management of
the fund and enlisting of new members. In addition to member contributions, some receive
donor funding. To cope with risks and vulnerability, the schemes exploit social capital
namely solidarity and reciprocity as a primary resource making it easier identify the
contributing population and collect the contributions (Carrin et al., 2005). These values are
inherent in the self-help strategies that have been employed by the poor for a long time
(DeRoeck, 1996; Musau 1999).
9
All over the world, community financing have been used as a strategy for mobilizing
resources to finance and deliver healthcare for the informal sector and the poor. Some
schemes have been successful in pooling resources and transforming money into effective
services efficiently (Preker, 2002). Germany, Japan, China, Korea and Taiwan for example
transformed their health insurance by enlisting the informal sector through small groups
which eventually merged to larger schemes (Criel & Van Dormael, 1999; WHO, 2010).
Thailand and Indonesia presents examples of countries that have made remarkable strides in
increasing service coverage and financial protection through gradual expansion and
integration of voluntary CBHIs into the broader healthcare system (Poletti, Balabanova,
Ghazaryan, Kamal-Yanni, Kocharyan, Arakelyan & Hakobyan, 2007). Ghana and Rwanda
have in the recent past successfully increased health insurance coverage to 60% and 91%
respectively by reorienting their health financing towards pro-poor prepaid schemes
(Fernandes et al., 2009 ; Durairaj, D‘Almeida & Kirigia, 2010; MOMS, 2011; Schieber,
Cashin, Saleh & Lavado, 2012). In Kenya CBHI schemes have evolved over time to take
care of healthcare financing requirements of the low income households who have been left
out of mainstream prepaid schemes (MoH, 2014).
In Kenya, the Schemes are registered by Ministry of Labour Social Security and Services
(MLSSS). They have an umbrella association called Kenya Community Based Health
Financing Association (KCBHFA) that assist the organizations and other key stakeholders in
promoting community based health financing initiatives. The number of CBHIs that were
registered with KCBHFA in 2015 was 115 (KCBHFA, 2015) with a membership of 94,000
(Munge, Mulupi & Chuma, 2016). They rely on social capital particularly social solidarity,
trust and social accountability to enhance community and ownership. They offer platform
for community participation through annual general meetings and other periodic meetings
where members voice their complaints, concerns and present their views on varied issues.
During these meeting, the management team gives feedback in form of reports on financial
performance (Munge et al., 2016). Despite the recognition of Community based health
financing by MoH National Health Policy (NHP) 2012-2030, the initiatives have not been
applied as a vehicle of identifying and subsidizing the poorest as well as strengthening
inclusion of the poor and the informal sector (MoH, 2012).
10
1.2 Statement of the Problem
In Kenya the poverty levels are high resulting to exclusion of almost half the country‘s
population from private and public prepaid health insurance schemes (MOMS, 2011; KNBS,
2010). Only 4% of all health resources are pooled through health insurance and risk pooling
mechanisms account for only 17.1% of the population. On the basis of wealth status, only
2.9% of the poorest quintile is covered (Chuma & Okungu, 2011; MoH, 2014). As a result
OOP pushes about 1.48 million Kenyans below poverty line while millions lack access to
essential healthcare services and many more are deterred from seeking healthcare services
(Chuma & Maina, 2012, MoH, 2015).
Limited capacity in raising healthcare funds through general and payroll taxes, declining
donor and NPISH funding (KNBS, 2016; MoH, 2015) are some of the challenges that
continue to undermine efforts of improving access to segments of population that are
excluded from mainstream prepaid schemes. Additionally, NHIF monthly contribution rate
for the informal sector does not reflect the ability to pay particularly for the poor and
vulnerable groups. The increasing pressure to financing healthcare from domestic resources
calls for employment of financing mechanisms that can circumvent the ensuing organization
and political challenges. CBHIs have emerged as an innovative healthcare financing
mechanism with a potential of addressing the existing health inequities particularly for the
poor and vulnerable segments of the population (Fernandes et al., 2009; Durairaj et al., 2010
& Schieber et al., 2012). Currently, CBHIs covers 1.3% of the Kenyan population (MoH,
2014).
Given, the high levels of exclusion that results in huge disparities (MoH, 2014) and high
levels of financial risk associated with healthcare costs (Chuma & Maina, 2012) coupled with
increasing pressure to finance healthcare from domestic sources (MoH, 2015) and the
urgency of moving closer to UHC (WHO, 2015a) there is need of examine innovative
healthcare financing and equity through CBHIs in Kenya within the health financing
functions.
11
Various empirical studies focusing on CBHI from the perspective of potential in inclusion
and sustainability have been conducted by Dror & Jacquier (1999); Musau (1999). Jutting
(2000), Preker (2002); Chuma & Okungu (2011); Schieber et al., (2012); Okech (2013) have
extensively studied CBHI as feasible financing options for the informal sector and poor
households. The issue of how healthcare financing functions are modeled within CBHIs for
realization of equity in healthcare has however not been expressly studied. This study seeks
to examine innovative healthcare financing and equity through CBHIs in Kenya within the
health financing functions of revenue collection, fund pooling and purchasing. The study also
looks at the moderating effects of government stewardship in healthcare.
1.3 General Objective of the study
The purpose of this study was to examine innovative healthcare financing and equity through
CBHIs in Kenya.
1.4 Specific Objectives
This study was guided by the following specific objectives:
1.4.1 To establish the effect of enrolment on equity in health care in CBHIs in Kenya.
1.4.2 To establish the effect of mix of contributions on equity in health care in CBHIs in
Kenya.
1.4.3 To establish the effect of risk pooling on equity in health care in CBHIs in Kenya.
1.4.4 To determine the effect strategic purchasing of health services on equity in health
care in CBHIs in Kenya.
1.4.5 To establish the moderating role of government stewardship on equity in health care in
CBHIs in Kenya.
1.5 Hypotheses: Null and alternative hypotheses
1.5.1 H0: Enrolment is not related to equity in health care in CBHIs in Kenya.
H1: Enrolment is related to equity in health care in CBHIs in Kenya.
1.5.2 H0: Mix of contributions is not related to equity in health care in CBHIs in Kenya.
H1: Mix of contributions is related to equity in health care in CBHIs in Kenya.
1.5.3 H0 Risk pooling is not related to equity in health care in CBHIs in Kenya.
H1: Risk pooling is related equity in health care in CBHIs in Kenya.
12
1.5.4 H0 Strategic purchasing is not related to equity in health care in CBHIs in Kenya.
H1: Strategic purchasing is related to equity in health care in CBHIs in Kenya.
1.5.5 H0 Government stewardship does not have a moderating effect on equity in health
care in CBHIs in Kenya.
H1: Government stewardship has a moderating effect on equity in health care in
CBHIs in Kenya.
1.6 Justification of the Study
1.6.1 Sectoral Policy Makers
This study will contribute towards policy debate and dialogue on the appropriate timings of
premium collection in CBHIs, the types and mix of contributions for greater risk equalization
and the suitable mechanisms of achieving equity in contributions in CBHIs as a vehicle of
reducing disparities in healthcare access and providing financial protection for uncovered
segment of the population. The study will therefore provide focal point on a new domain for
action on health financing for UHC.
1.6.2 The Ministries of Health both at the National and County levels, related
Government Ministries and Departments
This study will act as a focal point of the areas that require specific legal and regulatory
framework defining the role of CBHIs within the health financing framework including
enhancing the stewardship role. Additionally, be instrumental in guiding decisions on
resource allocation particularly from the ministries of finance and health and county
government with an aim of creating the right mix contributions in CBHIs.
1.6.3 Researchers and Academicians
The study will contribute to the body of knowledge on innovative health financing and equity
through CBHIs with particular focus on the informal sector. Hence the findings of this study
will be of interest to other researchers who seek to explore factors that influence how health
financing functions are executed in CBHIs. In particular researchers would want to on factors
that prevent CBHIs from achieving their targeted enrolment rates and the cause of
fluctuations in premiums collected and reimbursements with an aim of cushioning CBHIs
13
from financial deficits.
1.7 Scope of the Study
The study focused on innovative healthcare financing and equity through CBHIs in Kenya
within the health financing functions. The researcher observes that there are other community
financing schemes that offer other forms of community financing such payment of user fees
for healthcare at the point and time of use but are not the focus in this study. The target
population for the study was 82 CBHIs registered by the umbrella body (KCBHFA) in Kenya
that had complete and coherent data. The researcher collected pilot data from three CBHIs
under Bidii network, making the number of CBHIs from which data was collected from in
the main study to be 79 CBHIs. The researcher noted that most of the 79 CBHIs had an
average of four members of management team with vary roles. To make inference about the
CBHIs, responses were sought from four members of each CBHIs management team which
translated to 316 members. The sample size was adjusted to 306 using Yamane (1967)
formula. Data was collected between 23rd
March, 2016 and 14th
April, 2016.
1.8 Definition of Terms
1.8.1 Health Financing
Health financing involves mobilization of funds for healthcare, accumulation and allocation
of pooled funds in purchasing of health services efficiently and equitably (WHO, 2010a).
1.8.2 Equity in Healthcare
Equity in health refers to absence of systemic and possibly solvable disparities in healthcare
access, financial risk protection on health related costs and quality of care between social
groups emanating from social, economic and geographical differences (Starfield, 2011;
Mills, Ally, Goudge, Gyapong & Mtei, 2012; WHO, 2015a).
1.8.3 Innovative Financing
Innovative financing refers to non-traditional methods of financing that embrace public-
private partnerships including other financing mechanisms that tap new resources, deliver
14
value or ensure effectiveness in solving development problems on the ground (Taskforce on
Innovative International Financing for Health Systems Working Group 2, 2009).
1.8.4 Universal Health Coverage
UHC is an aspiration that all people have access to quality and effective promotive,
preventive, rehabilitative, palliative and curative health services at the time of need without
suffering financial ruin (WHO, 2010a; 2015a).
1.8.5 Revenue collection refers to the process employed by a health system in determining
and obtaining money to pay for health system costs from households, organizations and
donors (WHO, 2010a).
1.8.6 Enrolment
Refers to the uptake of pre-paid health insurance in CBHIs (WHO, 2010a).
1.8.7 Mix of Contributions
This is a desired mix of public and private healthcare funds for financing UHC (WHO,
2010a).
1.8.6 Risk Pooling
This involves the accumulation and management of financial resources in a way that permits
spreading of risk of payment for health among all members of a pool and not by individual
members of the pool who fall ill (WHO, 2010a).
1.8.7 Strategic Purchasing
Strategic Purchasing entails the search for most cost effective interventions for increased
healthcare access and financial risk protection (WHO, 2010a).
1.8.8 Government Stewardship
Government stewardship refers to the broader overarching accountability over the
performance of the entire health system and eventually over the health of the whole
population (WHO, 2000; Alvarez-Rosete, Hawkins & Parkhurst, 2013).
15
1.8.9 Community Based Health Insurance Schemes (CBHIs)
This is a form of nonprofit community health financing initiative that is characterized by
voluntary membership community involvement in designing and managing of the scheme
(Carrin et al. 2005, Soors et al., 2010).
1.9 Chapter Summary
The chapter articulates on the inequities in LIMCs and the systemic challenges in healthcare
financing that impede reduction of disparities in healthcare access and financial risks. It
describes the current financing systems in developed countries, Sub-Saharan Africa and most
importantly in Kenya. Its demonstrates the need of reducing out-of–pocket payments by
building a health financing system that enables people to pre-pay for healthcare based on
their capacity to pay while exempting the poorest groups. Further, it discusses that emerging
role of CBHI as a vehicle of inclusion of the informal sector and the poor. In addition, it
highlights achievements of other countries in expanding coverage for universal health
coverage that can be replicated in Kenya. This study seeks to examine innovative healthcare
financing and equity through CBHIs in Kenya within the health financing functions. Specific
objectives of the study are derived from the health financing and the stewardship functions.
The findings of this study will be of interest to policy makers, implementers, academicians
and researchers in different ways.
In the next chapter, literature related to health financing for UHC particularly the health
financing and the stewardship functions in CBHIs and their effect on equity in healthcare in
Kenya is reviewed. Chapter three details the research methodology employed in the current
study, detailing the methods used in data collection and analysis in the study. Chapter four
presents the results and findings. Discussions of the findings, conclusions and
recommendations are presented in chapter five.
16
CHAPTER TWO
2.0 LITERATURE REVIEW
2.1 Introduction
This section presents theoretical review, the conceptual framework and empirical review of
literature. Theoretical review describes theories relevant to this study while the conceptual
framework illustrates the relationship between the independent and the dependent variable as
mediated by the intervening variables. Literature review entails a review of studies that have
been carried out on performance of CBHIs.
2.2 Theoretical Review
The importance of a theoretical foundation in research cannot be understated. A theoretical
framework forms the foundation of any research since it is the basis on which knowledge is
construed (Grant & Osanloo, 2014). Lysaght (2011) postulates that a theoretical framework
acts as an anchor on which empirical literature is reviewed, research procedures are chosen
and employed. It also consists of selected theories that guide the researcher‘s thoughts and
understanding of the research topic. Whetten (1989) summarizes the critical components and
conditions that make a sound theory as ‗what‘, ‗how‘ and ‗why‘; and ‗who‘, ‗where‘ and
‗when‘ that give the study a well-rounded logical flow. In the first set constitutes of
elements, what specifies the factors and constructs that have been identified in a study, how
describes their causal relationship while why explains the plausibility of theorized
relationship. The second set constitutes of conditions that sets the range for the theory. In
essence they limit the applicability of a specific theory.
Saunders, Lewis and Thornhill (2016) categorizes theory development in research into three
approaches; deductive, inductive and abductive approach. In deductive approach, the theory
provides framework which research questions or hypotheses and data collection procedure is
carried out. The researcher therefore develops a theoretical framework by deriving theories
from existing literature in the beginning of the proposed study and then subjects it to rigorous
test with the objective of confirming or recommending its modification (Creswell, 2014).
Conversely, in inductive approach the researcher builds the theoretical framework in data
analysis phase. In abductive approach, the research uses cognitive and numerical reasoning to
explain surprising facts that cannot be explicated by existing theories. This process generates
17
new theories or modifies the existing ones (Saunders et al., 2016). This study employed a
deductive approach in developing a structured analysis for examining and understanding
innovative healthcare financing and equity through CBHIs in Kenya.
2.2.1 Social Capital Theory
Vast sources of literature posit that social capital which manifests as trust, cooperation and
reciprocity facilitates collective action and lowers opportunistic behaviour (Field, 2008;
Andriani, 2013). The contemporary implication of social capital was advanced by social
scientists including Bourdieu (1986), Coleman (1988) & Putnam (2000). Putnam defines
social capital as the connections among individuals that originate from social networks and
the norm of reciprocity, trustworthiness, and civic engagement that arise from these
networks. This definition underlines the significance of social capital as an essential catalyst
of coordination and achieving better social and economic outcomes. Operationally, Lin
(2001) view social capital as resources embedded in social ties that are accessed and utilized
by its members.
In the broader view of social capital theory, connections and higher associational activities
inside a community fosters a sense of civic engagement where cooperation, the norm of
reciprocity and mutual trust develops over time which is used to institute collective action,
reduce information asymmetry, reduce transaction costs and establish efficient transactions in
the community (Alesina & La Ferrara, 2002; Andriani, 2013). Recent research on the role
social capital on health outcomes have found that individuals behaviour and health outcomes
are to a large extent determined by social capital attributes rather than rational choices made
by the individuals (Eriksson, 2011). In particular, social capital influences positively on the
value people attach to their health (Donfouet & Mahieu, 2012). It can also promote sharing
and transmission of positive health messages that discourage negative health behaviour such
as alcohol abuse and smoking (Takakura, 2011). Communities that high levels of social
capital are easily persuaded to support health policies aimed at achieving equity in healthcare
(Donfouet & Mahieu, 2012).
18
Social capital is one of the key contextual factors that have greatly contributed formation and
success of CBHIs of extending equity in health care (Catherine & Salmen, 2000). Structural
elements of social capital facilitates formation of CBHIs, influences their composition and
practices. On the other hand cognitive elements of social capital whose key principles are
solidarity, reciprocity and trust informs the community‘s values, attitude, behaviour and
norms resulting to collective action (Krishna & Shrader, 1999). The two elements of social
capital are therefore critical for enrolment, pre-payments, risk pooling and purchasing of
health services in CBHIs (Chen et al., 2012; Mladovsky, 2014). They stir up members of a
community faced by uninsured health risks to pool their resources together for common use.
Mutual trust enhances transmission and assimilation of information hence reducing
information asymmetry (Donfouet & Mahieu, 2012; Tundui & Macha, 2014). Mladovsky &
Mossialos (2006) underline the significance of direct involvement of government through its
official in organizing its citizen and sustaining their involvement.
2.2.2 Stewardship Theory
The stewardship theory was advanced by Donaldson & Davis (1991) in search of a new
perspective of managerial motivation to agency theory. Traditionally the theory of
stewardship in literature is grounded in the principal- agent dichotomy. In contrast to the
agency theory, a steward of an organization is seen as the one who acts in the best interest of
an organization (Hill & Jones, 1992). More recent definition encompasses the welfare of all
stakeholders‘ interests as opposed to only the stockholders benefit. According to Donaldson
& Preston (1995) a steward transcends above an agent and goes beyond the stockholders
interest. He demonstrates unfaltering commitment to upholding the fiduciary and non-
fiduciary obligations of the organization as well as the moral duty to other stakeholders
impacted by organizations actions.
The stewardship theory holds that variations in executives‘ performance depend on the extent
to which the structural situation facilitates the executives‘ action. The question therefore is
whether or not the organizational structure aids the executive in formulation and executing
strategies for high organizational performance. A structure that delineates roles and assigns
authority is deemed to be facilitative (Donaldson, 1985). The stewardship theory shifts its
19
focus from the management to the facilitative and empowering structures of an organization
that deliver superior returns to all shareholders through enhanced effectiveness. According to
Donaldson & Davis (1991) & Donaldson (2008) some situational factors that incline an
individual towards becoming stewards include an involvement oriented environment, a
communalist and low-power distance culture. The theory presupposes that the stewardship
relationship is based on trust and reciprocity between the principal and the steward built
through long-term interaction. Additionally, the steward is presumed to be motivated by a
rational process (Van Slyke, 2006). Given its potential in furthering the interest of all
stakeholders‘ stewardship had a greater potential of achieving and maintaining long –term
benefits for all stakeholders in an organization, the theory of stewardship has been
successfully applied in corporate and social entities in reconciling economic and moral
aspects of the management (Nahapiet, Gratton & Rocha, 2005)
The concept of stewardship in health was introduced by the WHO report of 2000 (WHO,
2000). In health care, stewardship is defined as a broader and overarching accountability
over the performance of the entire health system and eventually over the health of the whole
population (WHO, 2000; Alvarez-Rosete et al., 2013). According to Alvarez-Rosete et al.
(2013) the distinguishing and conceptually useful facet of stewardship lies in its ability to
allocate ultimate responsibility for the health of the entire population. WHO (2000) puts
forward three distinct aspects of stewardship including policy formulation, regulation and
gathering and use of health information for decision making. Despite highlighting the
important role of the stewardship function in healthcare, this description fails to capture the
evolving role of the government as a steward and application of stewardship in different
contexts such as in developing countries. It also fails to identify the specific constructs of
stewardship within which different countries can strengthen stewardship (Alvarez-Rosete et
al., 2013).
WHO (2001) suggests a more expanded view of stewardship in the health sector. The report
from the WHO experts describes stewardship as a function that ‗fosters a culture self-
determination and self-direction among individuals and organizations in the system within an
overall framework of agreed norms and values‘ (WHO, 2001). This view lays emphasizes on
20
operational aspects of stewardship such as ethical, inclusive and proactive that were not
captured in the earlier WHO definition (Alvarez-Rosete et al., 2013). In line with the
expanded definition of stewardship, Veillard, Brown, Baris, Permanand & Klazinga (2011)
suggest an expanded list of domains within which the government can exercise the
stewardship function. They include defining health‘s vision and policy making, influencing
better health through advocacy, ensuring good governance in health systems, ensuring
alignment of health systems design with health system goals, directing health systems
through legal, regulatory and policy instruments and collection, dissemination and use of
health information and research.
While different aspects of this responsibility may be delegated to stakeholders in the health
sector, a country‘s government through its ministry of health remains the steward of stewards
(WHO, 2000). The stewardship theory is particularly important in healthcare because it
entails oversight that is exercised ethically, proactively and with inclusiveness (Alvarez-
Rosete et al., 2013). In effect, it has direct and indirect effect on the outcome of actions of
health system players in their advancement of UHC and its permanence from a financing
point of view; namely improving access healthcare services and financial risk protection
(WHO, 2000; 2010, Alvarez-Rosete, et al., 2013). CBHIs play a complementary role in the
health system particularly in health financing by extending health coverage to segments of
the population that are excluded from mainstream health insurance. Government stewardship
therefore a critical determinant of successful and sustainable health financing in community
based structures such as CBHIs (Preker & Carrin, 2004).
2.2.3 Diffusion of Innovations Theory
Healthcare systems had experienced an increase of innovations that are meant to enhance life
expectancy, quality of care, diagnostic and treatment options, delivery of care as well as
health financing options. Diffusion of innovation remains major challenge in all fields
including in healthcare. Healthcare is posited as one of the fields that have numerous
evidenced based innovations. Despite successful implementation of these innovations in
other areas, their adoption has been slow in other areas (Berwick, 2004). Diffusions of
Innovations theory seeks explain how, why and at what rate new ideas and technology
21
spreads through a social system. The theory was popularized by Rogers (1962). Rogers
(2003) defines diffusion as the process in which a new idea, product, practice, philosophy or
an object is disseminated through structures over a period of time among participants of a
societal system.
Rogers (2003) proposes four main elements that influence dispersion of innovation;
innovation, communication channels, time and social systems. Innovation refers to new idea,
practice or product to an individual or a unit of adoption. Communication channels refer to
the process of generating and sharing information with an aim of reaching a consensus, the
time dimension relates to the rate of adoption while the social systems represent interrelated
social units that are engaged in joint problem to achieve a common objective. Further, Rogers
(2003) identifies two types of innovative decisions that organizations use to adopt an
innovation; collective and authority innovations decisions. Collective innovation decisions
occur through a generally accepted mechanisms consensus while authority decision occurs
when an innovation is adopted by few individuals who possess position power in an
organization. Beyond its original domain, diffusion of innovation theory is applied in policy
transfer where administrative structures and ideas in an institution or a country influence the
development of policies in another (Marsh & Sharman, 2009).
CBHIs have emerged as innovation in health financing as result of poor government
spending in health, political instability and poor governance in the healthcare system (Preker,
2002). The rate at which various countries adopt CBHIs as a strategy for mobilizing
resources to finance and deliver healthcare for the informal sector and the poor varies. Many
countries have progressed towards UHC through gradual expansion and integration of
voluntary CBHIs into the broader healthcare system. Their adoption in some countries has
been slow, a situation that hampers their efforts of addressing the challenge of health
inequities (Criel & Van Dormael, 1999; Poletti et al., 2007; Durairaj et al., 2010; WHO,
2010a).
22
2.3 Conceptual Framework
Miles & Huberman (1994) describes a conceptual framework as a logical system of concepts,
assumptions, and beliefs that support and guide the research plan. It outlines the key factors
and constructs and the theorized relationships among them (Miles & Huberman, 1994, p. 32).
Camp (2001) views a conceptual framework as a configuration that best explains the
researcher‘s conceptualization of the natural progression of the observable facts that are
being studied. It presents the researcher‘s understanding of how the research problem will be
explored, the direction that the research problem is expected to take and how different
constructs in the study relate to one another (Grant & Onsoloo, 2014). Luse, Mennecke &
Townsend (2012) aptly posit that a conceptual framework provides the researcher an
opportunity to specify and delineate concepts in a problem.
This study is based on conceptualized relationship between health financing and equity in
healthcare and the moderating effect of government stewardship. This study proposes a
framework that combines a conceptual and an analytical framework. The framework depicts
the perceived relationships between the inextricably linked health financing functions as the
independent variables; enrolment, mix of contributions, risk pooling and strategic purchasing
and the dependent variable equity in healthcare, with the moderating variable being
government stewardship. The framework also offers a platform for analyzing the
performance of the CBHIs financing system along the financing functions. Achieving equity
depends on how the CBHIs combine the functions under the stewardship role of the
government as shown in figure 2.1.
23
Adapted from Kutzin (2001; 2008)
Adapted from Kutzin (2001; 2008); Mathauer & Carrin (2010)
Figure 2.1 Conceptual Framework
2.3.1 Equity in Healthcare
Equity in healthcare implies that all segments of the population should have a fair
opportunity to attain their full health potential and, more pragmatically, no one should be
disadvantaged from achieving this potential, if it can be avoided (WHO, 2008). Equity is
therefore concerned with bringing health differentials down to the lowest level possible.
UHC is a practical expression for equity in healthcare and the right to health (WHO, 2015a).
Moving towards UHC presents the best possible option of attaining the overall objective of
WHO, namely achieving the highest possible attainable standard of health for all as a
fundamental human right (WHO, 2010a); an indication of the basic principle of the
Independent Variables Moderating Variable Dependent Variable
Government Stewardship Design
Training
Monitoring Co-financing
H5
H1
H2
H3
H4
Risk Pooling -Social solidarity -Mechanisms for enhanced risk
pooling . Size of pools .
Purchasing - Contracting
- Provider payment mechanism
- Referrals - Waiting period
Mix of Prepaid
Contributions
Enrolment
- Affordability of contributions
- Unit of membership
- Timing of collection
- Trust
Equity in Health Care
- Increased access
- Equity in
Contributions
- Sustainability
- Quality of Care
24
Declaration of Alma Ata and the WHO Global Strategy for Health for All by the Year 2000
(WHO, 1978; 1981).
Gilson (2003) observes that health systems are intrinsicly relational. The systemic challenges
of access and financial risks associated with healthcare cost can therefore be construed to be
of relational and behavioural nature. The call for universality with particular concerns on the
health of the poorest and vulnerable segment of the population (Mills et al., 2012)
necessitates exploration of relationships and behaviour that influence equity in healthcare
(Glison, 2003). This study assessed equity based on the equity in contributions, increased
access to health care, quality of care and sustainability of CBHIs as explained in sections
2.3.1.1, 2.3.1.2, 2.3.1.3 and 2.3.1.4. Subsequently, the influence of enrolment, mix of
contributions, risk pooling and strategic purchasing on equity in healthcare in CBHIs and the
the moderating effect of government stewardship is discusses in sections 2.3.2, 2.3.3, 2.3.4,
2.3.5 and 2.3.6
2.3.1.1 Healthcare Access
Access to healthcare has featured prominently in global health policy literature (Xu et al.,
2010; WHO, 2010a). Guaranteed access to healthcare is one of the ways of reducing health
inequities as it is one of the two prongs of UHC. Despite its heighted level of importance and
attention, access to healthcare still remains a complex concept as demonstrated by the
divergent definitions of the notion by various authors. Levesque, Harris & Russell (2013)
conceptualizes five definitions based on the five dimensions in literature of access to health
namely; approachability, acceptability, availability and accommodation, affordability, and
appropriateness. The author further proposes five corresponding abilities of populations that
interact with these dimensions to generate access. These abilities include; ability to perceive
health needs, ability to seek health services, ability to reach health care, ability to pay
without suffering financial constraint and ability to engage through participation and
involvement in management and decision making on issues touching population healthcare
needs.
25
Looking at the dimensions, approachability implies that the sick are aware that some form of
services is offered, the services are within reach and seeking treatment will subsequently
have an impact on the individual‘s health. CBHIs membership rates are determined by the
distance of the household‘s home from the nearest health facility where insured services are
provided. Long distance to health facilities has been cited as a key barrier to equity in
healthcare particularly healthcare access. This barrier arises when patients cannot reach a
health facility due to long distance to health facility, huge transportation charges and lost
wages. The poor and vulnerable segment of the population is affected most (Parmar, Allegri,
Savadogo, Sauerborn, 2013; ILO, 2012). Geographical access can be measured using time
and distance required to access care. The results of these measurements are then compared to
benchmarks. Such standards need to be tailored to urban and rural settings, and should reflect
the existing standards in the community (Carrin, 2003).
In Kenya utilization of healthcare services was found to be negatively related to distance to
health facility (MoH, 2014). Franco, Diop, Burgert, Kelley, Makinen & Siampara (2008)
found that distance to the health facility was a significant negative predictor for healthcare
seeking, particularly regarding assisted deliveries. In Rwanda, Schneider & Diop (2004) also
established that members of 54 community schemes visited providers directly according to
their geographical ease of access. While limited in number, these results suggest that the
policies under consideration have not effectively addressed distance-based discrepancies in
access to insurance and service utilization. The CBHIs management team of Bamwanda in
the Democratic republic of Congo went a step further and implemented a differential fee
based on distance from households to hospital. This approach was however not effectual in
stimulating utilization of health services by members living furthest from the hospital (Criel,
1998). Studies in Nigeria and Burkina Faso established distance to be one of the factors that
influenced enrolment. Households far from health facilities were found to be more willing to
enroll than those near the health facilities (De allegri, Sanon & sauerborn, 2006; Ataguba,
2008). Conversely, earlier studies in Burkina Faso and India reported lower enrolment rates
among households travelling longer distance to health facilities (Mathiyazaghan, 1998;
Dong, Kouyate, Cairns & sauerborn, 2004; Dror, Radermacher & Koren, 2007; Panda,
Chakraborty, Dror, & Bedi, 2013). Distance was however not qualified in all these studies.
26
The barrier to access arises from both direct and indirect costs. Transport costs, a cost that is
rarely included in the benefit package forms part of direct cost. Evidence by McIntyre,
Thiede, Dahlgren & Whitehead (2006 p. 862) suggests that transport costs accounts for one
fifth of all direct cost of seeking healthcare costs. Direct costs arise from poor transport
infrastructure in rural settings which increases the time spent to and from the health facilities.
This results to high opportunity costs in form of foregone wages or income. Combined, the
direct and indirect costs present a high barrier to equity in health care. Evidence from
Burkina Faso shows that coverage by CBHI did not increase health services utilization if the
insured households are located more than 5 kilometers away from a health facility (Parmar et
al., 2013, p.5)
Secondly, acceptability relates to judgment on appropriateness of services based on the way
they are organized. For instance services that are equitable are acceptable to poor and
vulnerable. Thirdly, availability refers to the physical existence of the services that are
offered by qualified health personnel. Fourth, affordability reflects the economic capacity to
pay for health services, have time to travel to the health facility and the opportunity costs
associated with lost income, perceived quality of care and the health providers‘ attitude and
behaviour. Lastly, appropriateness denotes fit between the services provided and patient‘s
needs. It indicates the services provided and the manner in which they are provided. This
process should be integrated, consistence and timely (Levesque et al., 2013).
The concept which is frequently used as a proxy of utilization of health services or coverage
refers to ability to utilize health services when they are needed (Xu at al., 2010; Levesque et
al., 2013). Ability to utilize health services qualified by need is influenced by availability of
services and the ability of the sick to seek health services (Levesque et al., 2013). The
predisposing factors from supply side include location, availability, cost and effectiveness of
the services. On the other hand, ability of a client to use health services is influenced by
many factors. Firstly, people have differing perceptions of the health and as a consequence
have different expectations of their health. Secondly, ability to cater for direct and indirect
costs associated with seeking healthcare and thirdly, non-financial factors such as physical
27
acceptability and other forms of social exclusion and discrimination (Xu at al., 2010;
Levesque et al., 2013).
The foregoing intricacies present challenges in measuring access. In practice, many
researchers measure access in terms of health services utilization or coverage. This approach
fails to capture an important concept that qualifies access; the fact if the members receive the
services they actually need (Xu at al., 2010). Although, adjustments for need can be made,
self-reported data is fraught with a shortcoming of not reflecting the actual medical need.
This is especially true in communities that display heterogeneity in economic capacity. A
study by Salomon et al (2004) focusing on comparability of self-rated health in various
countries showed that the rich are likely to report greater need compared with the poor. This
is despite that fact the poor are in generally worse health since they are exposed to more
health risks in their work places and areas of residence (Smit & Mpedi, 2010; ILO, 2012 &
WDI, 2013).
This study employed reported information on utilization to measure access. This is because
the target population of the study is CBHIs which are composed of predominately the poor
segment of the population (Bennett et al., 1998; Atim, 1998; Jutting, 2000 & Hsiao, 2001).
Additionally, previous studies agree on the active role of community in mobilizing, pooling
and allocation of resources as well as strong solidarity mechanisms in CBHIs (Jakab &
Krishnan, 2001). The documented high level of involvement address the shortcomings on the
demand side by increasing ability of the community to perceive, seek, reach, pay and actively
participate in management of health services. This implies that they are empowered to report
on the ease with which they access health services. Regression analysis was used to
standardize for differences in factors thought to be influenced by the objective need as
recommended by Xu et al. (2010).
28
2.3.1.2 Quality of Care
Quality of care is expressed in terms of disappointment with the outcomes and through a
series of comparisons which they make between their original expectations and the reality,
between the care offered before and after they subscribed (Criel & Waelkens, 2003). From
the perspective of patient rights, patients from all socio-economic levels who seek healthcare
deserve correct and courteous treatment, safe medical conditions, and sufficient information
on their health status and treatment options (WHO, 2010b). It has also been argued that
providing high quality services can lead to increased equity in health care, service utilization
and, in turn, reduce unsupervised and often risky self-treatment (Tipke, Diallo, Coulibaly,
Storzinger, Hoppe-Tichy & Muller, 2008; Robyn, Sauerborn & Bärnighausen, 2013)
CBHIs have been seen as an attractive solution to the challenge of generating financial
resources for the formal health sector in developing countries (Fernandes et al., 2009;
Durairaj et al., 2010; MOMS, 2011; Schieber et al., 2012; Robyn et al, 2013). In particular, it
is a potential instrument to improve access to healthcare by reducing financial barriers to
health services, empowering enrollees through fostering dialogue between communities and
healthcare providers, and improving quality of care by introducing contractual arrangements
contingent on quality standards (Criel & Waelkens, 2003).
Research on quality of care in developing countries has continued to increase over the past
two decades (Baltussen & Ye, 2006). Evidence on the relationship between health insurance
and quality of care in sub-Saharan Africa is scarce. A recent systematic review conducted in
Asia and Africa using randomized controlled trials, quasi-experimental and observational
studies concluded that there was a weak positive effect of both social health insurance and
CBHI on quality of care (Spaan, Mathijssen, Tromp, McBain, Ten Have & Baltussen, 2012).
An earlier study of the Maliando scheme in Guinea-Conakry revealed that participants
viewed rapid recovery, good health personnel, good drugs and a nice welcome at the
participating health facilities as the most important features of quality. When membership
was discussed specifically, lack of quality of care was cited as the most important cause of
non-enrolment (Criel & Waelkens, 2003).
29
Despite the anticipated enrolment rates of approximately 50% in Nouna CBHIs in Burkina
Faso, membership remained low in spite of an upward trend over time. The enrolment rate in
the first year of operation was 5%, increasing marginally to 9% in 2010. The enrollee drop-
out rate declined substantially from 32% in 2004 over time but remains considerable at 16%
in 2010. In 2006, the most common reasons for dropping out, after affordability of the
insurance premium was patient dissatisfaction with the quality of care provided to CBI
enrollees. Patients judged the quality of care to be poor based on the health services they
received, such as drugs and medical staff behavior (Dong, De Allegri, Gnawali, Souares &
Sauerborn, 2009; Robyn et al., 2013). A recent mixed methods study on the relationship
between CBHIs provider payment and health worker satisfaction in Nouna found that
payment attributes such as insufficient level of capitation payments, infrequent schedule of
capitation payment and lack of a payment mechanism for reimbursing service fees strongly
affected service provider‘s satisfaction. It is possible that CBHIs provider payment
mechanisms result to dissatisfaction of health workers and in turn translates into a quality of
care differential between CBHIs enrollees and non-enrollees (Robyn et al, 2013).
This study uses patient experience measures to assess the quality of care offered by the
contracted service providers. Although this measure is relatively new, it is increasingly being
considered important measure of quality of care by experts as efforts to improve the quality
of care continues to evolve. Patient experience measures provide feedback on experiences of
patients‘ seeking health care including interpersonal aspects of care. Additionally, these
measures looks at various aspects of care ranging from clarity to existence of mechanisms
that check on patient perceived quality of care in contracted health facilities on issues
concerning waiting time, availability of staff, services, drugs and supplies. These measures
represent patients‘ needs, values, expectations and preferences and as result reveal critical
information on the extent to which the care offered by the contracted service providers is
truly patient centered. They therefore provide a robust and validated alternative to subjective
reviews (Robert, Berenson, Pronovost & Harlan, 2013).
30
2.3.1.3 Equity in Contributions
Catastrophic health payments can occur irrespective of the amount of the money paid for
healthcare services. Although catastrophic health expenditure occurs in both rich and poor
countries, poor countries contribute to over 90% of the affected (Xu et al., 2003). Shielding
households from catastrophic heath payments remains a principal objective in all nations at
various levels of development (WHO, 2010a; Chuma & Maina, 2012). Previous literature
posits that poorer households are subjected to higher financial risks due to ill health in the
face of competing needs. The case is worse for vast majority of vulnerable households who
require more financial protection to access healthcare (Xu et al., 2003; Su, KouyatЙ &
Flessa, 2006; Ahmed & Mesbah, 2015; Buigut, Ettarh & Amendah, 2015). These aspects of
vulnerabilities and inequities have to be taken into account when designing national and
global health financing policies (Saksena, Xu & Duraira, 2010).
Poor households are exposed to more risks since they work in small workplaces, live and
work in unsafe and unhealthy conditions often not under the purview of labor and health
regulations, are ill equipped in terms of skills and education, have limited access to health
education and prevention programmes and are not aware of their social entitlements (ILO,
2012). They earn low and unpredictable incomes, lack of access to assets, credit, finance,
training, information and technology (Smit & Mpedi, 2010; World Development Index
(WDI), 2013).
Health systems in low- income countries are financed predominantly from OOP (WHO,
2014c). This implies that everyone pays the same amount of money for health services
regardless of their income. When a member of the poor household is sick, the household
have to choose between paying for health services and paying for other basic needs such as
food, rent and children‘s education. Other coping strategies that employed by the poor when
such shocks arise include borrowing from relatives and friends, substituting household labour
supply by sending engaging children in workforce, selling assets and deferring seeking health
services (Asadul & Pushkar, 2012). When there is a compelling need to seek health services,
they risk impoverishment and in worse cases destitution. In addition, indirect health costs
such as transport cost in case of a chronic illness have been found to result to be more
31
prohibitive than direct cost of health services. Besides, the financial shocks arising from
direct cost of health care, such households also face income loss when the sick member is the
working adult(WHO, 2010a).
Equity in contribution relates to the principle of similar contributions for similar ability to
pay (horizontal equity) and the progressivity of a health financing system (vertical equity)
and financial protection. Vertical equity implies that everyone contributes the same
proportion of their income to the healthcare system. Various countries have re-oriented their
health financing system towards attaining progressivity, cross-subsidization and financial
protection. For instance, healthcare contributions in Ghana are based on people‘s ability to
pay. The large cross-sectional risk pools and high enrolment rates enhances cross-
subsidization (Durairaj et al., 2010 and WHO, 2010a). A systematic review of equity in
contribution and distribution of healthcare benefits in low and middle income countries
reveal mixed results. In Asia Pacific countries demonstrated high progressivity across all
financing sources (Asante et al., 2016).
On the contrary, Sub-Saharan Africa OOP and voluntary health insurance contributions were
regressive. Similarly, a study by Chuma & Maina (2012) revealed that Kenyan households
spend over one tenth of their household budget on health expenditure annually signaling high
reliance of OOP expenses among the poor. The poorest households spent a third of their
resources on healthcare payments each year compared to only 8% spent by the richest
households. About 1.48 million Kenyans are pushed below the national poverty line due to
healthcare payments. This implies that healthcare expenditure is not based on ability to pay.
This study measures equity in contribution based on the level horizontal equity in CBHIs.
The measure was appropriate for this study given its ability to measure the extent to which
payments are based on ability to pay and allocation of resources based on health needs (Tao,
Kizito, Qinpei & Xiaoni, 2014).
Additionally, the extent of cross-subsidization based on allocation of claim budget was
employed as a measure of equity in contribution in the studied CBHIs. This method is
preferred due to availability of information on horizontal equity and cross-subsidization.
32
2.3.1.4 Sustainability
CBHIs face myriad constraints attributed to their small size, limited management and
technical skills in insurance and inaccessibility of local service providers due to long
distances and poor medical services. Various studies have documented that CBHIs fail due to
limited managerial competence, financing or a combination of both (Tabor, 2005).
Consistence with other studies focusing on sustainability of CBHIs (Atim, 1998; Carrin,
2003), this study looks at sustainability in terms of financial and managerial sustainability.
Sustenance of equity in healthcare depends on long-term financial sustainability of schemes.
Availability of appropriate managerial skills in accounting, determination of packages,
setting of contributions and management information systems is critical for viability of the
schemes (Carrin, 2003).
The institutional arrangement of CBHIs has significant influence on performance of health
financing functions and more importantly the sustainability of the schemes. Sustainability of
the schemes is critical for scaling up and maintaining the achievements of CBHIs. The
management team usually negotiates with health service providers on behalf of the
beneficiaries. The team also manages the collection of revenue and ensures enhanced risk
pooling (Mebratie, Sparrow, Alemu, & Bedi, 2013). The strong community involvement in
management of schemes predisposes them to risks of sustainability due to limited
management skills available in the community (Preker & Carrin, 2004; De Allegri,
Sauerborn, Kouyaté, & Flessa, 2009).
Particularly, there is low utilization of management information system in CBHIs (Preker &
Carrin, 2004), while the management team lacks premium calculation skills, booking keeping
and accounting skills (Mebratie et al., 2013). Proper book keeping is necessary for effective
monitoring and control. According to LeRoy & Holtz (2012) control systems are necessary
for reduction moral hazard and fraud. They are critical in scrutinizing claims in fee-for-
service payment method where the scheme‘s management needs to verify suitability of
services provided and whether the provider charges reflect the services that were actually
provided. In addition, schemes need to monitor under servicing and provision of low quality
services under capitation payments system.
33
Besides increasing vulnerability of the schemes, these shortcomings hinder their growth. The
small pools are inefficient making it hard for them to achieve equity in health care. To
survive they adapt various techniques. These techniques include; first limiting total payouts
to reduce financial vulnerability related to common shock covariant risks. Second, they
introduce waiting periods to curb adverse selection. Third, they introduce gate keeping
mechanisms, ceilings and co-payments to reduce supplier induced moral hazard. All these
survival tactics reduce attractiveness of the schemes hence impacting negatively on
enrolment. Further, the limited risk pooling hinders cross- subsidization of the poorest and
vulnerable groups (Jacobs et al., 2008, p. 141).
Studies have shown that few management teams negotiate with service providers on price
and quality of health services. As a result this impact negatively on the quality of health
services delivered, which further reduces the attractiveness of the schemes. Small pools
worsen the negotiating position of the management with healthcare providers (Jacobs et al.,
2008). The management‘s capacity to improve members‘ awareness, knowledge, skills and
attitude toward pre-payment and build an insurance culture is also critical in increasing
enrolment (Aggarwal, 2010, p. 28; Donfouet, Makaudze, Mahieu, & Malin, 2011).
Expansive literature on coverage of CBHIs documents that CBHIs fail to cover the target
population. A study by Musau (1999) in Kenya, Uganda and Tanzania found coverage of
1%-7%. According to Tabor (2005) CBHIs cover less than 10% of the targeted population. A
study conducted by Ndiaye et al. (2007) revealed that 95% of the 580 schemes that were
studied had less 1000 members.
Limited managerial and administrative skills critical for viability of schemes was found in 22
CBHIs drawn nine West and Central African countries. The CBHIs management teams were
deficient in pricing, premium collection, and contribution enforcement, determination of
benefit package, contracting of providers, marketing, communication, management
information systems and accounting skills. Even though the schemes has been in existence
for over two decades promotion and marketing of the schemes is still largely driven by
external organizations despite (Tabor (2005). The schemes are faced with challenges of late
34
policy renewals (Carrin, 2003) making long-term planning difficult (De Allegri et al., 2009,
p. 591).
The capacity of the management team in achieving the goals they have committed to has also
been found to influence enrolment rates of CBHIs. Various studies indicate member CBHIs
hardly ever question the integrity and competence of the CBHIs management team (Criel &
Waelkens, 2003; Criel et al., 2002). On the contrary, in a study conducted by De Allegri et
al. (2006) in Burkina Faso majority of the complaints against the schemes management was
linked to failure in achieving commitments particularly in quality of healthcare and failure to
advocate for favourable relations with healthcare providers.
The meaning of financial sustainability varies widely depending on the organizational
structure, revenue structure and the encompassing goal of the organization (Bowman, 2011).
The overarching goal of CBHIs is to achieve their social mission of offering social protection
to the poor and vulnerable by reducing poverty and vulnerability through provision of
community based mechanisms to the low-income households in their efforts to manage risks
in exchange for regular premium payments proportionate to the likelihood and cost of the
risk involved (Bowman, 2011; ILO, 2012).
Broadly, financial sustainability refers to the ability to sustain financial performance over
time relative to risk and capital commitments (Bowman, 2011). One of the distinctive
characteristic of CHBIs is the community base. The community members covered by the
CBHIs determine the benefit package that need. The package is based on the premiums that
they are ready to commit during the policy period. On the contrary, commercial insurance
determines the premiums and benefits offered by the insurance while in social health
insurance the rates and benefit package is decided by the government (Tabor, 2005).
Financial sustainability in the context of CBHIs does not necessarily mean self-financing.
Expansive literature is in agreement that besides the members‘ contributions, contributions
from other partners namely; the government both national and county, donors and other
insurers form a significant source of revenue for the CBHIs (Carrin, 2003; Mebratie et al.,
35
2013). Hence, it is important to look at financial viability within the broader context of the
combined sources of funds. Financial viability of CBHIs can also be evaluated by comparing
the range of benefits offered across CBHIs. Some may not cover chronic cases that attract
repeated and sometimes high payouts. Such schemes report a high percentage of cost
recovery (Carrin, 2003). This undermines the spirit of UHC of access to need health services
for all without exposure to financial ruin (WHO, 2010a).
In addition, financial sustainability of CBHIs can be improved through subsidies and
reinsurance. The two are inextricability linked since subsidies encourage enrolment and
increase the size of the risk pool through increased membership. Besides, subsidies enhance
inclusion of poorer and vulnerable households, thereby reinforcing the social network and
social solidarity; increasing the impetus for UHC. On the other hand, risk transfer
mechanisms through reinsurance enhance viability of CBHIs particularly small pools typical
of CBHIs (Carrin, 2003; Carrin, 2011; Mebratie et al., 2013). Carrin (2003) observes that the
risk pool can be solidified by encouraging the small pools to merge into bigger one by
creation of a network or a federation.
Various studies agree that CBHIs are often unable to mobilize sufficient resources because of
the limited income from the target population. In addition, the pools often small undermining
chances of broader risk spreading and financial risk protection function; one of the key
components of UHC. A study by Private Sector Innovation Programme for Health (PSP4H)
(2014) in Kenya revealed that the limited resource base from the community impacts CBHIs
ability to raise adequate resources. For that reason the pools are intrinsically small hampering
broader risk spreading across the population. The small size of schemes and resource
constraints puts the viability of the CBHIs at risks (Chen et al, 2012; Mebratie et al., 2013).
Government, and its development partners, can support the growth of CBHIs by ensuring
that there is a satisfactory supply of appropriate health services, by subsidizing start-up costs
and the premium costs of the poor, by assisting CBHIs build technical and managerial
competence, by helping to foster development of CBHI networks, and by assisting CBHIs
establish and strengthen links with formal financial institutions and healthcare providers to
36
better manage covariate shocks and catastrophic health risks (Tabor , 2005, Mebratie et al.,
2013).
2.3.2 Effect of Enrolment on Equity in Helathcare
This study explored the relationship between enrolment and equity in healthcare by looking
at the affordability of contributions, unit of membership, timing of collections, and trust
influence access to affordable healthcare through uptake of health insurance in CBHIs. The
sub- constructs were adopted as a measure of enrolment since they are key drivers of
decisions to enrol and not to enroll in community based health insurance schemes (Carrin,
2003).
2.3.2.1 Affodability of Contributions
Wealth quintile is one of the major demographic factors that influence access to affordable
healthcare and increases finacial risk protection from healthcare cost. The higher quantile is
more willing and able to pay more for a health cover than the lower quintile (De allegri et al.,
2006; Basaza et al., 2008; MoH, 2014; Adebayo, Uthman, Wiysonge, Stern, Lamont &
Ataguba, 2015). Dror et al. (2006); Ahmed et al. (2016) & Babatunde et al. (2016) observe
that low income households are willing to pay for healthcare insurance. According to
Dercon, Gunning, Zeithlin, Cerrone & Lombardin (2012) low income households‘ display
high price elasticity due to low and irregular income, a factor that influences demand for
health insurance. The premium set by CBHIs is therefore a critical determinant of
households‘ decisions to enroll. Uptake of health insurance in CBHIs is influenced by pricing
strategies that responds to low income ability to pay including; affordable premium and
flexible premiums (Carrin et al., 2005), payment of premiums in kind (Jakab & Krishnan,
2001), savings linked premium payments (Tabor, 2005), subsidized and exemption of
premiums (Gustafsson-Wright & Schellekens, 2013).
McCord, Steinmann, Tatin-Jaleran, Ingram & Mateo (2012) suggest that premium should be
proportionate to income level low income households for it to stimulate willingness to pay.
Setting premium as a percentage of annual household income encourages purchase of health
insurance (Tabor, 2005). Varied premiums for different family sizes premium structure is
37
also used to encourage signing up. Large families pay less per person compared to smaller
families (Atim, 1999). On the other hand, sliding scale premium is employed among poor
households (Tabor, 2005). Mobile money transfer platform have been used to save money for
premium payments for health cover (Matul, Tatin-Jaleran & Kelly, 2011).
The poorest and vulnerable groups are unlikely to sign up for health insurance in CBHIs due
to lack of financial means to pay for premiums (Tabor, 2005). Basaza et al. (2008); Atagabu
et al. (2008) alluded that lack of financial capacity was the primary reason for non-enrolment
in CBHIs. Premium subsidies and exemptions given to lower quintile influences inclusive
health coverage and equitable healthcare access (Chriatian Aid, 2015). Governement
premium subsidies targeting the lower quintile in CBHIs aids in reducing health inequities
and ensures that the benefits packages address the needs of the poor (Tabor, 2005). The
extent of progressivity of premiums in CBHIs is a key determinant of affordability. Akazili,
Gyapong & McIntyre (2011) observe that flat rate premiums charged by CBHIs do not take
into consideration the disparities in ability to pay among the rich and poor in the community.
In effect, both the rich and the poor pay the same amount of premium. Plausibly, the poor are
not disposed to enroll in CBHIs (Adebayo et al., 2015).
2.3.2.2 Unit of Membership
The unit of membership targeted by CBHIs in critical for increasesing uptake of health
insurance and for controlling adverse selection (Atim, 1998). Most CBHIs have adopted
famiy as a unit of membership. The rational of defining housholds as the unit of membership
is to dissuade hosueholds from enrolling most vulnerable or sick family members. This
requirement creates risk subsidies that are essential for sustainabe risk pooling where the
health subsidizes the sick (Atim, 1998; Criel & Waelkens, 2003; Wodtke et al., 2012).
According to Donfouet, Mahieu, & Malin (2013) recruiting from pre-existing mutual benefit
associations in a target popualtion influences uptake of health insurance in CBHIs It allows
CBHIs to extend membership beyond those who can join voluntarily by exploiting the pre-
existing social networks. Donfouet et al. (2013) & Adebayo et al., (2015) postulate that
households which are part of communal associations such villages; cooperative societies and
38
development projects are more willing to enroll in CBHIs. Defining the percentage of
households in a village would be required to enroll before providing insurance is highly
associated with inclusiveness and sustainability of CBHIs Bennett et al. (1998) & Carrin et
al. (1999) observe that establishing a minimum requirement of enrollees from the poor and
vulnerable members of the community enhances access to healthcare and financial
protection. Desmet et al, (1999); Musau (1999) & Carrin et al. (2001) also put forward that
drawing a defined or allowing an automatic percentage of membership from mutual
associations creates risk subsidies besides increasing membership rates.
2.3.2.3 Timing of Collections
Peridocity of enrolement fee and premium paymnet is a major determinant of enrolment in
CBHIs. According to Tabor (2005) & De Allegri et al. (2006) the main hindrance of
enrolment in CBHIs is the requirement to pay enrolment and or annual premium as a single
payment. High enrolment rates have been registered in CBHIs that a have a flexible policy on
collection of premium, periodicity of premium payment and enrolment procedures (De
Allegri et al., 2006; Defourny & Failon, 2008). To obtain a CBHIs membership, one is
required to pay enrolment fee and premium as a single payment. This obligation acts as a
negative driver of enrolment particularly for large families that may not be able to pay
premium for all members of the family (De Allegri et al., 2006; Defourny & Failon, 2008;
Adebayo et al., 2015).
The period of the year when the CBHIs require households to pay enrolment fees and
premiums can be a driver or a barrier to enrolment. This is in view of the fact that cash
inflows vary from one time of the year to the other. For instance, households are likely to
have more capacity to pay during harvest or livestock sales, making it a more appropriate
time to collect fees (Criel and Waelkens, 2003; De Allegri et al., 2006; Matul et al., 2013).
According to Chen, Liu, Hill, Xiao & Liu (2012) deferring premiums payments to periods of
increased household liquidity increased uptake of health insurance by 6% in China. Mathauer
et al. (2007) observe that a lump sum premium payment influences the modest renewal rates
of poor households in Kenya. The preffered time of premium collections varies between
urban and rural. While spread payments constitute a driver to enrolment in urban areas;
39
annual payments at harvest time are preffered in rural areas (Bennett et al. 1998). Events
such communal meetings have proved to be perfect time for enrolment fee and premium
collection in Uganda (Carrin et al., 2001).
2.3.2.4 Trust
Given the relational nature of health systems the success of interventions put in place to
address challenges in healthcare access and financial risk protection is dependent on the
extent to which the interventions harness the already existing social capital which manifest
into both structural and congnitive elements (Krishna & Shrader, 1999; Catherine & Salmen,
2000; Gilson, 2003). Structural elements manifest in form of institutions created in response
to challenges, their composition and the practices that guide the collective actions. On the
other hand, cognitive elements are composed of trust, cooperation and reciprocity that inform
the values, attitudes and social norms rallying people to collective action and lowers
opportunistic behaviour (Catherine & Salmen, 2000; Putnam, 2000; Gilson, 2003; Chen et
al., 2012; Mladovsky, 2014).
Trust is therefore vital for formation as well as for success of CBHIs in their quest of
reducing health inequities (Krishna & Shrader, 1999; Mladovsky, 2014). Human contact and
connections encourages formation of CBHIs which are high relational activities compared to
informal risk management strategies. Additionally, trust influences enlisting and renewal
decisions. Chen et al. (2012) posits that enrolment rates in CHBIs is affected by three
dimensions of trust, namely, trust among CBHIs members, trust between CBHIs members
and contracted service providers and trust in CBHIs management and in the scheme.
Trust among members of the community increases optimism making members more open to
behaviour change (Takakura, 2011; Chen et al. 2012). High levels of trust persuades
members to try new ideas together such as CBHIs in order to address challenges of uninsured
health risks and barriers of heathcare access. Trust provokes collective actions that speak to
values; attitude and behaviour of members of a community encouraging them to pool their
resources together to cross subsidize each other and improve healthcare access. Trust also
enhances sharing and assimilation of information (Chen et al., 2012)
40
Trust between CBHIs members and contracted healthacare providers is experiential in nature
(Chen et al., 2012). The relatioship is determined by past and current expeience with a
healthcare providers. Quality care and reliability of healthcare providers facilities formation
of strong trust relationships between CBHIs members and health service providers which in
turn enhances enrolment and renewal rates (Jutting, 2004). A health facility that has
adequate and properly trained healthcare personnel, offers a wide range of health services
and has a regular supply of presciption medication is more likely to encourage enrolment,
renewal and utilization health services due to its reputation of trustworthy and dependable
relationship (Jutting, 2004; Chen et al., 2012). Close proximity to contracted health facilities
results to greater interactions between CBHIs members and management of the health
facility‘s. Frequent interactions increases exposure to information and establishes efficient
transactions (Alesina & La Ferrara, 2002; Chen et al., 2012; Andriani, 2013).
According to De Allegri (2006) trust in CBHIs management team is a key determinant of
enrolment rates in CBHIs. Trust in CBHIs management team is dependent on the team‘s
technical competence and collective character. Provision of adequate information and
documentation when enroling and positive past experiences with scheme‘s management team
or mutual benefit associations reinforce community‘s trust in CBHIs management team and
in the scheme itself. Further, the ability of the scheme‘s management to negotiate favourable
contracts with providers, enforce the contracts and achieve set goals strengthen trust in the
scheme and its management. Such contracts relate to quality of care and the range of health
services offered by contracted healthcare providers (Criel & Waelkens, 2003; De Allegri,
2006 & Chen et al., 2012).
2.3.3 Effect of Mix of Prepaid Contributions on Equity in Healthcare
Levies paid at the point of use deter millions of people worldwide from using health services
and makes them to defer checkups when chances of cure are cheaper and high. This promotes
overuse by those who can pay leading to inequity and inefficiency in the way in which health
resources are allocated (Xu, Evans, Carrin, Agila-Rivera, Musgrove & Evans, 2007).
Prepayments and pooling mechanisms present a viable alternative that de-links healthcare
utilization from the need to pay. It requires people to pay for healthcare when they are
41
healthy and more productive and allocate the payments for the duration of their life. Indeed,
some form of prepayments and pooling approach is a critical determinant of the progress
made towards UHC. The level of prepayments and pooling in health system is a critical
determinant of how well a system meets the populations‘ health needs and desires. Countries
that have made quick progress towards UHC have health financing systems that require
individuals to contribute to healthcare through taxation and or insurance based on their
ability to pay. This implies that the rich should pay more while the poorest and vulnerable
groups need to be exempted (WHO, 2010a).
Countries with large informal sectors face difficulties of collecting payroll based health
insurance taxes due to underdeveloped tax systems. Many countries aspiring to achieve UHC
use a targeted approach that focuses on the formal sector. A common resultant of this
approach is a two tier system; a small formal sector that is covered and a large uncovered
informal sector (WHO, 2010a). Additionally, studies focusing on health insurance and equity
in healthcare show that voluntary and private health insurance payment in Sub- Saharan
Africa are moderately regressive (Asante et al., 2016). CBHIs represent one of the
prepayments approaches that have been employed in the journey towards UHC in many
countries. The limited resource base of their target population and their small size influences
their ability to raise adequate resources necessary for greater risk sharing (PSP4H, 2014).
Inefficient prepayments systems in low income countries are a major obstacle to raising
adequate resources through prepayments and risk pooling systems. Government spending
and or donor funding is critical for increased access and financial protection in these
countries (WHO, 2010a). According to Durairaj & Evans (2010) & WHO (2015a)
government spending in healthcare influences access to healthcare and the level of financial
risk protection since it‘s a steady and sustainable source of funds. Moreno-Serra & Smith
(2012) posit that health outcomes are greatly influenced by financing health from pooled
public resources. Globally, there was a marginal increase in overall government spending
from 10.9% in 2002 to 12% 2013. Government expenditure varies greatly; OECD countries
spent 15.6% while Sub Saharan Africa, Middle East and South Asia spent 11.1%, 8.7% and
8.3% respectively. The level of health inequities in the three WHO regions is reminiscent of
42
the low government expenditure in healthcare. Despite the dire need of progressivity in
government spending many LIMCs face challenges of increasing their fiscal space for health.
The amount of funds available from taxes is influenced by the capacity of tax system to
collect taxes. According to WHO (2010) majority of low income countries have a large
informal sector which makes it difficult to collect both payroll and general taxes. The narrow
tax base results to fiscal rigidity. According to Durairaj & Evans (2010) expansion of space
for health is a viable option in developing countries that can be achieved by collecting
additional revenue through tax measures. According to WHO (2015a) many countries have
not exploited available scope for expanding their fiscal space for health. Additional revenue
raised from taxes can be used to fully or partially subsidize the contributions of the poorest
and vulnerable segment of the population. In Gabon, a tax on money transfers raised
approximately US$ 30 million in 2009. The money raised is used to subsidize low income
households. Similarly, Pakistan finances part of its health expenditure from taxes levied on
profits of pharmaceutical companies (WHO, 2010a). Within the context of CBHIs, Rwanda
and Ghana offers examples of African countries that have been able achieve equity by
subsidizes the poorest segments of their population through CBHIs (Humuza, 2011; Atim,
2011).
Some countries have moved back and forth on the best strategies of achieving the right mix
of contributions to meet their population health needs and expectations. In itself, political
goodwill influences the decisions that are made to expand fiscal space for health. Plans of
expanding benefits package in Philippines require expansion of fiscal space for health
through payroll tax reforms. These reforms have proved to be politically unfeasible slowing
the country‘s progress to UHC (Honda et al., 2016). This highlights the importance of
political goodwill in pushing for a country‘s health agenda. The great strides that have been
made by many African in increasing their fiscal space for health is owed to the Abuja
declaration that rallied political support for increasing government spending as a percentage
of THE (WHO, 2013). Rwanda is one of the African countries that have met the funding
requirements for both the Abuja Declaration targets and the HLTIIFHS owing to a
43
combination of multiple sources of funding that are supplemented by donor funds (WHO,
2013).
Taxes on harmful products such as alcohol and tobacco have dual benefits of raising
additional funds and mitigating adverse health and economics effects of their consumption.
The sin taxes are more politically acceptable since they provide progressivity in tax
collections while expanding the fiscal space for health (Bird, 2015). Sin taxes on Tobacco are
more common and have been introduced by Thailand, Nepal, Mongolia and Bulgaria where
all or a percentage of the tax revenue is earmarked for funding healthcare. Thailand has also
earmarked a portion of sin tax from alcohol to health. Similarly, in 2012, Philippines enacted
a new tax law on tobacco and alcohol products that seeks to increase sin tax revenue by 60%
(World Bank, 2014). The sin tax is earmarked for subsiding massive insurance premiums for
poor households (Honda et al., 2016).
Donor funding play a critical catalytic role of supplementing government and prepaid funds.
In 2012 external aid accounted for a quarter of THE in low income countries (WHO, 2015f).
The 2014 CHATHAM House report accentuates the need for additional funding for
realization equity goals. Predictability and harmonization of donor funds with national
priorities and systems remains a major challenge. The International Health Partnership
represents some of the initiatives that seek to rally donor countries and development partners
to adopt SWAp in aid disbursement (WHO, 2010a). Such initiatives would direct donor
support to a country national health strategy that allows the recipient country to fund priority
policy areas in health. For instance, harmonization would enable developing countries to
strengthen their capacity in tax collection and expand their prepayment and risk pooling
system (WHO, 2013).
The mix of contribution varies across countries; majority of developed countries health
systems are tax funded with subsidies for the poor. France has created two types of funds that
are used to cushion the elderly, disabled and the poor against healthcare costs. A national
solidarity fund funded by the social health insurance funds and from proceeds of the
solidarity day is used to pay for elderly and disabled peoples‘ social and health services. The
44
poor are covered through a public supplementary insurance programme that that pays service
providers directly for the cost that they have incurred. The fund pays for co-payments, optical
and dental services. Canada continuous to maintain a stable mix of contributions since 1997,
with the government funding 70 percent of total healthcare cost and the private sector
spending 30 percent of the healthcare bill. In Germany, healthcare is predominantly tax
funded with majority of the population contributing through mandatory statutory insurance
system. The population is divided into three tiers; one of the groups represents the poor
people whose premiums are paid through government revenue (Bidgood, 2013).
Switzerland operates a compulsory social insurance system composed of multiple health
insurance schemes and a government approved plan. To reflect the ability to pay across
different socio-economic groups premiums are set at the community level. Subsidization
policy varies across schemes, the poor are exempted from paying premiums and members co-
pay for doctor‘s visits. On the other hand, health insurance in US is voluntary with employer
based tax subsidies. Majority of the country‘s population receives services funded through
private insurance. Low income children and the elderly benefit from subsidized care and
several safety nets funded through government and private insurance payments. The US
health system is however the most expensive and inequitable with a significant population
still not covered (Bidgood, 2013).
All African countries operate a pluralists financing system with varying mix of contributions.
Healthcare funds originate from government, donors, employers, households and non-
governmental organizations. Capacity to collect taxes remains a big challenge in most
African countries owing to the informal nature of their economies. Health continues to
receive low priority in government allocations (WHO, 2013). According to IMF (2012)
almost half of the countries in the WHO African region receive substantial amount of
revenue from natural resources; in some cases the revenues from natural resources surpass
other government revenues. Despite the substantial revenues collected public funding to
healthcare remains insufficient (Costa-Font, Gemmill & Rubert, 2009). The capacity of
raising additional tax revenue exists in many countries. Ghana funds part of its national
health insurance scheme through a 2.5% increase in value added taxes. Gabon raises funds to
45
subsidize healthcare costs for low income through new taxes imposed on remittances and
mobile phone operators. The funds are channeled through the national health insurance and
social security schemes (WHO, 2013). According to WHO (2010) a 50% increase in tobacco
taxes has a potential of raising US$ 1.42 in 22 low income countries.
Given the foregoing funding challenges donor funding plays a critical role of funding health
initiatives. The amount of donor funds received by African countries varies from less than
20% to more than 40%. For instance, donor funds in Malawi accounted for more than 40% of
the THE between 2001 and 2010. Additionally, the trend of flow of the donor funds varies
from one country to another. Some countries have experienced a decrease in donor funding
while others have registered a boost in donor funding. Burundi and Malawi registered a
substantial increase in donor funding between 2005 and 2010 (WHO, 2013). On the other
hand, Kenya registered a decline in donor funding from 35% in 2009/2010 to 26% in
2012/13 (MoH, 2014). Donor funding is fraught with challenges; it is unpredictable and fails
to meet set targets in most cases (WHO, 2013).
Various African countries have established financing mechanisms that improve access to
healthcare and offer protection against catastrophic health expenditure. Most of these
initiatives benefit children under five years and pregnant and lactating women. For instance,
Uganda operates voucher schemes for pregnant women while Kenya offers free services for
children under five years and pregnant women. Additionally, pro-poor financing mechanisms
entail offering free services in primary healthcare facilities. Kenya offers free health services
in level 2 and 3 health facilities which are funded through a combined government and donor
fund. The insufficient levels of government and donor funding combined with high poverty
levels exposes the poor and vulnerable groups to a disproportionate economic burden. As a
result, healthcare costs continuous to pose significant barriers to healthcare access while
millions are driven below poverty line (WHO, 2010a; Chuma & Maina, 2012).
Within the context of CBHIs, a financing pool composed of different mix of contributions
drawn from membership premiums, government and or donor funds provides subsidies and
exemptions to poorer and vulnerable households, a practice that is significant for achieving
46
an equitable and sustainable resources allocation. In Rwanda, has the mix of contributions in
CBHIs is composed of approximately 50% membership contributions while government,
donors and insurance transfers contribute the remaining amount. Revision of the proportion
contributed by each party is critical sustenance of equity goals. As donors exit, the
government and CBHIs members should take up responsibilities of generating more revenue
(Christian Aid, 2015).
Government and donor funding can be combined to form a health equity fund is used to
cover healthcare costs incurred by the poor and vulnerable segments of the population. The
fund provides cross subsidies across the population since more healthcare resources are used
to cater for the cost of the neediest segment of the population. In Cambodia and Laos health
equity fund is used to purchase CBHIs premiums for the poor. Introduction of the fund was
associated with increased utilization of health services and increased financial risk protection
(WHO, 2010a). Health equity fund beneficiaries recorded lower cases of borrowing money
for healthcare compared with families that pay for healthcare at the point of use. Health
equity funds can be use a basis of lobbying for increased government funding. In Cambodia
the fund has attracted more funding from the government over time (WHO, 2010a).
Illegal payments results to cash flow problems for the provider, which eventually leads
provision of low quality health services. This results to service rejection and client
dissatisfaction making it difficult for the CBHI scheme to attract and retain clients. For
example, perception of poor quality service at Kitovu Hospital, in Uganda, made it difficult
for the KPPS micro-insurance program to obtain new clients (Basaza et al., 2010). Successful
CBHIs build long-term, stable relations with trusted partners. Due-diligence is undertaken
before partnership relations are entered into, and clear written agreements summarize the
responsibilities and obligations of all the partners. Therefore, there is need for increased
support for CBHI schemes and for the establishment of a co-financing scheme that would
complement premiums paid by individuals toward their health insurance with government or
donor funding (WHO 2001; Tabor, 2005).
47
This study uses the mixture of financial sources in CBHIs as ameasure of mix in
contributions. This measure was deemed appropriate for the study since a mix of contribution
in CBHIs have been recommended by various authors given the limited corrections from
members contributions, the sustainability of governement revenue and importance of donor
funds in the formative stages of CBHIs which implies the important role played by each
source in furtherance of equity in healthcare (Preker, 2002; Carrin, 2003 & Durairaj &
Evans, 2010)
2.3.4 Effect of Risk Pooling on Equity in Healthcare
Risk pooling makes an individual‘s health expenditure more knowable and manageable
(Davies & Carrin, 2001; WHO, 2015a). According to Smith & Witter (2004) the rationale of
risk pooling in healthcare in relation to equity and efficiency can be viewed from two
perspectives. First, the equity aspect entails sharing of some or all individual‘s health risks
associated with healthcare expenditure across a risk pool. The equity objective is therefore a
strong determinant of healthcare access and financial risk protection for a risk pool
membership with similar health needs irrespective their individual circumstances. The level
of risk pooling in developing countries is still low due to predominantly high levels of
poverty that diminishes the ability of the poor to pay for health insurance (Smith & Winter,
2004, Parmar, et al., 2013). Smith & Witter (2004) argue that communicable disease
commonly associated with the poor have been predominant in developing countries. This
implies that poor have higher health needs and less ability to pay.
Secondly, risk pooling increases efficiency in the manner in which health funds are utilized.
It facilities transfer of health resources to the poor since their demand for healthcare is higher
given their exposure to more risks in their workplaces and areas of residence (Smith &
Witter, 2004; Smit & Mpedi, 2010). Additionally, risk pooling encourages timely access to
healthcare when treatment options are cheaper and reduces chances of loss of income due to
illness. Risk pooling therefore is a therefore a critical determinant of improved health status
of a population and more importantly in reducing health disparities (Davies & Carrin, 2001;
Parmar, et al., 2013).
48
The current double burden of disease in developing countries results the increase of
uncertainty of illnesses and healthcare costs driven by costly medicines, procedures and
technologies (WHO, 2015a). These costs present a greater burden to the poor in absence of
risk pooling mechanisms (Smith & Witter, 2004). Attitudes towards making choices that
ensure tradeoffs between equity and efficiency is a key determinant of impetus towards
UHC. The level of social solidarity in a society influences the readiness of the rich to cover
the poor and the health to cover the sick, a situation that ensures risk sharing (WHO, 2010a).
Such system allows individuals to spread healthcare costs over their life cycle by
encouraging them to make payments when they are young and healthy and drawing on them
in times of sicknesses. Such arrangements reduce barriers to access and lower the incidence
of catastrophic health expenditure (WHO, 2010a). According to James & Savedoff (2010)
most societies possess a degree of social solidarity from which they can draw on when
establishing redistributive systems. Existence of an aspect of social influences the level of
inequities that a population can tolerate (WHO, 2010a). Previous experience with mutual
organizations influences willingness to share risks and in turn the realization of equity goals
(Goudge et al., 2012).
The composition and size of contributions to a pool influences the amount of resources raised
and in turn the attainment of equity from the benefits drawn by members (WHO, 2015a).
For a pool to raise adequate resources it must draw its membership from large group of
population. This move should be meticulously planned to avoid preclusion of the poor and
vulnerable groups who are desperate for healthcare. According to WHO (2010a) compulsory
contributions are more favorable given their ability to draw resources from a wider social
economic background resulting to risk equalization (WHO, 2010a). Voluntary schemes such
as CBHIs play a critical role of extending the benefits of risk pooling in countries that
experience high levels of exclusion (Davies & Carrin, 2001).
Limited organizational capacity in establishing a social health insurer results in
fragmentation of risk pools. Fragmentation raises the questions of sustainability. Multiple
pools draw their membership from different population groups which limit their size. They
49
also attract tend to attract covariant risk, duplicate benefits and efforts, incur high transaction,
administration and information systems cost and display low negotiating power with
contracted healthcare providers. As consequence of such duplication and inefficiencies they
are not sustainable options of addressing equity goals (WHO, 2010a; Langenbrunner &
Somanathan, 2011; Kutzin, 2012).
Establishing larger risk pools should be part of a health financing strategy from the start. In
order to achieve higher coverage a risk pool can enroll a larger population by targeting a
larger geographical area. For instance, instead of targeting village populations, for example,
the district population could be targeted. Expansion of risk pools can be achieved through
establishment of a federation or network of CBHIs. In that case, there is also greater
likelihood of having cross-subsidies between rich and poor households. This has proved
successful in the case of the Bwamanda Scheme in Congo and in the Nkoranza Scheme in
Ghana where the enrolment was high. It was however not so in the case of the CHBIs
established at district level in Tanzania where the consequent of fragmentation was low rates
enrolment (Davies & Carrin 2001; Aryeetey, Jehu-Appiah, Spaan, Agyepong & Baltussen,
2012).
One of the common themes in recent publications on African and Asian countries that have
realized great impetus towards UHC is pool consolidation (Lagomarsino, Garabrant, Adyas,
Muga & Otoo, 2012). Rwanda started with multiple schemes before consolidating the
contribution from all insurers (Makaka, 2012). Similarly, Latvia, Estonia, Lithuania and
Poland began their compulsory health insurance systems with multiple regional funds before
embarking on progressive consolidation and transformation of the territorial funds into
branches of the national funds (Kutzin, 2010).
A country can decide to consolidate or not to consolidate pools into a national pool or to
maintain separate pools each reflecting distinct needs of a population segment or to
encourage competition through multiple separate pools is based on national priorities as spelt
out in the national health policy. For instance, the Republic of Korea chose to consolidate
over 300 separate pools into one national fund. On the contrary, in 2007, Swiss citizens chose
50
a system encourages market competition by maintaining multiple pools rather than
consolidating them into a single caisse unique. Rwandan tax system is still in nascent stages;
it operates three health insurance organizations. However, there is some degree of risk
equalization where money is transferred from pools that serve low risk groups to those that
serve high risk groups (WHO, 2010a).
Geographical location can have a profound impact on the choice of risk-pooling
arrangement. For example, if development of rural healthcare is a priority, it would seem
sensible to design a system which establishes separate the rural and urban risk pools, with a
robust revenue allocation mechanism that transfer more funds to the rural area. Such a
mechanism will ensure that rural areas have access to a secure stream of finance that is not at
risk from increased demands for healthcare services from their urban counterparts (Smith &
Witter, 2004; Shimeles, 2010; Dutta & Hongoro, 2013). Risk equalization can also be
achieved through allocation of funds from general tax revenues to health facilities in poorer
regions where health needs are higher. By doing so, the government allocates money for the
poor who have less ability pay while the wealthier people who contribute more and have
fewer health needs receive less (Smith, 2008).
The choice of risk-pooling arrangement may be influenced by levels and distribution of
income, and the nature and magnitude of potential revenue bases. Where large differences of
income and healthcare needs exist, inter- pool transfers shield equalizes risks where the high
income members‘ cross- subsidize the low income. This requirement is particularly important
in a system in which there has been substantial purchasing devolution to a large number of
small risk pools. On the other hand, risk pools that have almost similar per capita income and
needs, a less integrated system will be more pragmatic (WHO, 2010a).
Fragmentation of risk pools is more widespread in devolved health system where purchasing
is carried out by the devolved health unit. Fragmentation leads equity and efficient problems
which policy response is to establish an aspect of integrated risk pools for all the devolved
units. This arrangement involves some extent of risk transfer at the national level where
financial transfers between pools are used to offset the variations (Smith & Witter, 2004).
51
Two main approaches have been employed in risk equalization arrangements among
devolved units risk pools. First, capitation payment is used as base of funding of the
combined risk pools. Capitation payment can be defined as actual contribution to a risk
pool‘s revenue that can be associated to a specific member of a pool at a point in time. One
of the advantages of the capitation method is its ability to correct variation emanating from
differences in size of pool while its inability bridge per capita needs variations between pools
is one of its downside (Smith, 2008).
To address the problem inherent in the capitation method, many countries have developed a
second, which adjust the capitation that is affiliated to an individual, based on the
individual‘s characteristics, health status and social economic ranking. The risk adjusted
capitation takes into consideration the variation in risk exposure among the risk pools. Such
arrangement may involve a central system that collects revenue and distribute it to risk pools
depending on their approximated expenditure needs. In an alternative system, risk pools
collect revenue by themselves and make financial inter-pools transfers from low expenditure
to high expenditure pools based on needs of each pools. When the variations cannot not
satisfactory offset among the pools extended transfers can be implemented. Differences in
when the revenues base may result to large discrepancies that will require further adjustments
for adequate risk transfers. Such circumstances arise where considerable differences in
income level exist requiring an additional transfer. The two sets of transfers match the risk
pooling and desired income reallocation needs. In its first round of reforms, Estonia faced
equity and efficiency challenges as a result of failed risk transfer from wealthier to poor
regions. This necessitated creation of tax based equity fund dubbed as the central sickness
fund from which per capita transfers are made to local sickness funds (Smith, 2008).
The choices made at various steps along the path to UHC and at different levels influence the
level of social solidarity exhibited by a population and it in turn influences the equity goals.
Chile enacted a health reform that permits the wealthy people to opt out of the national social
health insurer and instead enroll with competing private risk- based insurers. This reform
resulted in creation of two separate systems distinguished by income and individual risk
characteristics of the people. The ‗opt out reform‘ eroded social solidarity between the rich
52
and the poor. The envisioned goal of reallocating of resources to the poor who are also high
risk was not achieved. Instead the national insurer, which was largely composed of poorer
people, ended up subsidizing the private insurers resulting to equity and efficiency challenges
(WHO, 2015a).
While CBHIs provide risk pooling benefits, their small reserve base makes them vulnerable
to insolvency when faced with large claims resulting from covariant risks. The extent to
which the risk of insolvency can be mitigated depends on the scheme‘s management
technical expertise in risk management right from the design to implementation of their
activities (Fairbank, 2003). Social reinsurance provides risk transfer mechanisms that
guarantee the survival of small risk pools. The viability of social reinsurance is determined
by the schemes efforts in reducing the risks. This is in the view of the fact that social
reinsurance cannot address all the risks facing CBHIs. The practical requirements of
reinsurance mechanism are extensive and involve a complex task of identifying the losses
that are transferrable (Smith & Witter, 2004; Dror & Armstrong, 2006).
Fairbank (2003) proposes two dimensions of financial risks; first, the frequency of risk
occurrence and second, the amount of the financial cost that could be incurred in an event of
risk incidence. Risk pooling in CBHIs involves a combination of the two dimensions usually
at varying degrees. Extreme cases would involve either the CBHIs covering only expensive
but rare medical occurrences or insuring members against inexpensive medical conditions
that happen reasonably frequently. Although some CBHIs are able to insure the two
extremes, it is rather common to find CBHIs that cannot covers the two dimensions without
risking their own sustainability (Fairbank, 2003; Wang and Pielemeier, 2012).
CBHIs income is influenced by the resources base where they draw their premiums from;
low income households. Based on the low and irregular cash inflows to the targeted
households, CBHIs are constrained on the funds they can raise from the premiums. For
CBHIs to spread risks equitably and efficiently, they must improve access and reduce
financial risks to their members. In essence the kinds of risks insured by CBHIs are
influenced by the members‘ needs. Whatever the combination degree and frequency of risks
53
chosen by CBHIs, the risks should be insurable for the schemes to remain viable in the long
run (Fairbank, 2003; Dror & Armstrong, 2006).
Policymakers are exploring alternative risk pooling mechanisms as part of their efforts to
expand the availability and affordability of health insurance cover. From proposals that
would create health insurance exchanges to those that would include an individual mandate,
these alternatives have the potential to significantly affect the composition of health
insurance risk pools and subsequently affect premiums (Scheiber, et al., 2012). In other
words, rules governing health insurance attempt to balance the tradeoffs between access to
coverage and premium affordability (Makaka, 2012). The primary purpose of any insurance
scheme is to share risk between individuals and hence extend financial protection to
members of the scheme (Mills, Ally, Goudge, Gyapong & Mtei, 2012). Different
stakeholders in a risk pooling scheme may have different perspectives on the objectives of a
scheme, and stakeholder objectives will also vary according to the type of CBHI scheme.
In the current study, risk pooling is measured by size, composition, distribution and risk
transfer mechanisms that have been put in place in CBHIs. These measures are enhances risk
equalization among members (WHO, 2010)
2.3.5 Effect of Strategic Purchasing on Equity in Healthcare
As countries transform their health systems towards reducing barriers to healthcare access
and removing financial risks associated with illness, their decisions are not only informed by
enrolment strategies for various social economic groups, revenue collection approaches, mix
of contributions and risk pooling mechanisms but also on critical decisions on purchasing
mechanisms (Carrin, Mathauer, Xu & Evans, 2008; WHO, 2010a). WHO (2000) identifies
two distinct approaches of purchasing health services; strategic and passive. Strategic
purchasing is the active and evidenced based process of identifying a package of healthcare
services that needs to be purchased, select healthcare providers and deciding on the agreeable
methods of purchasing healthcare services (WHO, 2010a). On the other hand, passive
purchasing involves following a pre-determined budget. The design and implementation of
purchasing of health services influences the quality of care, resources allocation, equity and
54
responsiveness in the way health services are delivered and in so doing, it provides an
impetus for UHC (Munge, Mulupi & Chuma, 2016).
In strategic purchasing, the purchaser involves three main parties namely; healthcare
providers, members insured by an insurers or a scheme and the government through the
ministry of health. The purchaser should source for the most cost effective purchasing
mechanisms that addresses the insured populations‘ needs, desires and values and expressly
specifies sanctions and actions agreed on by the purchaser and providers (Munge et al.,
2016). As the purchaser buys health services for people, it is important for the purchaser to
ensure there are effective mechanisms in place to determine and reflect people‘s needs,
preferences and values in purchasing, and hold health providers accountable to the people
(Honda, McIntyre, Hanson & Tangcharoensathien, 2016)
The government role through the ministry of health in strategic purchasing of healthcare
services can be viewed in the light of its role as a steward of health and well-being of a
country‘s population. This is because the purchasing arrangements are influenced by the
policy framework within the country. The ministry of health should therefore set clear policy
guidelines that guarantee redistribution of pooled funds through strategic purchasing with an
aim of addressing distinct population needs (Preker, Liu, Velenyi & Baris, 2007; Honda et
al., 2016). For instance, lack of an integrated regulatory framework undermines realization of
equity and efficiency benefits of strategic purchasing in Kenya (Munge et al., 2016).
Ambiguous policy and regulation on purchaser- provider by the government weakens
efficiency in strategic purchasing. For instance, while the New Rural Cooperative Medical
Scheme in China is entrusted with purchasing of health services and supervision of providers
the government makes the final decision on the providers and has the powers to penalize
poor performing providers; this weakens efficiency and accountability in strategic purchasing
(Honda et al., 2016).
Achieving equitable access to needed healthcare services is dependent on the purchasing
arrangements in a health system. An important task in an active purchasing arrangement is
determination of healthcare needs of the population, the best set of interventions and their
55
suitable mix to meet these needs and desires based on the available resources (Honda et al.,
2016). With respect to delivery of health services equitable access should be addressed from
three fronts; physical, economic and needs responsive services. Imbalances in geographical
distribution of health facilities and human resources are more pronounced in LIMCs (WHO,
2010a). To reduce these disparities, an efficient strategic purchasing system should consider
incentivizing service providers to locate in the underserved areas, compensate members from
underserved areas in order to ease the burden of indirect health cost that they incur as well as
utilize the country‘s disease database as a reference for determination a population health
needs (Honda et al., 2016).
Contracting delivery of health services is problematic since it is difficult to observe and
confirm the delivery of services. Efficient purchasing is therefore hard to achieve since the
outcome of care delivered is determined by circumstances that are not easy observe and
verify (Honda et al., 2016). Capping payments is used to induce providers to deliver
outcomes that are not observable. However, capitation payments leave the provider exposed
to uncertain caseloads. Providers can practice active selection in order earn high rents by
serving low cost patients or by offering low quality services (Preker et al., 2007).
The underlying costs of each provider also determine how efficiently the pooled funds are
allocated for the best outputs possible from the service providers. Paying the same rent to all
providers does not consider the opportunity cost of different providers. An insurer can price
discriminate potential providers by requiring them to commit to different contract based on
the quality of their services. Czech Republic pays a premium price per procedure to high
quality providers with a cap while low quality providers are paid a low price without limit
(Preker et al., 2007; Honda et al., 2016).
The organizational structure of pool may also act as a hindrance to efficient strategic
purchasing. Small pools limit the scale and scope of strategic purchasing function and the
package of health services that can be bought. Over time, the schemes acquires wide
networks with service providers creating a rigid hierarchical connections which weakness the
benefits of mutual a relationship between the purchaser and the provider. Additionally,
56
fragmentation of pooling systems is extended to the purchasing function where different
pools purchase healthcare services separately. This weakens resource allocation and in turn
undermines equity and efficiency in healthcare (Preker et al., 2007).
Contracting in CBHIs is determined by the desire to ensure uninterrupted supply of health
services by contracted service providers. This has seen a departure from the tradition of
contracting local, low cost, public and faith based service providers to contracting some high
cost private hospitals to avoid frequent interference of service provision occasioned by health
workers industrial action. Simplified and easy to understand contract document are
commonly used in CBHIs to enable the CBHIs management committee to comfortably steer
the negotiations and at the same time ensure that CBHIs members understand the contract
and the benefit package. As such it fails to capture the expectations of the purchaser and
provider such particularly specifics of remedial actions in case of breach of contract (Tabor,
2005; Munge et al., 2016). Honda et al. (2016) alludes that members‘ views are not only
important in ensuring that the benefit package reflects their needs and expectations but they
also increase their awareness on the quality of services that they should expect from the
providers.
The payment policy adopted by a purchaser is associated with improved healthcare access
and reduced exposure to financial risk. Some of the advantages associated with fee for
service method include; encouraging providers to be more productive since it rewards the
efforts of a provider and its flexibility in offering premium prices and below cost prices to
influence the quality of services offered (Preker et al., 2007). Fee for service methods
increases utilization of health services as demonstrated by empirical evidence from health
systems in Western Europe, Canada and United States (Bloomberg & Price, 1990; Honda et
al., 2016). The method of payment adopted should reflect the desired efficiency outcomes of
a specific health system. For instance in Philippines weak gate keeping and referral system
necessitated a shift from fee for service payment method to case based payment (Honda et
al., 2016).
57
Capitation curbs adverse selection and hence aids in cost containment. It encourages the
provider to seek low cost health delivery methods such as low cost care, alternative
treatments and technologies. It also encourages providers to focus on preventive measures
which are cheaper than curative. Promotive and preventive health services are particularly
beneficial for low-income groups due to their cost-effectiveness (Preker et al., 2007; Honda
et al., 2016). Additionally, capitation offers flexibility in its administration. Besides its
administration as a flat fee, capitation fee can be tailored to individual demographic,
economic and risk characteristics. Such adjustment can be used in ensuring resource re-
allocation during purchasing of health services; as a consequence achievement of equity. In
Germany, capitation fee in adjusted based on age, gender, family size and disability
(Barnum, Kutzin & Saxenian, 1995).
The extent to which the ministry of health monitors the conduct of health workers influences
whether or not health services are actually delivered at the health facilities. According to
Preker et al. (2007) the policy actions on health workers abseentism and reward of
hardworking health professionals are some of the approaches that can be used to curb moral
hazard during healthcare delivery. In the same way, adverse selection can be addressed by
establishing the underlying cost incurred by cost providers in order to avoid overpaying those
whose cost are low. Close monitoring of the underlying cost is necessary to avoid
misrepresentation of cost information. Moreover the government can establish treatment
protocol as a guide to diagnosis and means of determining the needs of patients and the
course of treatment. Diagnosis based payments can be applied where such mechanisms exist
(Preker et al., 2007; Honda et al., 2016).
According to WHO (2010a) countries should always look at areas of improvement as they
progress towards UHC. Such improvement is sometimes influenced by presence of skills and
technical capability of the steward. Meng & Xu Ling (2014) notes that despite high
population coverage in China, gaps in service and cost coverage weakened healthcare access
and the extent of financial protection due to rising medical costs. Regressive premium
payment risks exclusion of the poor families while schemes covering many low income
families face financial instability. The schemes management were not fully aware of the
58
potential benefits strategic purchasing, an arrangement that could enhance equity and
improve sustainability of the schemes (Honda et al., 2016).
The role played the ministry of health and regulatory bodies in providing direction on
strategic purchasing and in ensuring that strategic purchasing decisions in healthcare reflect
the national health priories and population needs. Equally important is clarification of roles,
uncertainty in delegation to avoid duplication of duties and resources and addressing of
capacity gaps that may exist in various players (Picazo, 2014). Recent evidence from
Indonesia and the Philippines shows that roles, authority and capacity between the steward
and health services purchasers should be clarified. Additionally, the stewardship function at
the national level should be strengthened to enhance the capacity of the government in
promoting efficiency and equity through strategic purchasing particularly with regard to
quality standards, payment methods and systems and regulation of prices (Honda et al.,
2016).
An integrated regulatory framework is a critical determinant of adoption of efficient
purchasing approaches in health care systems. Regulations involve replacement of traditional
systems characterized by hierarchical relationships where reimbursements are based on
services offered. The traditional systems are substituted with formal contracts, decentralized
management and wide variety of providers drawn from public and private providers. Policy
makers should strike a balance between stimulating entrepreneur behaviour and regulating
the conduct of entrepreneurs (Honda et al., 2016). For instance despite their growing role in
extension of coverage to the excluded groups, CBHIs in Kenya report to the department of
Social Services. Additionally, unlike other private insurers CBHIs have no capital
requirements. This arrangement undermines the accountability requirements in strategic
purchasing of health services whose responsibility lies within the ministry of health (Munge
et al., 2016).
Enacting of laws and regulations that guide the operations of players in the healthcare sector
is critical for enhancing their roles in promoting equity in healthcare through design and
implementation of strategic purchasing. Additionally, despite CBHIs accounting to 1.3% of
59
the covered population and their long existence in the country, there no specific rules and
regulation that guides health purchasing arrangements in CBHIs in Kenya. Lack of
government stewardship particularly in controlling the conduct of providers limits the
bargaining power of CBHIs (Munge et al., 2016).
While formulation of policy and enacting rules and regulations is imperative, their
implementation is equally important. Clear role allocation is critical for guaranteed
realization equity priories set in place by the overall steward (Honda et al., 2016). Despite
existence a clear certification policy on identification and enrolment of members who should
benefit from exemption of premiums from the Philippines national government, many
members of the sponsored program were unaware of their right to benefits. To address this
shortcoming PhilHealth mandated health facilities to identify potential beneficiaries and
ensure that they benefit from the program (Picazo, 2014).
The political will and support to health reforms influences the manner in which the role of
the government as a steward is viewed. Support for reforms is often influenced by several
socioeconomic factors including cultural affiliations, historical background, political will and
ideological outlooks (Preker et al., 2007). Similarly, countries follow different paths as they
move towards UHC, some progress faster than others. Strategic purchasing is a critical
determinant of how quickly a country can achieve greater access to health services and
expand financial risk protection. Resources should be allocated based on population needs
and product benefits. Such decisions should be informed by accurate population surveys and
good information systems for timely information management and analysis (WHO, 2010).
A mechanism for receiving and responding to complaints and membership is an enabler of
efficiency in strategic purchasing of health services. The purchaser of health services should
establish a robust complaint and feedback mechanism through which members should air
their complaints on quality of care, availability of health workers, services and prescribed
medication in contracted health facilities. Information technology offers a vibrant platform
for channeling and responding to members concerns. In CBHIs, members express their views
and concerns during meetings (Munge et al., 2016). Although PhilHealth, the national health
60
purchasing agent in Philippines has established a website through which members and
providers views should be registered and addressed, complaints are not addressed on time.
Generally, many insurers have not exploited the social media as a channel for enhancing
communication for providers and insured members (Honda et al., 2016).
Gatekeeping measures such as a referral system improves geographical access to health
services and cost effectiveness in utilization health services. According to Munge et al.
(2016) CBHIs in Kenya have a well-established referral system where members can only
access specialized services in higher level hospitals through referral from lower hospital.
Where referral systems are weak provider payments measures can be used in curbing self-
referral. In Philippines, case based payment method was used to improve efficiency through
rational provision and utilization of services (Honda et al., 2016). This study measured
strategic purchasing by assessing the extent to which CBHIs purchasing decisions influence
providers‘ behaviour and encourage contracted service providers to pursue equity, efficiency
and delivery of quality of care (Munge et al., 2016).
2.3.6 Moderating effect of Government Stewardship on Equity in Healthcare
According to WHO (2000) all health systems perform four functions irrespective of how they
are organized or where they are located. The functions include financing, resource
generation, service delivery and stewardship (WHO, 2000). The renewed focus on the
significance of health system in improving a population‘s health has put the role of state in
healthcare and in the governance of health systems while at the same time recognizing the
growth and diversity of players involved in healthcare provision of into limelight (Alvarez-
Rosete, 2008; Hafner & Shiffman, 2012). Stewardship in healthcare entail an overarching
obligation for guiding the health system, a role which has implications in reducing inequities
in access and financial risks (WHO, 2010a; Alvarez-Rosete et al., 2013).
According to Veillard et al. (2011) there six domains within which the government can
exercise the stewardship function. They include defining health‘s vision and policy making,
influencing better health through advocacy, ensuring good governance in health systems,
ensuring alignment of health systems design with health system goals, directing health
61
systems through legal, regulatory and policy instruments and collection, dissemination and
use of health information and research. The manner in which each stewardship domain is
implemented in health systems influences the realization of equity agenda in a country
(Alvarez-Rosete et al., 2013, WHO, 2014b).
WHO (2014b) elucidates that the policy direction set by a country influences the equity
agenda in a country. Policy formulation entails identifying and clarifying priorities in
healthcare systems including health financing. UHC priorities include provision of essential
health services for all. This ensures fairness, an issue that is at the heart of the equity agenda
(WHO, 2010a). Primary healthcare was endorsed in 1978 by WHO members as a platform of
reducing persistent inequities in healthcare. According to WHO (2010a) various countries
have formulated strategic policy direction that have reduced financial access barriers.
Investment in primary healthcare was highly associated with thrust towards UHC in Thailand
where reforms in health financing focused largely on nationwide expansion of primary
healthcare (Prakongsai, Limwattananon & Tangcharoensathien, 2009). Mebratie et al. (2013)
point out that implementation of CBHIs is getting heighted policy attention in many
developing countries as an alternative financing mechanism for extending coverage to the
poor and vulnerable groups. In practice the use of evidence based information influences
policy formulation and regulation particularly in regard to progress towards UHC (WHO,
2010a).
The inclusion of generation and use of intelligence as feature of stewardship emanates from
its potential of influencing evidence based planning and decision making in a country‘s
health sector. As a mandated steward of health, the ministry of health should establish
institutional arrangement to harness health related information and utilize it in decision
making within the health sector and other related sectors (Mebratie et al., 2013). According
to WHO (2010a) national statistical agencies play a key role of channeling information to the
health and national planning sectors. Information gathered by the agencies on the proportion
of population that have access to needed health services and the level of financial risk
protection influence strategies put in place for realization of equity in healthcare.
Additionally, national health accounts and households health expenditure surveys provide
62
information that is essential for assessing OOP and financial risk protection. Different
contextual situations in different countries require that country customize information
gathering based on their ability to gather, monitor, analyze and interpret information for
decision making and policy formulating.
With regard to provision and purchasing of health services, efficiency and quality of health
services can be improved by considering the set of services needed by each segment of the a
country‘s population, the interventions that can best meet the needs, the appropriate mix of
services, the best purchasing arrangement and the service providers that can best deliver
these services. In reality, such active purchasing influence access to healthcare. Additionally,
allows health consumer in encouraging and enforcing set standards of quality and efficiency
(WHO, 2010a).
The high cost of gathering health performance data from small groups makes its impractical
for CBHIs to gather intelligence (Tabor, 2005). The government can play a supplemental role
of collection and analysis of information. The information gathered can be used in
stimulating establishment of CBHIs, detect problems in existing CBHIs and recommend
practical solutions to the problems (Carrin et al., 2005). A review of CBHIs in Nepal shows
that intelligence gathering was non -existence in all the schemes despite having been initiated
by the government (Deutsche Gesellschaft fur Internationale Zusammenarbeit (GIZ), 2012).
For the ministry of health to ensure that the policies designed to attain equity goals are
implemented, it should possess the requisite tools for steering the entire health sector. These
tools include powers, incentives and sanctions that influence the behaviour of different
players in the health sector (Veillard et al., 2011). The mix of rules, incentives and sanctions
with a regulatory framework influences how successful the ministry of health is able to
influence the behaviours of actors towards achieving equity goals (WHO, 2000).
It is common for the ministry of health to delegate some stewardship responsibilities to other
actors in the healthcare system. Irrespective the stewardship arrangements that exist the
ministry of health remains the stewarded the stewards, in that it takes the ultimate
63
responsibilities for its population‘s health. The powers given to each actor must be
commensurate to the delegated responsibilities (WHO, 2000; Alvarez-Rosete et al., 2013).
Where the ministry of health is expected to implement a national health policy funds should
be allocated for execution of the tasks. Similarly, the ministry of health should ensure that the
implementation of delegated is coordinated (WHO, 2012).
The ability to incentivize the actors in the health sector is another important tool that
influences the implementation of national health policies including addressing disparities in
healthcare access and financial risk protection. In order to secure interest active participation
of sectoral actors the ministry of health should address the discrepancies in resource and
power allocation (WHO, 2012). WHO (2000) advices governments to consider the barriers
that impede achievement of the health system goals and then align the actors‘ behaviour
towards achievement of those goals. Evidence shows that incentives and regulation aid in
influencing the interest of the actors‘. For instance, inequities in access and lack of a
responsive contribution system are some of the problems that impede achievement of UHC.
In recognition of the critical role played CBHIs in extending health coverage to the informal
sector, the Rwandan government uses performance based financing as an incentive to
improve the quality of care offered by healthcare providers. In this case performance based
financing in health facilities encourages health workers to ‗offer quality care to CBHIs
members (Humuza, 2011).
According to WHO (2012) the extent to which the ministry of health deals with the challenge
of actors‘ hidden agendas and possible conflict of interest influences how successfully the
health policies are implemented. As a steward mandated with coordination of partners, the
ministry of health should take a lead role in managing dissonance and mediate divergences
among actors by maintaining continuous dialogue and collaboration with the actors. It
imperative that the steward to clarify and define the role and responsibilities of each actor to
mitigate unnecessary conflicts. Actors who deviate from the norms and standards that are set
by the steward should be sanctioned. Brinkerhoff (2003) identifies self-policing measures
such a professional code of conduct among healthcare providers and incentives to service
64
users that allow them to switch from low quality healthcare providers to high quality
providers.
According to WHO (2008) health equity is affected by wide range of forces including social,
economic and political factors. These forces have both direct and indirect impact on health
outcomes, some over which ministry of health and other sectoral actors have no direct
control. For instance some social determinants that influence equity in healthcare include
water and sanitation, infrastructure, food security, working and living conditions, social
economics status as well as power and resources. This illustrates that the ministry of health
needs to build wider of relationships for effectiveness of its stewardship role (WHO, 2000).
The ministry of health should build different types of relationships ranging from simple
affiliations to formal associations. Similarly the amount of resources and time involved in
establishment and maintaining such relationships vary considerably based on their nature.
As an effective steward, the ministry of health should be flexible and practical in creating and
maintaining partnership with an understanding of the broader determinants of equity in
healthcare. The decision on whom to build partnerships with is influenced by their inspiration
and connections among other factors (WHO, 2000; 2008).
Various countries through their healthcare steward have established formal partnerships with
community financing schemes as an alternative of extending healthcare to the informal
sectors. Germany, Japan, China, Korea, Taiwan, Thailand, Indonesia, Ghana and Rwanda
have moved closer to realization of equity in healthcare by enlisting the excluded segments
of the population through community driven schemes. More importantly, they have defined
the place of community financing initiatives within the context of the national health
financing policy (Preker, 2002; WHO, 2010a; Fernandes et al., 2009; Durairaj et al., 2010;
Schieber et al., 2012).
Accountability in healthcare delivery is critical for improving the performance of health
systems and is so doing contributes to achievement of equity (WHO, 2000). Accountability is
critical given the huge access, information and expertise asymmetries that intrinsically exist
65
among oversight bodies, users and providers. It is therefore important to first, clarify roles on
which that each player will be held accountable. Actors in health sector include from the
users, providers, professional bodies, donors and government. They represent diverse groups
which reflect the complex interdependences among the players. The government as a steward
bears the responsibility of establishing institutions and mechanisms that ensure that actors
adhere to acceptable conduct (Alvarez-Rosete et al., 2013).
Such mechanisms include reduction of corruption, fraud and misuse by requiring
transparency in execution of social security policies and openness to public scrutiny in health
system players‘ activities; ensuring compliance with laid down processes, procedures and
standards and sanctioning unacceptable conduct through the professional licensing, judicial
and legal framework .The accomplishment of these activities depends on the extent to which
line ministries implement governance principles in sectors under their ministries (WHO,
2000; Alvarez-Rosete et al., 2013). For instance, despite the patients being the focus of
service delivery, their role in ensuring accountability is not clear and enforceable. Huge
information and expertise asymmetries and the power of health services provider makes it
hard for the users to obtain and interpret health information. These constraints make it hard
for them to demand for accountability (Brinkerhoff, 2003).
Secondly, clarity of specific areas in which the actors should demonstrate and account for
performance based on the agreed targets. Some of the determinants of performance include
services, outputs and results of programs and government agencies. Performance
accountability focuses on outcomes such as increased access, quality of care, patient
satisfaction and equity. For instance, the ministry of health is responsible for setting priority
actions, regulations and overall performance of health systems in realization equity goals
(Travis et al., 2002). Various appraisal tools are used measure performance at various levels;
from service delivery points to the top management. At service delivery point customer
satisfaction surveys and complaints and feedback mechanisms are more appropriate and
valid. A strategic plan which clearly defines expected outcomes and performance
measurements is a suitable tool for top management (WHO, 2000).
66
The extent to which equity in healthcare is achieved depends on the level of congruence
between organizational structure of the actors and national health policy (WHO, 2000). As
the main actor, the government should ensure that the linkages among the actors are seamless
with clear roles and efficient communications. A fit between the organization structure and
policy objectives should enhance equitable and efficient utilization of resources and engender
a supportive management culture. Organizational misfit may arise from gaps that occur when
separation of functions are not complemented with the necessary organizational changes. It
may also arise when the structures created are not backed by law or when there duplication of
tasks (WHO, 2000; Alvarez-Rosete et al., 2013).
Communication facilitates exchange of information critical for planning and implementation
of actions that key for achievement of equity (WHO, 2000). For instance communication
between information generating agencies and the purchasers ensures that the purchaser
focuses on the populations‘ health care needs and allocates adequate resources for the
neediest segment of the population. Similarly, the communication between the donors and
government facilitates clarification of policy priorities for funding (WHO, 2010a).
Interministerial communication particularly between the ministry of finance and the ministry
of health helps reorienting governments spending towards healthcare. Closer collaboration
between the two ministries is critical for making a case for increased allocation to healthcare.
According to WHO (2013) interministerial committees helps in demystifying the myth of
health being an unproductive sector. Involvement of the ministry of finance from planning
through implementation and monitoring aids in setting the stage for evidence based dialogue
about health and its contributions to national development. Continued exchange of ideas and
information between bilateral and multilateral donors and the ministries of finance and health
help in making a case funding priority areas in health (WHO, 2001; 2013).
The desire for change at the community is evident in the periodic population surveys (WHO,
2010b). At the grass root level service providers and CBHIs have be used as channels of
change for achieving universality. Chen et al. (2012) point out that success in execution of
CBHIs products is dependent on effective communication among stakeholders particularly
service providers and the beneficiaries. Advocacy and information campaigns create
67
awareness on the importance of health insurance. Additionally, trust is critical determinant
for enrolment and renewal of policies at the community level. Various studies have shown
that trust is built and maintained through frequent communication. Meetings and field visits
increases helps in nurturing trust since they reduce physical proximity. Civil society groups
helps in eliciting political support needed for enacting requisite laws for expanding fiscal
space for health needed for achieving universality (WHO, 2010a). This study adopted
schemes design of CBHIs, monitoring CBHIs related activities, as a trainer and as co-
financier as measures of stewardship in CBHIs. The four tasks were deemed critical for
steering CBHIs in the direction of equity in healthcare (Carrin, 2003).
2.4 Empirical Review
A review of the past literature has identified a variety of equity in healthcare and health
financing resources. Policy makers and researchers recommend that countries aspiring to
achieve equity in healthcare to embrace innovative health financing mechanisms that
diversify domestic sources of funding for heath for a guaranteed rapid coverage of the
informal sector and inclusion the poor. CBHIs have been used as a strategy for mobilizing
resources to finance healthcare for the informal sector and the poor. This section presents
empirical literature on equity in healthcare within the health financing functions and related
concepts.
2.4.1 Effects of Enrolment on Equity in Healthcare
Equity in healthcare can only be achieved when the excluded segment of the population is
enrolled in a health insurance scheme. The percentage of population that has signed up for
CBHI compared to the target is critical given the voluntary nature of CBHIs. A low
membership rate is an indication of adverse selection while broader membership is critical
for long-term viability of CBHIs (Jütting, 2004, Chen et al., 2012). The presence of
willingness to pay has been documented by various authors (Dror et al., 2006; Ahmed et al.,
2016; Babatunde et al., 2016). The enrolment strategies employed by CBHIs should
therefore respond to willingness to pay. Uptake of CBHIs health insurance is influenced by
affordable premiums, flexible and varied options of payments, trust and unit of enrolment
(Carrin et al., 2005).
68
Generally, premium should be proportionate to income level low income households
(McCord et al., 2012). However, establishing an affordable price for low-income households
is an intricate task (Churchill, 2006). This is in the view of the fact that low income
households‘ exhibits high price elasticity for demand as a consequence of low and irregular
income (Dercon et al., 2012). Correspondingly, given voluntary enrolment of CBHIs,
affordability of premium is often cited as the main determinant of membership (Carrin,
2003). Leftley (2005) rule of thumb recommends that insurers targeting the poor should work
with members to establish the cash they can spare on an average day before making cost
adjustment. Low income holders‘ purchase decision is influenced by their perception of
products cost and benefits (ILO, 2012). Furthermore, pricing products for this market
requires that the CBHIs achieve a delicate balance of equitable, affordable premium, benefits
and sustainability. Choosing Healthplan All Together (CHAT), an innovative model offers a
quick and practical tradeoff between costs and benefits since it assists low-income
households in choosing benefits based on their ability to pay (Dror, 2007).
Various studies have focused on the issue of premium affordability. An extensive study
conducted by WHO targeting 82 community health insurance schemes covering populations
outside formal employment in developing countries within the context of membership and
coverage found that the premiums in the Nkoranza scheme in Ghana varied from 5 to 10% of
annual household budgets. In the Rwandan Project Study, premium varied from 5.6% to
7.7% in the lowest and highest income category, respectively. Flat rates contributions are
regressive and disadvantage the poorest while sliding scale contributions are sometimes not
statistically significant like in Rwandan case. One indication though in this study is that
affordability matters, is that large households with more than five members had a greater
probability to enroll in the CHIs than others (Schneider and Diop, 2004). A Thiès Study in
Senegal that sampled four villages in which CBHIs operated through a two stage sampling
procedure found income to be a significant factor in explaining enrolment. Belonging to
lower and upper income quintile decreased and increased the probability of enrolment,
respectively. 31% of the wealthiest quintile was insured compared to 8% in the poorest
quintile (Jütting, 2004; Diop, 2005). A recent survey of CBHIs in Kenya focusing on
purchasing decisions in CBHIs shows that there is a high relationship between price and
69
perceived product benefits. In effect members sometimes shun product that are low priced
due to the perception they many not deliver value (Munge et al., 2016).
CBHIs have modified their pricing policies to increase access of healthcare to the poor and
vulnerable segment of the population. In order to improve affordability, membership
enrolment periods are typically long and follow harvest times, thereby maximizing the
probability that households have cash available (ILO, 2013, n.p.). A multiple methods study
aimed at establishing the lessons learnt by CBHIs in Ghana and Nigeria within the context of
enrolment, partnership and policy framework found that some CBHIs in Ghana and Nigeria
sometimes allow installmental and in-kind premium payment (Christian Aid, 2015 &
Atagabu, 2008). Experience from a CBHI in Ghana suggests that allowing in-kind payment
has increased enrolment rates since households could afford to pay in-kind (Chankova,
Sulzbach & Diop, 2008, p. 268).
Non-affordability of the premiums by the poorest segment could be addressed by subsidizing
or exempting their premiums. Poor people are willing to pay a part of their premium if their
contributions are supplemented by a government subsidy. Payments based on the
households‘ ability to pay are deemed to be progressive. Lack of reliable data on households‘
income for the informal sector is cited as the main reason why CBHIs charge a flat premium
which is regressive for poorest and vulnerable households. An income dependent fee
redistributes income and is therefore a more equitable way of raising healthcare funds
(WHO, 2010a). Using a health demand model to estimate the price-related change for
various types of healthcare Dong et al. (2009) found that premium adjusted for income or
subsidies for poor increases enrolment in CBHIs and in effect equity in enrolment in Burkina
Faso. Data was collected from a household survey of 988 sampled using a two stage cluster
sampling approach.
In Rwanda, CBHIs were able to achieve progressivity with time; they started with a flat fee
but later changed to income based fee backed by donor subsidies. Descriptive analysis of
data obtained from CBHIs based on evaluation of experiences gained through years of
membership indicate that access of health care improved drastically when subsidies were
70
used to ease financial access of the very poor, resulting to access of one in every six
Rwandans (Kalk, Groos, Karasi, & Girrbach, 2010). Further, an empirical study by Dhillon,
Bonds, Fraden, Ndahiro & Ruxin (2012) investigating the impact of subsidized enrolment
payment in Mayange CBHIs in rural Rwanda with approximately 25,000 people found that
100% coverage was realized through subsidies introduced between February 2007 and April
2007. Regression analysis was used to analyze the data from this survey research.
A wide range of experiences from China, Thailand, Vietnam, Korea and Indonesia shows
that incentives are critical for stimulating enrolment. These countries have successfully
scaled up CBHIs (Poletti et al., 2007). Government subsidies have longer lasting effect given
their higher potential of sustainability (Ridde, Haddad, Nikiema, Oudraogo, Kafando &
Bicaba, 2010). Adequate financial resources are necessary to cover the subsidized premiums.
Although, the law on national health insurance exempts the poorest from paying the premium
in Ghana although,, the proportion of the poorest among the insured decreased from 30% in
2005 to 1.8% in 2006 due to lack of enough resources to cover the subsidized premiums
(Oxfam International, 2008).
An empirical study by Desmet, Chowdhury & Islam (1999) on community‘s mobilization for
participation in organizing and managing health care delivery and financing using data
collected from two largest rural and non-governmental health insurance schemes in
Bangladesh in a survey research design that used descriptive analysis found that a pro-poor
policy that differentiates contributions according to socio-economic groups where the
contributions for the destitute were 1/10th
of the contribution proposed to the highest income
category increased membership rates among the two lowest socio-economic groups. Renewal
contributions and user fees for consultations and medicine, and caesarian section were also
differentiated: the poorest categories pay the smallest co-payment or face no charge as in the
case of medicine (Morestin & Ridde, 2009).
Another option would be to exempt the indigent from all co-payments. This is theoretically
the case in Rwanda, but in reality recent observations in the field shows that this exemption
is rarely respected (Durairaj et al., 2010). In the WHO study only 13 of the 44 schemes
71
surveyed in the WHO Study had exemption policies to allow the poor households to join
(Carrin, 2003). For instance in one of the three districts in the Rwandan Project, the local
church paid for the contributions of about 3,000 orphans and widows with their family
members. However, after 15 years of operation of the GK scheme, 20% of the ‗destitute‘
group and more than half of the ‗poor‘ group had still not been reached. The contribution
levels and other payments are still said to be too excessive especially for the poor and
vulnerable (Desmet, Chowdhury & Islam, 1999; Morestin & Ridde, 2009). Similarly,
exemption mechanisms were found to be ineffective in Tanzania in an empirical study
conducted by Msuya, Jütting & Asfaw (2007). The study evaluated the impact of community
health funds in lowering barriers to health care access. In effect, like loans, installment
payments mostly benefit the moderately poor.
Achieving adequate membership rates is likely to be easier when households or even
villages, cooperatives or mutual benefit societies are taken as the basis of membership.
Adopting an appropriate unit of membership is critical for increased uptake of CBHI.
Targeting households as opposed to individuals extends schemes membership beyond
voluntary membership. A study analyzing the potential of 50 Mutual Health Organizations in
contributing to healthcare access and extending social protection to disadvantaged population
in six countries in West and Central Africa conducted by Atim (1998) using an inventory
survey and case study methodology found that using family as an obligatory unit of
membership boosted enrolment rates. The study used both qualitative and quantitative
methods of data analysis. This mitigates the problem of adverse selection. Given the high
association of poverty and higher family size (Wodtke, Elwert & Harding, 2012), flat rates
have been found to increase the probability of enrolment of poorer and vulnerable families.
Various researches focusing on unit membership have shown the majority of CBHIs had the
family as the unit of membership (Carrin, 2003; Carrin et al., 2005). Carrin (2003) found that
a number of schemes had actually switched to this type of membership, after experiencing
problems of adverse selection, as a result of families signing up ill family members or family
members most prone to consume health care. Also, most of the case studies (14) reviewed in
the WCA study had an automatic family coverage (Atim, 1998).
72
Some schemes have gone beyond adapting family as a unit membership and set a minimum
percentage of households in a village would be required to enroll before providing insurance.
For instance Kasturba Hospital scheme in India set at minimum enrolment of 75% of poor
households in a village while the Vietnam Health Insurance programme recommended
insuring adequate numbers of children by establishing a minimum of 50% per class (Bennett
et al., 1998, Carrin et al., 1999). Correspondingly, some CBHIs in Uganda have defined a
minimum of 60% membership from mutual benefits societies (Carrin et al., 2001). The same
is true for Grameen Health plan in Bangladesh and Mburahati scheme in Tanzania. In
Bangladesh, participating in Grameen Bank credit programme guarantees automatic
membership of the scheme while Mburahati scheme targets the entire membership of co-
operative societies (Desmet et al., 1999; Musau, 1999).
The requirement to pay one fixed annual installment is a common practice in many CBHIs.
The periodicity of the payment of premiums seems to influence the decision to enroll
especially for the poor and vulnerable groups. Indeed, it appears that the obligation to pay the
enrolment fee and/or the yearly membership premiums in one payment constitutes an
important obstacle, in particular for the poor and vulnerable. De Allergri et al. (2006)
demonstrates that enrolment in CBHIs is related to higher social economic status using data
collected from a population based case control among 15 community offered insurance in
2004 in rural Burkina Faso. The former used conditional logistic regression to explore the
relationship between the variable. Adoption of approaches that make premium payments
more flexible is critical for enrolment and renewal of policies in CBHIs given the low and
irregular income earned by the poor households (Smit & Mpedi, 2010; ILO, 2012).
Spreading enrolment fee and renewal fees are some of payment policies that encourage
enrolment (Criel, 1998; Criel, 2002; De Allegri et al., 2006).
Flexible payments which correspond with cash inflows from harvest and livestock sale are
likely to significantly boost insurance enrolment and renewal rates (De Allegri et al., 2006;
Chen et al. (2012). An empirical study by Chen et al. (2012) focusing on the effect of
deferred insurance premiums offered through credit vouchers established that enrollment
rates rose by 11 percentage points. Using self-administered questionnaires to collect data
73
Mathauer, Schmidt & Wenyaa (2007) evaluated 23 focus group discussions. Data analyzed
using content analysis and descriptive statistics revealed that inability to pay lump sum
premiums when they fall due influenced the poor not to enroll according to a study
conducted in Kenya.
From the WHO Study, it was observed that schemes in urban areas were more inclined to
establish monthly or quarterly contributions so as to match the income patterns of urban
informal sector workers. Annual contributions, collected at the time of harvest of cash crops,
seem to be prevalent among schemes in rural areas (Bennett et al. 1998). However, empirical
study by Ron (1999) on the role played by CBHIs in Philippines and Guatemala in reducing
financial barriers to seeking care with data collected from 3000 families, analyzed using
descriptive statistics found that the ORT Health Plus Scheme (OHPS) in the Philippines
flexible premium payment plan increased enrolment. The payments are made monthly,
quarterly or semi-annually. Other schemes link the time of payment of the contribution with
a suitable event in the community. For instance, burial societies in Uganda use their monthly
meetings for the collection of premiums, for both first-time members or for those who renew
their membership (Carrin et al., 2001).
An empirical study by Basaza, Criel, Van der Stuyft & Basaza (2007) on reasons for low
enrolment in two CBHIs using a case study design with data collected through key informant
interviews, exit polls on both insured and non-insured and review of schemes records found
that revealed that spreading their premium payments over the year greatly increased their
membership. The study employed a framework method for data analysis. A study
investigating the determinants of mutual health organization in Ghana, Mali and Senegal by
Chankova, Sulzbach & Diop (2008) found that the level of household wealth had more
influence on insurance membership in Ghana, where the CBHIs being studied collected the
premium in a lump sum, and less influence in Senegal and Mali, where payments were
spread over the year. The household data was analyzed using multiple regression analysis.
The period of the year when enrolment fees and/or membership premiums are collected can
also favour enrolment or, on the contrary, constitute an obstacle, according to whether it
74
takes into account or not the seasonal fluctuations of income (Criel, 1998; Atim, 2000; De
Allegri et al., 2006). An empirical study by Criel & Waelkens (2003) evaluating the reasons
for declining subscriptions to Maliando mutual health organization on Guinea-Conakry using
data collected from current and past enrollees and non-enrollees validated through focus
group discussions found that collecting fees during the harvest period, which seems more
appropriate in some contexts, does not guarantee individuals' capacity to pay the due amount.
Due to their local presence CBHIs can offer innovative approaches of premium collection
such as paying premiums in kind makes insurance accessible to everyone, including
chronically poor. This increases inclusion of the poor, boost renewal rates and increases
willingness to pay; and as a result promotes equity in health care. Another possibility is to
pay the premium by giving work time to the insurance (for example, in a field from which
the harvest is then sold). This alternative should however not be used to exploit the already
vulnerable households (Jakab & Krishnan, 2001). A survey conducted by Asfaw & von
Braun (2004) on the role of CBHIs in protecting against the downside health effects of
economic reforms in rural Ethiopia using household data and double-bounded dichotomous
choice contingent valuation method of data analysis revealed that the poorer households
preferred to pay their premium with work. The households are nonetheless subject to
exploitation in view of the fact that the premium in kind they give as work is higher than the
amount they would have paid in cash.
Trust is an important factor when considering CBHI enrollment, given the amount of risk
that is inherent in the nature of insurance schemes. Catherine & Salmen (2000), refer to trust
and social cohesion as ―the most important single factor determining the success of any
external intervention was the degree of trust already existing in a particular community‖.
CBHI enrollment rates are likely affected by three manifestations of trust relationships—trust
in others within the community, trust in health providers covered by the scheme, and trust in
the CBHI scheme and management team (Chen, Daukste, Przybyl & Fechter, 2012).
Trust among individuals in a community can effectively be assessed through examining
social networks. Woolcock (1998) & Fay (2005) postulate that the poor in both urban and
75
rural settings rely on existing social networks to manage risk, although the formation of these
networks and the outcomes they are designed to fulfill can vary greatly. In rural
environments, trust is based primarily on the relationships created by traditional customs,
ethnic groups, and common occupations, rather than other social arrangements (Fay, 2005).
Conversely, in urban settings, individuals look to build trust relationships with one another
according to the degree of reciprocity and/or mutually beneficial support that can be derived
from those relationships, rather than through kinship ties as often found in rural settings
(Jellenik in Fay 224, 2005). Generally, urban networks tend to form with greater diversity
and are generally larger in size; however urban networks are usually more susceptible to
instability as a result of the transient and impermanent nature of urban dwellers (Jellenik qtd
Fay, 2005).
In terms of trust in health providers covered by the CBHI scheme, factors such as the
availability, quality, and reliability of health providers have been found to be significant
determinants of enrollment. Trust relationships between health providers and individuals are
generally influenced by experiential lessons which are mainly composed of the past and
current experiences. The trust relationship between individuals and health providers covered
by the CBHI scheme is based largely on experiential knowledge comprised of past and
current experiences. An analysis of qualitative data from 13 rural and urban CBHIs drawn
from multiple regions across the world established that there exist high level of trust between
the members and the CBHIs management team. Further, the perception of fairness and
transparency in schemes is better positioned to nature trust relationships with the community
(Chen et al., 2012).
The main challenges associated with the formation and facilitation of trust relationships
between individuals and organizations in urban and rural settings are similar to the
challenges faced in building trust relationships among individuals, namely the role of
geographic proximity and residential transience. For trust to arise there must be consistent
and sustained behavior. Transience impedes the formation of trust, as it undermines the time
and effort individuals and/or health service providers are able to invest to build the
foundation upon which trust can develop (Chen et al., 2012). The high levels of transience in
76
urban settings make it difficult to build trust between urban residents and contracted health
service providers despite their closer proximity to health providers. Empirical study by
Tibandebage & Mackintosh (2005) on the constitution and destruction of trust within
Tanzanian healthcare transactions using data from patients interviews and secondary data on
their social economic status; charges paid and payment methods within the context of service
providers revealed that transience among health providers particularly private health
providers in Dar es Salaam, Tanzania, notably dispensaries, opened and closed regularly in
an attempted to remain financially sustainably in a fiercely competitive low-income market.
While physical proximity proves to be a relatively insignificant impediment to the formation
of trust between urban residents and health providers, there is evidence that indicates the
reverse may be the case in rural settings. In a study of the Vimo Self Employed Women‘s
Association (SEWA) CBHI scheme in Gujarat, India using a prospective cluster randomized
controlled trial study design and data collected from a baseline of 713 claimants and 1440
claimants after two years from CBHIs drawn from 16 rural areas found that non-financial
barriers, primarily the distance to health service provider, were found to exclude the poorest
of the poor when left unaddressed. The survey data was analyzed using principal component
analysis and the impact distance had on preventing trust from forming between rural resident
and micro health organizations is reported in terms of those excluded. It was found to have a
disproportionate effect on rural SEWA membership (Ranson et al., 2007).
The trust that potential CBHI members have in the health insurance scheme is a significant
determinant of whether or not community members will decide to enroll in the scheme.
Ozawa & Walker (2009) suggest that that respondents who were newly enrolled or had
renewed their membership in the local CBHI scheme expressed higher levels of trust towards
insurers they had good past experience with than those that had never enrolled or had
dropped out. In addition, the research showed that despite the high levels of trust in
Cambodian villages, individuals with poor previous experiences with other organizations
were less willing to trust CBHI insurers. The analysis was based on data collected using a
questionnaire that was administered on 560 households and focus groups selected by
stratified and population proportion to size sample. Factor analysis, test retest and
77
multinomial logistic regression model were used to identify dimensions of trust, measure
trust levels and explore association between enrolment in CBHIs and trust in insurers.
Schneider (2005) evaluates trust in micro health insurance in Rwanda with data collected
using a questionnaire from a survey of 24 focus groups in three districts. The study used an
exploratory research design and descriptive statistics to analyze the data. Findings reveal that
trust influences enrolment decisions in micro health insurance. Given the importance of
client trust in the insurer and the CBHI scheme, it is crucial to build an understanding of how
to best to foster individual and community trust in a CBHI scheme. One way in which trust is
fostered relates to the management team and the enabling processes found to repair, create,
and develop trust among potential CBHI members. CBHI managers have at their disposal a
number of tools that can enhance the likelihood of enrollment among targeted communities.
These tools incorporate the stakeholder feedback to determine schemes benefits package and
pre-payment amounts, the acceptable payment mechanisms and the schedule of payment
collections (Chen, Liu, Hill, Xiao & Liu, 2012).
The following hypothesis was proposed from empirical literature:
H0: Enrolment is not related to equity in health care in CBHIs in Kenya.
2.4.2 Effect of Mix of Contributions on Equity in Healthcare
In most low-income countries, a large proportion of the population pay for health services at
the point of use from its own pockets. The high level of out of pockets presents myriad
harmful effects. The poor and vulnerable are deterred from using health services or from
continuing with treatment because they cannot afford to pay. Utilization of health services for
poorer people means cutting spending on the basic need such as food, clothing, shelter and
education in order to pay for health costs (WHO, 2010a). Empirical study by Xu et al. (2007)
on protecting the poor from catastrophic health expenditure by reducing reliance on OOP
with survey data analyzed using regression analysis from 89 countries covering eighty nine
percent of the world‘s population found that each year, approximately 150 million people
suffer financial catastrophe, indicating that they are forced to spend more than 40 % of the
78
income available to them on healthcare while 100 million of those people are pushed below
the poverty line.
WHO report of 2010 postulates prepayments as the most efficient and equitable method of
raising funds for healthcare. Hypothecated, general and payroll taxes, insurance or a
combination of two have been hailed as the most progressive ways of funding healthcare for
universal coverage (Doetinchem, 2010; Durairaj & Evans, 2010). It allows people to
contribute when they are health then draw on services funded by these sources when they
need them rather than paying out of pocket for them. It also spreads contributions through
one‘s life time and enhances sharing of financial costs of ill health across the population
(WHO, 2010a). In general, the greater the proportions of prepayment in overall health
financing, the more households are protected from financial catastrophe and impoverishment.
Although Xu et al. (2007) found no difference between the protection offered by prepayment
system and tax based system.
The WHO constitution asserts the right of entire population in all countries to access all
ranges of needed health services (WHO, 1948). This fundamental right can only be
guaranteed when all segments of the population are guarded against severe financial risks
associated with OOP payments. UHC evokes equity in healthcare through guaranteed
protection against financial risk associated with ill health and is therefore closely associated
with equity in financing (Soors et al., 2010). The best approach is to develop a health
financing system which facilitates contributions before healthcare is needed (WHO, 2010a).
Countries world over continue to face the challenge of raising adequate resources for health
care. Healthcare costs continue to rise amidst the financing shortfall. The uptick is driven by
communicable diseases that are predominant in low income countries and the global upsurge
of prevalence of non-communicable diseases. The trend is aggravated by technological
advancement, infrastructure and procedures improvement and development of sophisticated
medicine (WHO, 2010a; 2013). Advancement towards UHC is therefore dependent on
raising adequate funds from a sufficient number of individuals.
79
The prepaid funds from the population should be supplemented with general government
revenue and or donor support where necessary. The WHO report of 2010 urges countries at
all levels development to adapt innovative financing mechanism that exploit novel resources
particularly from domestic sources. Different countries have taken different paths to UHC
depending on the starting point and the choices they make along the way. Similarly, various
countries are at different points on the path to UHC and at different stages of developing
responsive financing systems. For instance countries that have come closer to attaining
universal health coverage employ tax revenue; payroll and general government taxes or both
to cover the needs health needs of its population, hence guaranteeing access to all at the time
of need (WHO, 2010a; HLTIIFHS, 2009).
This has proved problematic especially for low income countries where a large population
works in the informal sector, making it hard to collect income taxes and wage-based health
insurance contributions. Majority of those who work in the informal sector have no form of
financial protection against health related costs. They have to pay for health services at the
point of use hence they risk facing financial hardship and even impoverishment when faced
with sickness (WHO, 2010a). Policy makers across the world recommend that countries
aspiring to move away from OOP payments at the point of use can adopt three
interconnected options; first is to replace OOP payments with forms of prepayments; second,
is to consolidate the funds into large pools and third is to ensure that the funds are employed
efficiently to purchase needed health services. Where the economic context and fiscal space
is constrained, voluntary schemes have proved to be valuable starting point. Besides
extending financial protection against costs of ill health to the excluded, they are instrumental
in familiarizing people on the benefits of pre-payments and risk pooling (WHO, 2010a).
In Kenya health insurance coverage is limited; only at 17.1% (MoH, 2014). This can be
attributed to large informal sector and a comparatively small and stagnant formal sector. The
informal sector in Kenya accounts for approximately 82.8% of jobs in Kenya (KNBS, 2016).
This presents practical difficulties in collecting tax and health insurance contributions
particularly from the informal sector due to lack of institutional capacity to collect taxes
(WHO, 2010a). In addition, the health financing system is highly disintegrated with OOPs
80
being the main form of fragmentation. Other forms of fragmentation include NHIF, CBHIs,
private insurers and donor funding (Chuma & Okungu, 2011). This fragmentation
jeopardizes cross-subsidization necessary in pooling system (WHO, 2010a).
Like other countries that have espoused UHC, Kenya faces continuous challenges of trading
off and balancing competing demands as it moves along stages towards its realization and
sustenance (Carrin et al., 2007). Faced by increasing demand for healthcare services and
limited fiscal space for health, Kenya began by targeting the formal sector through the
national insurer, NHIF. This strategy has resulted into a two tier system making the
conditions worse for the uncovered. The formal sector enjoys financial protection against
health cost while majority in the informal sectors are not covered. Chuma & Maina (2012)
found that each year Kenyan households spend one tenth of their household budget on health
expenses with the burden of OOP being highest among the poor. As a consequence the
incidence of catastrophic health expenditure is approximately 1.48 Kenyans each are driven
below the national poverty line as a result of catastrophic health expenditure. This analysis is
based on data from 8414 households in a national household and expenditure survey. The
incidence of catastrophic health expenditure was estimated from a sample with the health
care costs as a share of the cost. The existence of such a broad based system calls for a
targeted approach (WHO, 2010a).
CBHIs have emerged in the backdrop of poor government spending in health, political
instability and poor governance in the healthcare system. One of the distinguishing
characteristic of CBHIs is the enlisting community‘s participation that allows community‘s
involvement in setting the premium. This is quite the opposite in commercial or social health
insurance where the premium is decided by the insurer and the government respectively.
CBHIs face a challenge of mobilizing sufficient resources attributable to their small size and
the contributory capacity of the target population (Tabor, 2005; Schieber et al., 2012).
Another distinctive feature of CBHIs is the social solidarity. CBHIs employ traditional self-
help and social mobilization strategies that have been embraced by the poor in low income
(Jakab & Krishnan, 2001; Preker et al., 2002). They exploit willingness and ability to pay for
81
healthcare and try to build local risk-sharing arrangements based on solidarity. They can
therefore be employed innovatively in identifying pockets of exclusion and subsequent
subsidization of the poor and exemption of the poorest and socially excluded groups.
According to Fuchs (1996) one of the requisite conditions for universality is subsidization of
the poor. Subsidies drawn from general revenues and donor funding can channeled through
CBHIs to further enhance equity in health care. In addition cash transfers and vouchers can
be used to lessen access barriers associated with indirect health costs such as transport,
accommodation and lost work time (WHO, 2010a).
The pledge to advance towards UHC requires reprioritizing of budgetary allocations with
health in mind (WHO, 2010a; Kutzin, 2012). Given the history of volatility of aid flows, the
level of total government expenditure on healthcare is crucial for guaranteed equity in
healthcare to essential health services and financial risk protection for all particularly for the
poor and socially excluded groups. This is in view of the fact that government spending is a
stable and sustainable source of financing. There are numerous estimates on how much
financing is needed for realization of UHC. The 2014 CHATHAM House report, ―Shared
responsibilities for health, a coherent global framework for health financing‖, recommends
that all countries should spend at least US$86 on health per capita, and strive to spend 5% of
GDP on health (WHO, 2010a; Chatham House, 2014). To ensure increased level of
government spending, countries in the WHO African region have realized the pressing need
of reprioritizing their government expenditure in line with the 2001 Abuja declaration that
requires countries to allocate at least 15% of its TGE to health (OAU, 2001). The WHO
High-level Taskforce on Innovative International Financing for Health Systems (HLTIIFHS),
2009) recommends that countries should allocate at least $44 per capita for delivery of an
essential package of healthcare services. WHO (2013) report on the state of financing in the
WHO African region identifies Botswana as one the countries that have met the Abuja and
HLTIIFHS target through ring-fencing of state budget. In Kenya, the percent of government
spending on health have fluctuated from a base of 7% in 2001/02, rising to 8.6 % in 2002/03,
then falling to 4.6% in 2009/10 and then rising to 6.1% in 2012/13 (NHA, 2012/13).
82
Although discussions on innovative financing have in the past focused on donor support,
many middle income countries have successfully employed innovative financing in raising
domestic resources. By so doing they have proved that innovative financing is not a preserve
for high income countries. For example, Gabon was able to raise US$ 30 million for health in
2009 after imposing a 1.5% levy on the post-tax profits of companies that money transfers
and a 10% tax on mobile phone operators. Similarly, Pakistan government has for many
years imposed a tax on profits of pharmaceutical companies to fund healthcare expenditure
(WHO, 2010a). In the earlier Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) /
MOH missions, the government of Kenya had committed to earmark 11% of Value Added
Taxes (VAT) for the proposed National Social Health Insurance Fund (NSHIF). The
commitment was later reduced to a general commitment for financial support (sixth Deutsche
Gesellschaft für Technische Zusammenarbeit (GTZ) / MOH mission report, 2004).
Additionally, sin-taxes on products that are harmful to health have a potential of raising
additional funds. They have dual advantages of improving health by reducing consumption
while raising additional funds. In Kenya, Tobacco and Alcohol have always been targeted for
raising additional government revenue. Despite the perennial tax increases, no taxes have so
far been hypothecated for health. An increase in health budget is seen as a more pragmatic
approach of raising additional funds (WHO, 2010a). The WHO High-level Taskforce on
Innovative International Financing for Health Systems (HLTIIFHS), 2009) recommends that
countries should implement options that suit their economies and are like have political
backing (HLTIIFHS, 2009).
Many low income countries are characterized by underdeveloped pre-payments and pooling
structures. According to the Organization for Economic Co-operation and Development-
Development Assistance Committee (OECD- DAC, 2010), these countries will require
substantial financial support from external partners in the short to medium term for them to
achieve UHC. The 2014 CHATHAM House report highlights the absolute need for donor
funding in low income countries. The report estimates that the provision of essential
preventive and curative package focusing mainly on communicable diseases would require
83
approximately $86 per person. In 2012, the low-income countries spent only $21, $6 of
which was financed through external aid (WHO, 2015f)
External aid can have a key catalytic role in strengthening of prepayment and pooling
system; a financing method that would propel the countries towards UHC. Although Kenya
registered a decrease in donor funding from 35% in 2009/10 to 26% in 2012/13, the first
decline of donor funding in health financing history, the donor funds continue to play a
critical role in health financing particularly in HIV/ AIDs, Malaria and Non-communicable
diseases programmes (WHO, 2010a; Stenberg et al., 2010; MoH, 2015f).
Within the context of CBHIs, donor support is critical in subsidizing and exempting those
who are unable to pay the CBHI premiums particularly the poorest and vulnerable.
Expansive literature on the potential of CBHIs in offering social inclusion and financial
protection through community financing documents that CBHIs extend coverage to rural and
low-income segments of the population excluded from other forms of healthcare pre-
payments. However, the poorest and socially excluded groups are excluded as a result of
extreme poverty (Atim, 1998, 1999; Bennett et al., 1998; Jakab & Krishnan, 2001; Musau,
1999; Chuma & Okungu, 2011;; Okech, 2013). The Ghanaian National health program
which was upgraded from smaller CBHIs has achieved inclusion of the poor and vulnerable
from earmarked funds (Schieber et al., 2012).
To achieve equity in healthcare, the poor will need to be subsidized, while the poorest and
socially excluded will require to be exempted from paying the premiums. Donor funding
channeled through government pre-payment and pooling structures reduces fragmentation
and duplication of Official Development Assistance (ODA) and other forms of international
aid efforts. To ensure effectiveness of donor support in realization of UHC, it is imperative
that the support is given in spirit of 2005 Paris Declaration on Aid Effectiveness. The
declaration urges donor countries to honour their pledges and ensure sustained mobilization
of resources. It also requires donors to adapt a sector wide approach (SWAp) in disbursement
of funding to avoid fragmentation. This approach gives the government a leeway to fund
priority interventions including subsiding and exempting the poorest and vulnerable
84
segments of the population. For instance, the poorest category of population in Rwanda is
exempted through a national risk pool which is funded by the government, development
partners and other health insurers (WHO, 2010a).
The small size of the CBHI pool however makes many CBHIs vulnerable to failure. Indeed,
the realization of one single large risk might lead them to bankruptcy. Moreover, most
schemes are especially subject to covariant risks, because in a limited geographical area, an
individual‘s health is not independent from the health of his or her neighbours, especially
when an epidemic or a natural disaster occurs (Tabor, 2005). Several alternative strategies
exist for greater risk-pooling aiming at protecting schemes from bankruptcy and sustaining
the financial protection of insured households. Using data from national socioeconomic
survey and health and welfare survey data from Thailand Limwattananon et al., (2011) found
that general taxation is the most sustainable and progressive method of financing health
services for the poor and the informal sector equitably.
The infancy stage of CBHIs is costly making it difficult for new CBHIs to consider
reinsurance in their formative stages. Their success from the beginning is therefore
influenced by both the technical and financial assistance they receive from the government or
and donors in mitigating the preventable and the unavoidable risks. In the infancy stage,
assistance on managing avoidable risks is more desirable while support on addressing
unavoidable risks should be reflected in long term plan for social reinsurance (Fairbank,
2003).
The following hypothesis was proposed from empirical literature:
H0: Mix of contributions is not related to equity in health care in CBHIs in Kenya.
2.4.3 Effects of Risk Pooling on Equity in Healthcare
Risk management have always been a concern to individuals and wider society. This has
given rise to informal risk coping strategies which are based on the rule of balance
reciprocity and trust in the social circles. According to FinAccess report of 2009, the
informal sector relies heavily on informal risk coping strategies. The report documents that
85
25% of the respondents relies on family members, 12.7% on savings, 10% on loan and only
1% on insurance claims. While these strategies are flexible and plausible for small loss
events that occur less frequently and predictably, and affect only a few members of the
community at a time, these mechanisms break down and wipe out years of progress in an
event of covariate shocks (Maleika and Kuriakose, 2008). Industrialization and urbanization
presents an additional challenge: a gradual collapse of the informal risk sharing strategies
(Chen et al., 2012).
The cost of health care relative to individual‘s income remains an impendement to UHC
(James and Savedoff, 2010). Individuals from low-income households earn low and irregular
income and often lack access productive assets (Smit and Mpedi, 2010). In addition, the poor
are exposed to more risks, have limited access to health education and prevention
programmes and are often not aware of their social entitlements (ILO/STEP – GTZ, 2006;
ILO, 2012). This makes the notion for equity in health care and by extension UHC uncertain
particularly for the poor. Worldwide, risk pooling has been lauded as the most effectual
mechanism of protecting people from financial barrier to health care (WHO, 2010a). Risk
poling de-links utilization of health services from direct payments as a result protecting poor
households from relying on coping strategies that push households into poverty (Carrin et al.,
2005). Rashad and Sharaf (2015) examines the incidence and intensity of catastrophic health
payments and the poverty impact of OOP using data from national surveys from Egypt,
Jordan and Palestine. The concentration index revealed in 2011 more than one fifth of the
population is pushed into financial catastrophe by OOP while 3% is driven into extreme
poverty in Egypt. Catastrophic health expenditure was found to be more profound in
wealthier households in the three countries.
Health insurance schemes are supposed to reduce unforeseeable or unaffordable healthcare
costs through calculable and regularly paid premiums. Sparse monetary/financial resources,
minimal economic development, public sector restrictions and low organizational capacity
explain the reasons for unavailability of health financing systems in less developed countries
(WHO, 2010b). In particular, the low income countries remain significantly needing hence
the significance of the debate regarding CBHIs in them (Chen et al., 2012). In response to
86
restrictions on government expenditure, most developing countries introduced user fees as a
form of cost recovery. Many researchers have documents the negative effects on equity in
healthcare as result of user fees. For instance, demand for health particularly among the
poorest households. Jütting (2004) evaluates the effect of health insurance in rural Senegal
using data collected through hosuehold interviews in four villages. Descriptive statistics
reveal a decrease in out of pockets payments for members (Jütting, 2004).
Proponents of equity in healthcare encourage governments to establish alternative health
financing methods that can de-link healthcare utilization from payments at the point of use.
Direct payments including user fees are identified as the major obstacle to UHC.
Prepayments and pooling mechanisms is regarded as the most efficient and equitable
approach of expanding health coverage. Abundant evidence shows that progress towards
UHC is largely dependent on raising sufficient funds from large pool, supplemented with
general government revenues and where necessary donor funding (WHO, 2010a).
For a risk pool to remain viable, it must be of sufficient size and comprised of a broad cross
section of risks. Health insurance risk pools are large groups of individual entities (either
individuals or employers) whose medical costs are combined in order to calculate premiums
(Preker et al., 2007). The pooling of risk is fundamental to insurance since it allows the costs
of those at higher risk of high medical costs to be subsidized by those at lower risk. Large
pools of similar risks exhibit stable and measurable characteristics that enable actuaries to
estimate future costs with an acceptable degree of accuracy. This, in turn, enables actuaries to
determine premium levels that will be stable over time, relative to overall trends (WHO,
2010a).
The most viable option of financing healthcare is through general tax revenues and
contributions to health insurance. As such, risk-pooling becomes a core characteristic of such
health insurance systems necessary for enabling provision of health services according to
people‘s as opposed to their individual capability to pay for the services (Bitran & Giedion,
2003; WHO, 2010a). These are based on payments contributed prior to illness; individual
contributions are one pool and used to purchase health services for all members based on
87
their needs. In tax-funded systems, the population contributes indirectly via taxes, whereas in
social health insurance systems, households and enterprises generally pay in via contributions
based on salaries or income (WHO, 2010a). One of the perquisites of establishing a tax
funded health system is a robust tax base. In addition to institutional capacity to collect tax,
strong tax compliance is also critical for developing a robust tax base. A tax funded system
responds to the fairness in financing healthcare in that it ensures equity in contributions. The
beneficiaries pay according to their means while guaranteeing them the right to health
services according to need (Gilson et al., 2001; Fronstin, 2008).
When a CBHI scheme proposes a health insurance contribution based on average healthcare
costs of the target population, a number of households usually the healthier ones may not be
interested in signing up. They find the proposed premiums too high in view of low expected
healthcare costs. The less healthy may be interested in signing up for the opposite reason. In
a voluntary scheme, adverse selection and its impact on healthcare costs and contributions
may lead to the discontinuation of a CBHI scheme (Preker et al., 2007). In addition,
voluntary schemes tend to attract enrollees from similar socio-economic background, lack the
necessary regulatory framework relating to contribution and health insurance
reimbursements. For example, funds may be organized along professional lines, for instance
farmers versus formally employed population (Preker et al., 2007; WHO, 2010a).
The practice of risk pooling is an indicator of fairness of contribution and of equity in
healthcare to health services. A solid risk pool capable of insuring its members adequately
should also consist of a sufficient number of members. Creating a large risk pool, however,
does not necessarily translate into lower premiums. Just as a pool with more low-risk
individuals can result in lower premiums, a large pool with a disproportionate share of high-
risk individuals will have higher premiums. When healthier individuals perceive no
economic benefit to purchasing an insurance cover, the membership becomes increasingly
skewed high risk members (Preker et al., 2007). Countries and societies must choose the
extent to which individual financial contributions depend on financial means, healthcare
utilization, or other factors. Whatever system is chosen, a crucial constraint is that the
revenues received must be sufficient to provide the desired system of healthcare (World
88
Bank, 2014). The failure of social and private insurance programs to achieve coverage
beyond 20% of the population has renewed concern in smaller scale risk pooling for instance
through community financing initiatives. Schemes like the Bamako Initiative, which was
widely promoted in the early 1990s, tended to pool finances regionally (WHO, 2000).
A pertinent issue at the core of universal health coverage is inclusion of the precluded groups
particularly the poorest and vulnerable segments of the population (WHO, 2010a). To
adequate resources to finance healthcare services, some form of contribution must be levied
on the population. The contributions may exclude the poorest and vulnerable groups if they
are not based on the ability to pay. This highlights the need for health financing policies that
support risk pooling mechanisms that protect people from financial barrier that inhibit access
to health and catastrophic health expenditure especially among the poor (James & Savedoff,
2010). The inference of such risk pooling mechanism is that it enables financial risks of ill
health to be spread across the population. They engender solidarity between the risk and the
poor and the healthy and the sick. The contributions from the rich are reallocated to the poor
who pay less for more services. Similarly, the healthy cater for all or some healthcare costs
incurred by the sick. Additionally, pooling and prepayments systems ease the financial
burden associated with contribution since they enable individuals to distribute the payments
throughout their life time (WHO, 2010a).
The design of risk-pooling arrangements may be heavily influenced by various aspects the
existing health insurance providers in a given setting. CBHIs are based on informal risk
sharing mechanisms which are built and guided by the rule of balanced trust and reciprocity
where members of risk pooling groups or community expect a return on their contributions
when their time comes. The risks are decentralized in that each individual‘s risk is borne by
others through a direct risk exchange mechanisms which is bound by the strong mutual ties
(Debebe, et al., 2012). Most of the CBHI schemes are small and seem to cover relatively
homogenous populations within a single pool (Jacobs et al., 2008, p. 141).
In some situations the existing public institutions which form a natural basis for community-
financed risk pools defined by geographical demarcations. In some other situations, providers
89
are predominantly unregulated and focused on making profit where more formal system of
purchaser-provider contracts need to be put in place (Mclntyre, et al., 2006; Durairaj, et al.,
2010). Pooling does not necessarily imply a single fund; different funds with different
financial capacities may exist at a lower level with a consolidation at a higher level for risk
transfer. Consolidation at a higher level involves creation of risk equalization fund funded by
multiple pools. Pools that suffer deficits are cushioned by pools that have financial surplus.
For instance, while Rwanda maintains three distinct pooling systems at the micro-level, they
are consolidated at the national level (WHO, 2010a). In Kenya, there is minimal risk pooling
in Kenya and hence very little cross-subsidization. Apart from tax funding, other forms of
pooling include NHIF, private health insurance, CBHIs and donor funding where the funds
are channeled through the general budget support. Only 4% of all health funds are pooled
through health insurance with NHIF operating the largest risk pool in the country (Chuma &
Okungu, 2011).
The Kenyan health financing system is fragmented. OOP payments present the main form of
fragmentation in the Kenyan health system. Other forms of fragmentation exist in the form of
NHIF, CBHIs, private insurance and donor funding. The NHIF mainly covers people
working in the formal sector; private health insurance companies cover the high income
groups, while most CBHI members are small scale farmers (Carrin, et al., 2007; Muiya &
Kamau, 2013). There is an extremely narrow revenue cross-subsidization in CBHIs and other
privately owned health insurance schemes since they draw their membership from members
whose socio-economic backgrounds. The poorest segment of the population which accounts
for a considerable percentage at bottom of the economic pyramid is left out. Even though the
NHIF enjoys large number of members, its funds are not pooled together with the
contributions from the CBHIs and with tax funding (Chuma & Okungu, 2011).
This situation is seen to be changing as CBHIs customer base increases by mainly drawing
membership from the NHIF scheme. Donor funds are also very fragmented where most
projects operate independently. It is therefore common to find two different donors funding
similar health projects within the same area with little cooperation in terms of financing,
operations and service delivery (Chuma & Okungu, 2011). Funding specific health
90
programmes independently undermines efficiency and equity goals envisioned by an
effective health financing system. According to WHO (2010; 2013) SWAp provides a
coordinated and harmonized structure that aligns donor funds to countries priorities and
within the broader objective of UHC (WHO 2010). Failure to pool donor resources in Kenya
results to inefficiency in allocation of resources since resource allocation does not reflect the
region‘s needs; this promotes inequities in access to care and financial risk protection (Wang
& Pielemeier, 2012; Muiya & Kamau, 2013).
Previous attempts to establish a social national health insurance scheme that offers financial
protection to all Kenyans have not been fruitful. Provision of health services to the informal
sector remains a major challenge for UHC in Kenya (Carrin et al., 2007). As a part of the
preparation towards implementing the new financing strategy for universal coverage, WHO
(2010) proposes a financing mechanism that has an equity outlook particularly for the
precluded segments of the population. A tradeoff between economic growth and reducing
health inequities should be evaluated carefully given that increased healthcare access and
financial protection for the poor is well documented means of achieving sustainable
economic development (Carrin et al., 2007; WHO, 2010a).
Efforts to increase NHIF coverage among those working outside the formal sector have
achieved limited success. Consequently, there is very limited income cross-subsidization in
NHIF. The proposed national health insurance scheme is regarded as the main mechanism
towards universal coverage (Chuma & Okungu, 2011; Muiya and Kamau, 2013). Large co-
payments undermine the financial risk protection provided through health insurance. It is
important to design an affordable and sustainable benefit package with minimal or no
copayments (Kamau & Holst, 2008; Kimani et al., 2012). A major limitation to most of the
past and present policy developments in Kenya is the failure to involve the public in the
identification and implementation of policy interventions. The government should engage
with the public when designing policies to promote universal coverage in order to ensure that
their preferences are adequately considered (Chuma & Okungu, 2011).
The amount of political goodwill and community commitment towards redistributive
mechanisms is critical for UHC particularly in the initial stages of UHC journey. Despite the
91
distinct routes taken by different countries based on the contextual and ideological
differences the prevailing attitudes towards solidarity and self-reliance remains a common
feature propels them towards UHC (WHO, 2010a). James & Savedoff (2010) defines social
solidarity as the willingness and readiness of the rich to subsidize the poor and the health to
subsidize the sick in health risk pooling systems (James & Savedoff, 2010; WHO, 2010a).
Goudge et al (2012) takes a broader view at the definition of social solidarity. They define
social solidarity as a communal property of a social- political culture that is acquired through
the process of collective historical learning. A degree of social solidarity is therefore
necessary for achieving equity in healthcare given that any effective system of financial
protection for the entire population relies on people‘s willingness to share healthcare costs
(James & Savedoff, 2010).
Current thinking among health policy makers‘ construes CBHIs as a transitional mechanism
of attaining UHC in low –income countries. This policy link between CBHIs and UHC is
informed by historical experiences in countries such as Germany, Japan and Thailand. Ghana
and Rwanda presents more recent examples (Mladovsky & Mossialos, 2006). The
emergence, scaling up and success of these schemes provides a critical learning process on
plausible strategies that low and middle income countries can embrace in their efforts to
achieve equity in health care. The health financing initiatives has emerged as a result of
governments‘ failure in meeting healthcare needs of the poor population (Jutting, 2004).
James & Savedoff (2010) hypothesizes that individuals without a health insurance cover are
likely to be worried about possible financial losses that may arise from future illness. The
individuals are therefore likely to exhibit social solidarity. CBHIs leverage on social
solidarity, reciprocity norms and cohesion to overcome problems of small risks, exclusion,
moral hazard, fraud and cost escalation (Preker et al., 2002; 2004; ILO, 2014).
Given the emphasis on health financing for equity, variation in social solidarity has been
viewed as the main explanation of variation on health insurance coverage in societies
worldwide (Jost, 2008; van Leeuwen, 2008; James & Savedoff, 2010). Empirical study by
Goudge et al. (2012) on social solidary and willingness to share health risks in Ghana,
Tanzania and South Africa found differences in willingness of cross-subsidization across the
92
countries driven mainly by past experience with health insurance, mutual ties and degree of
disparities in healthcare. With regard to subsidizing the poor, majority respondents indicated
that such subsidies should cover half of the cost or less. Data was collected from household
surveys using questionnaires from households sampled using different sampling methods in
each country. Similar, empirical study by James & Savedoff (2010) on attitude towards
solidarity conducted in 24 countries demonstrates that 31% of Kenyans favour greater cross-
subsidization of the poor by the entire population. With regard to solidarity with the sick the
average response in Kenya was 3.1, implying existence significant support for risk sharing
with the sick. Analysis was conducted using descriptive statistics on data collected from
randomly selected households in 2002-2003. Greater solidarity was demonstrated in poorer
countries than in the richest one. Correspondingly individuals with lower educations levels,
the sick and households‘ heads expressed greater solidarity with the poor (James & Savedoff,
2010). The nature and the extent to which communities are willing to take part in
redistributive schemes act a limit of acceptable inequities (WHO, 2010a).
Diverse studies have documented the importance of social solidarity in risk pooling. A study
in Guinea-Conakry shows that members exhibited a higher degree of social solidarity.
Consequently, the schemes members endorsed the redistributive mechanisms of the CBHIs.
A qualitative study by Criel & Waelkens (2003) conducted in Guinea- Conakry on the
reasons for low subscriptions in Mutual Health Organizations found that the enrollees value
redistributive effects of insurance which surpasses social circles including households, next
of kin and villages. The conclusions were drawn from analysis data collected from 147
villages through focus group discussions in March 2000. The responses were translated
before the analysis of the transcripts. Similarly, Hsiao (2004) found that members who
express high levels of social solidarity are more likely to accept cross-subsidization study in
China found that members that express high levels of solidarity are more likely to accept
cross-subsidization (Hsiao, 2004). Mladovsky & Mossialos (2006) argue that social solidarity
may reduce problems associated with insurance such as moral hazard and adverse selection.
Benefits of social solidarity in risk pooling have also been documented by Desmet et al
(1999) in Bangladesh and Schneider (2005) in Rwanda.
93
Social reinsurance offers risk transfer to CBHIs for social reasons rather than for commercial
motives. This implies that social reinsurers are concerned about the survival and possible
thriving of CBHIs (Fairbank, 2003). The goal of reinsurance is to offload risk and reward to
the re-insurer in return for more stable operating results, but the provider's additional costs
make this impractical (Wang & Pielemeier, 2012). Reinsurance is thus attractive because it
expands the size of the risk pool (Giedion, et al., 2013). Multiple insurance companies share
risk by purchasing insurance policies from other insurers to limit the total loss the CBHI
scheme might experience should a disaster occur. By spreading risk, a CBHI scheme can take
on clients whose coverage would be too great of a burden for the single insurance company
to handle alone (Gnawali et al., 2009). Reinsurance would pool the risks of several schemes,
thus granting them greater financial stability. More often however, there is very limited
experience with and capacity to undertake reinsurance (Mladovsky & Mossialos, 2006; Soors
et al., 2010; Debebe, 2012).
A partnership with local and or central government may be established so as to adequately
finance the health service benefits from the agreed upon benefit package. The average CBHI
scheme involves partnerships with health providers, pharmaceutical suppliers, financial
institutions, NGOs, local governments, donors, and in some instances, licensed insurance
companies. Lack of high quality service provision on the part of any partner has a negative
impact on all of the others (Tabor, 2005). In practice, the implications of multiple risk pools
will depend considerably on the extent to which there is clear market segmentation between
them, and the extent to which behaviour of CBHI schemes is regulated (Giedion et al., 2013).
In Tanzania, multiple risk pools already exist, with the Community Health Fund operating in
parallel with a social security scheme, and with different Community Health Funds in
different districts. However, there is fairly clear market segmentation between these risk
pools, and limited or no competition (Bennett, 2004).
Progressive scaling up of CBHI schemes and eventual mergers leads to larger risk
pools. Real merging of small-scale groups was achieved in hospital-based schemes in
Uganda where pre-existing groups such as dairy co-operatives, rather than individual
households, constituted the basis for enrolment (McCord & Osinde, 2003). To be viable
94
however, the introduction of professional management may well require external subsidies.
Merging of CBHIs in the same may take time (Usoroh, 2012). CBHI schemes can be unified
through risk-adjustment or equalization mechanisms. Equalization mechanisms would bring
about monetary support for those CBHI plans that face more than average risks. This support
would be financed via transfers from those CBHI schemes that face lower than average risks.
Thus, CBHI schemes in relatively poor areas with high health risks would be able to set
contributions at an affordable level, in view of subsidies received via equalization
mechanisms (Mathauer & Carrin, 2010; WHO, 2010a). Mechanisms are also required to
ensure the equitable allocation of funds pooled via tax revenue. Both mechanisms for risk-
equalization between insurance schemes and for the allocation of general tax resources
ensure that the relative risk of ill-health or likely health-care needs of the population served
are taken into account (Mclntyre et al., 2006; Durairaj et al., 2010).
Government may subsidize the cost of social health insurance where tri-partite contributions
to the Social Security Scheme are made by employer, employee and government. In
Tanzania, the government matching grant was introduced when it was realized that estimated
premium costs would be too high for the average household to enroll. A quasi experimental
study targeting enrollees of Hygeia Community Health care program by Gustafsson-Wright
& Schellekens (2013) on tri-partite contributions found that matching grants reduces
premiums, therefore increasing enrolment. Besides making scheme membership more
affordable, subsidies may be used to offset risk differences between schemes or compensate
for regional income inequities. However, in practice these other rationales for government
subsidy to schemes have not been observed in developing countries where CBHI schemes
receive some external donor support. Sometimes this has supported technical assistance to
the scheme, or has covered certain operating costs and on some occasions this has been used
to bail out failing schemes (Bennett, 2004; Makaka, 2012).
Macro-level risk is the risk associated with medical expenses generally. Micro-level risk is
the risk associated with incurring losses associated with particular medical treatments or
services (Usoroh, 2012). Individuals with health insurance pool their macro-level risk, while
those who lack health insurance of any kind retain the risk of loss associated with medical
95
expenses. The only way to increase macro-level risk sharing is to increase the number of
individuals with health insurance coverage (Carrin, 2003; Mills et al., 2012).While
individuals with health insurance pool their macro-level risks, the particular scope of their
insurance contracts determines which micro level risks are pooled. By enacting mandated
benefit laws, federal and state governments regulate micro-level risk pooling by requiring
coverage for certain benefits in all contracts of health insurance (Gnawali et al., 2009; Soors
et al., 2010.
The risk of loss associated with any service or treatment not covered by a standard contract is
retained at the individual level. In order to increase micro-level risk pooling, the scope of
health insurance coverage would need to be broadened (Fronstin, 2008; Gustafsson-Wright
and Schellekens, 2013). Some schemes cover in-patient but not outpatient drugs. This
disparity can have adverse consequences for other policies such as reducing length of stay
and reducing hospital utilization for minor ailments. Using data from household expenditure
and insurance enrolment surveys from seven states in Mexico, Galarraga et al. (2010) studied
the effect of insurance on catastrophic health expenditure among 36000 insured and
uninsured households in mid-2006. Using econometric method, sensitivity analysis and bi
variate probit model found that an insurance cover that incorporates an inpatient and
medicine cover is more effective in reducing catastrophic health expenditure.
Managing demand is often more difficult for drugs than for other healthcare technologies, so
some element of copayments is probably necessary in most health systems, in conjunction
with exemptions for the poor and an active regulatory policy that reduces the use of
medicines that are not cost-effective (Smith & Witter, 2004). A case study of New Rural
Cooperative Medical Scheme, an insurance scheme for the rural poulation by Honda et al.
(2016) on strategic purchasing in China, found that China has a standard an essential drug list
to check overuse of medical treatment. Data was gathered through documents review, actors
interviews and focus group dicussions after which both deductive and inductive methods of
data analysis were used to analyse the data.
96
The optimal size of risk pools is a central design consideration. The choice will to some
extent be dependent on the purposes of the risk-pooling scheme. For example, the nature of
the healthcare package under consideration has important implications for risk pooling. If it
is confined to relatively routine care of common conditions, expenditure is predictable, and
care can be delivered at a local level, so small risk pools may be satisfactory. However,
coverage of less common, more expensive care may require pooling at a higher level to
ensure that random expenditure variation can be managed and providers can be properly
regulated (Carrin, 2003; Fronstin, 2008; Debebe et al., 2012).
According to Aryeetey et al. in 2012, small risk pools introduce important additional
managerial incentives that may adversely affect system performance in terms of both equity
and efficiency, particularly if the pools are subject to very hard budget constraints. These
arise because the importance of the unpredictable random element of expenditure grows as
the size of the risk pool contracts. Small risk pools that perceive that their expenditure will
fall below their budget may spend up to protect their budgetary position in future years. Risk
pools that perceive that their expenditure will exceed their budget may be thrown into crisis,
leading perhaps to serious unplanned rationing, as they seek to conform to the budget
(Drechsler & Jutting, 2010). Different small risk pools will be under different budgetary
pressures, and so may adopt different treatment practices. Moreover, within a risk pool,
choice of treatment may vary over the course of a year if the risk pool‘s perception of its
budgetary position changes (Dong et al., 2009). A small CBHI scheme not only implies poor
financial viability and danger of bankruptcy, it also has implications on the managerial
capacity.
Small schemes may be unable to set aside the financial resources needed to hire professional
management. Managers are often voluntary members and may lack the skills as well as the
time to improve the performance of the scheme. Given the limitations to the size of the
association due to voluntary management, assisting voluntary managers to carry out certain
administrative tasks may promote expansion of the schemes (Smith & Witter, 2004;
Drechsler & Jutting, 2010) Risk pools may adopt a variety of defensive stratagem such as
cream skimming or insuring with a third party against overspending their budget. The nature
97
and magnitude of these managerial responses will depend heavily on the nature of ownership
and governance arrangements in place. They impose implicit or explicit costs on the system
which need to be weighed against any benefits brought about by devolving responsibility to
small risk pools (Soors et al., 2010).
The following hypothesis was proposed from empirical literature:
H0 Risk pooling is not related to equity in health care in CBHIs in Kenya.
2.4.4 Effects of Strategic Purchasing on Equity in Healthcare
Purchasing is defined as the process by which pooled contributions are used to pay providers
to deliver a set of health interventions. According to WHO (2010) strategic purchasing
involves an active search for the most cost effective interventions that best serves the
healthcare needs of the meet the target populations healthcare. The healthcare financing
function of purchasing encompasses a set of decisions. First, active identification of the
targets population health needs their preferences and values by assessing their health needs
and spending patterns. Secondly, searching the best health services based on the target
population‘s needs, preferences, available resources and health sector‘s priorities. Thirdly,
searching for services providers taking into consideration the quality, efficiency and equity as
well as determining the best payment methods and contractual arrangements (WHO 2000;
2010; Munge et al., 2016). The mandate may comprise the right of the CBHIs to purchase a
set of health services at the best price from pre- selected providers (WHO 2000; 2010;
Munge et al., 2016).
In many countries, lack of geographical access to inpatient facilities remains a major barrier
to health care access. A case study of New Rural Cooperative Medical Scheme in China,
Jaminan Kesehatan Nasional in Indonesia and PhilHealth in Philippines by Honda et al.
(2016) on strategic purchasing in revealed that there are a limited number of service
providers particularly those offering inpatient services in areas that are geographically hard to
access and the ensuing costs of transportation can also be a major impediment to inpatient
care in all the three countries. Data was gathered through actors interviews and focus group
dicussions after which both deductive and inductive methods of data analysis were used to
98
analyse the data. There is a case then for considering transportation as a possible benefit so as
to help avoid or reduce catastrophic expenditure associated with ambulance transport.
Incorporating ambulatory care in the benefit package also has a financial advantage. In cases
where ambulatory care would not be fully accessible, lack of effective ambulatory treatment
may result in urgent needs for more expensive inpatient care (Carrin et al., 2005). Where
ambulance services and ambulatory care are not cost effective, a compensation policy for
members from remote areas can be used as an alternative (Honda et al., 2016).
Empirical study by Munge et al. (2016) on strategic purchasing practices of 96 CBHIs in
Kenya found strategic purchasing to be a strong indicator of access to health services and
financial risk protection. The data was collected through documents review and actors
interviews and analyzed using descriptive statistics. Purchasing of health services can be
done in three main ways; first, an integrated purchaser provider approach where the
government allocates funds to government health service providers. The funds are usually
raised from general government revenue and or form insurance contributions. The second set
up involves a separate purchasing institution which purchases health services on behalf of a
country‘s population. The purchaser provider split is usually a national health insurance
agency or a government authority. Thirdly, individuals pay service providers for health
services (WHO, 2010a). The motivation behind such separation has been the desire to
develop a market for providers, which can lead to all the putative benefits associated with
market competition.
In practice, the merits of creating markets in healthcare provision remain a subject of debate.
Service providers can operate in two types of markets; a vertical or a horizontal integrated
market. Vertical integration occurs when a provider acquires or develops competences that
allow it to reduce its dependence on other providers. An examination of contractual
relationships by Baker, Bundorf & Kessler (2014) using hospital claims from Truven
Analytics MarketScan for the nonelderly privately insured in America in the period 2001–
2007 revealed that vertical integrated enables the provider to cut cost and have a direct access
to customers. The results were derived from regression analysis models that were used to
calculate the probable changes in prices, volume or spending index associated with standard
99
deviation change of hospitals market share holding other variables constant. On the other
hand, horizontal integration involves a merger of two or more service providers that provide
similar services to different members of the population. By consolidating their customers the
providers increase their market share, have more negotiation power and gain economies of
scale. Vertical integration may result in some loss of incentives for provider efficiency. If
open access to a provider is guaranteed for all in the locality, vertical integration is an
implicit way of creating a local risk pool (Smith & Witter, 2004).
Like other countries, Kenya has employed a combination of health purchasing methods.
Health services are purchased by different organizations with Ministry of Health, which
operate 191 government hospitals, 465 health centers and 2122 dispensaries being the main
purchaser. Other purchasing organizations include NHIF, CBHIs, private health insurers and
employers. Government owned health facilities get budgetary allocations based on a
historical incremental approach while staffs are paid salaries using general government
pooled tax funds. Lack of a coordinated and harmonized policy through SWAp has resulted
to disintegration of health funds. Only a small percentage of donor funds is channeled
through the Kenyan government to support the employment of health workers in remote rural
areas. A consolidation of purchaser provider is also used to provide free services in level 2
and 3 government owned facilities through a HSSF composed of pooled general government
revenue and donor funds (Chuma and Okungu, 2011).
Benefits packages under private health insurance are premium rated and vary from basic
packages tailored for middle income groups to sophisticated packages that are mainly
designed to meet the needs of the richest populations. Benefits packages for CBHI members
mainly involve inpatient care and are often linked to specific healthcare providers, usually
private-not-for profit and public health facilities. A study of purchasing practices in 96
CBHIs in Kenya by Munge et al. (2016) revealed that the services purchased by CBHIs are
dependent on premiums contributed by each member and in most cases high cost services,
specialist services and chronic care services are not covered. These findings were derived
from descriptive statistics analysis of data collected through documents review and actors
100
interviews. In addition, Private-for-profit services are rarely provide services to CBHIs due to
high cost.
Macro- level risk pooling influences access to health services; a realization of which requires
mandated benefits laws. An increase in micro-level risk pooling results to increased access to
health services through an expanded benefit package by health insurance contracts. The
downside of an expanded benefit package through macro-level risk pooling is the potential of
increase health services prices that decreases coverage rates. As a result, there is significant
interest in eliminating or decreasing mandated benefit laws. The purchaser might negotiate a
block contract with an independent provider, which implies that the provider will give all
necessary care to pool members for a fixed sum, regardless of the volume or severity of
demands. This arrangement effectively shifts the relevant part of the risk pool from the
purchaser to the provider (Honda et al., 2016).
In strategic purchasing, provider payment mechanism is an important element. One of the
most prominent methods of provider payment in most CBHI schemes is the use of salaries
and budgets. The payment mechanisms are expected to be beneficial for cost containment but
they may also lead to rationing, as a result of the enforcement of hard budget (Bennett et al.,
1998). Fee for service is also used to induce the performance of providers, certainly in a
situation of under-provision of health services (WHO, 2010a). A case study of Pereang
District in Cambodia by Soeters & Griffiths (2003) found that fee for service was part of an
incentive system that was geared towards increasing the quantity and quality of care. To
some extent, fee-for-service method of payment resulted to reduced OOP expenditure over a
period of time. Data was collected from health services providers in Pereang District. These
controls can be costly to implement since they require skilled human resource and
infrastructure to measure and monitor the use and possible overuse of services (WHO,
2010b).
Capitation involves payment of a fixed sum per person enrolled with a provider or facility in
each in each time period regardless of the services provided. To cushion the patients against
sub-optimal care, reports are made available to the public as a measure of health quality and
101
can be sued as a basis of financial rewards. Capitation rates are developed using local costs
and average utilization of services and therefore vary from one region of the country to
another. In many plans, a risk pool is established as a percentage of the capitation payment.
Money in this risk pool is with-held from the service providers until the end of payment or in
some cases contract period. When the primary care provider signs a capitation agreement, a
list of specific services that must be provided to the patients is included in the contract
(Preker et al., 2007). A case study of the Organizing for Educational Resources and Training
(ORT) Health Plus Scheme (OHPS) by Ron (1999) in the Philippines concluded that
capitation payment service providers was financially viable for risk reduction in CBHIs since
it reduces provider generated demand for inpatient care.
The amount of capitation is determined in part by the number of services provided and varies
from health plan (Preker et al., 2007). In cases where the purchaser compensates the
remotely located primary provider for referral or ambulance services, the primary service
provider uses this additional money to pay for the referrals. Obviously, this predisposes the
primary care provider to greater financial risk if the overall cost of referrals exceeds the
capitation payment. However, the potential financial rewards are also greater if diagnostic
referrals and sub-specialty services are controlled. Alternatively, some plans pay for test and
sub-specialty referrals through fee for service arrangements based on contractually agreed
upon fee schedules. Empirical study by Munge et al. (2016) found that some CBHIs in
Kenya do not negotiate for lower prices in to avoid compromising the quality of care
provided to its members.
A clear referral system is one of the methods used in controlling cost in strategic purchasing
of health services. A referral system influences the flow of patients to the right health facility.
In absence of a clear referral system, enrolled clients suffering from minor ailments will by-
pass the nearest health facilities and go direct to high level hospitals. This clogs higher level
health facilities with primary healthcare patients (Honda et al., 2016). This is a common
occurrence where an insurance scheme offers coverage only for higher hospital levels and
omits primary health care facilities. Gatekeeping through a referral system remains a
challenge since many patients prefer to go directly to high level hospitals where referral
102
system is weak because they expect the quality of care to be superior at that level (Creese &
Bennett, 1997; Galarraga et al., 2010). Though mandatory referral can be required by the
insurance scheme, the controls have to be implemented at the hospital. This can eventually
lead to the problem of adverse selection where some schemes exclude high risk population
groups such as the elderly and patients with pre-existing conditions. A more pragmatic
approach of containing costs through gate keeping involves introduction a broad benefit
package (Preker et al., 2007). Empirical study of Bwamanda Hospital Insurance Scheme in
D. R. Congo by Criel, Van Dormael, Lefevre, Menase & Van Lerberghe (1998) found that
the scheme uses broad benefit package as a method of gate keeping. The results were derived
from cross case analysis of data that was collected from ten focus groups discussions held in
Bwamanda District in March –April, 1996.
A referral system can adopt a one way or a two way referral approach. A one way referral
policy involves referring of patients to higher level facilities from lower level facilities. On
the other hand, two way referral systems involve referral of patients to both high and high
level facilities. For instance, a patient will be referred to a higher facility for specialized care
while a referral to a lower health facility is done where less specialized care such as
rehabilitation is required. A two way referral system leads to efficient utilization of health
resources and reduces patients‘ health care costs. The mutual system requires a well-
coordinated regulatory and communication structures all levels. A case study of New Rural
Cooperative Medical Scheme in China‘s Qinghai and Henan provinces by Honda et al.
(2016) found that adoption of the two referral system in Qinghai achieved greater efficiency.
The results were derived from inductive and deductive methods of data analysis that was
carried out on data gathered through documents review, actors interviews and focus group
dicussions.
The following hypothesis was proposed from empirical literature:
H0 Strategic purchasing is not related to equity in health care in CBHIs in Kenya.
103
2.4.5 Moderating Effects of Government stewardship on Equity in Healthcare
A moderating variable is a variable that influences the strength and direction of relationship
between a predictor or independent variables and a dependent variable. In the current study
the moderating variable was government stewardship. The concept of stewardship in health
was introduced by the WHO report of 2000 (WHO, 2000). Since then, the function has
received considerable attention with many authors debating its varying technical and political
perspectives ranging from the role of government in policy formulation to implementation,
from innovation to regulation, from performance and social equity to public-private
partnerships in healthcare (Travis, Egger, Davies & Mechbal, 2002; Alvarez-Rosete,
Hawkins & Parkhurst, 2013; Mladovsky, 2014). Despite significant disciplinary and
ideological differences, a common thread of interest has emerged, ‗the ultimate responsibility
for the performance of a country‘s health system lies with government‘.
Although responsibilities for different aspects of stewardship may be delegated to
stakeholders in the health sector such as county or national governments‘ parliamentarians,
professional associations, insurance funds and other purchasing agents, some providers, and
ministries such as finance and planning, a country‘s government through its ministry of
health remains the steward of stewards. As envisaged by WHO, the government role as a
steward is particularly focused on taking responsibility for the health and well-being of the
entire population as well as guiding the entire system (Travis et al., 2002). Governments
therefore has a responsibility of ensuring continuous progress towards UHC and permanence
of the achievements (WHO, 2000; 2010, Alvarez-Rosete et al., 2013).
The term stewardship as it relates to the state has been defined in numerous ways. The WHO
report of 2000 views stewardship as the careful and accountable management of well-being
the population (WHO, 2000). Armstrong views stewardship in an equally ethical and
efficiency- oriented manner. He defines it broadly and comprehensively as ―the willingness
to be accountable for well-being of the larger organization by operating in service rather than
in control of those around us‖ (Armstrong, 1997). Alvarez-Rosete, Hawkins & Parkhurst
(2013) take a nuanced view of stewardship that delineates stewardship from the related yet
distinctive concept governance. They define stewardship as a broader overarching
104
accountability over the performance of the entire health system and eventually over the
health of the whole population. They posit that the distinguishing and conceptually useful
facet of stewardship lies in its ability to allocate ultimate responsibility for the health of the
entire population. In healthcare debates, stewardship therefore occupies a special place
because it entails oversight of all the functions and in effect it has direct and indirect effects
on their outcomes.
Governments world over have the ultimate responsibility for ensuring all segments of the
population obtain services they need without suffering financial ruin associated with their
utilization. Stewardship entails maintaining a delicate balance between competing influences
and demands. It encompasses the task of defining and maintaining a strategic vision and
direction of the health policy, exerting influence through legislation, regulation the behaviour
of players and advocacy, and collecting and using intelligence (WHO, 2000; 2010; Alvarez-
Rosete et al., 2013). In countries that receives substantial amount of ODA, stewardship will
be concerned with leadership in channeling the donor funds to national health plans that are
informed by national priorities.
Beyond the formal health structures government stewardship is often hypothesized as a
critical determinant of successful and sustainable health financing in community based
structures such as CBHIs (Preker & Carrin, 2004). Various authors have different views how
appropriately the state can play the role of stewardship in CBHIs. Bennett et al. (1998)
construes that even where a clear government policy does not exist, the schemes are still
likely to play a critical role of increased equity in healthcare. Empirical study by Criel et al.
(1999) found that Bwamanda Scheme in the Democratic Republic of Congo has managed to
generate stable revenue for purchasing of health services for its members in a context where
government stewardship and external support was practically absent. Data was collected
from ten focus group discussions in Bwamanda District in March- April 1996. Cross case
analysis was employed in data analysis. The role of the scheme in the broader health system
remains largely undefined.
105
Carrin et al. (2005) view stewardship as critical to encouraging enrolment across different
income categories. He points out that in absence government stewardship the schemes will be
associated with certain population groups. Mladovsky & Mossialos (2006) views government
stewardship as critical to the success of schemes on condition that the government adapts
CBHIs as a strategy for achieving its equity and UHC objectives. On the contrary, Pauly et al
(2006) advocates for minimal government regulation citing an increase of cream skimming
and adverse selection in present of government subsidies. This study proposes four
government mechanisms for supporting the health financing functions in CBHIs namely
stewardship in schemes design of CBHIs, monitoring CBHIs related activities, as a trainer
and as co-financier.
Majority of CBHIs in Africa were created in response to and survived regardless of a vacuum
in government stewardship (Criel & Van Dormael, 1999). Most of these schemes are owned,
designed and managed by the community that they serve (Diop et al., 2006). Although most
studies on the performance the schemes are in agreement that they increase access to
healthcare and reduce catastrophic health expenditure, numerous studies have also
documents that the schemes are far from perfect. Their penetration levels are low raising the
question of their viability in the long-term. Additionally, abundant literature indicates that the
poorest and socially excluded segments of the community are not automatically covered by
these schemes (Jakab & Krishnan, 2001; Carrin et al., 2005; Rashad & Sharaf, 2015).
Empirical study dubbed as the WHO study carried out in 1998 focusing on nonprofit health
insurance schemes for populations outside formal sector in developing countries observed
that only few schemes were able to cover the targeted population. Based out on results
derived from descriptive statistics 44 covered 24.9% of eligible population, 13 schemes
covered below 15% while 12 schemes registered coverage above 50%. These challenges are
related to the context in which they are designed such as absence of formal insurance culture,
poverty and regressive health insurance premiums which lead to low revenue levels that can
be mobilized by the CBHIs (Jakab & Krishnan, 2001; Chen et al., 2012). Given the
aforementioned challenges and pitfalls, CBHIs requires government support for them to
expand their coverage to the excluded segments. It is here that government‘s stewardship in
106
the design of CBHIs becomes critical since it is ultimately responsible for the overall health
system performance of the country health system.
Limited technical capability and managerial skills in the community is well documented in
literature (Preker & Carrin, 2004; De Allegri et al., 2009, p. 591). Tabor (2005) put forward
that during the design, particularity at the start up and early operational phases it is
imperative for government and development partners to provide technical support and
management capacity development. Technical support in form of feasibility studies,
assistance in setting premiums, determining benefit packages, provider payment methods,
providing tools and skills that are essential for designing insurance related policies and
procedures in CBHIs such dealing with and mitigating moral hazard and adverse selection. In
addition, the government through legislation can recommend the minimum number of target
population that should enlist before starting up as well as enrolment of households as
opposed to individual. Further, the government can encourage integration into a regional or
national federation to ease provision of technical and managerial support on design and
management. Such networks can also facilitate re-insurance and political advocacy (Wang &
Pielemeier, 2012). CBHI schemes perform a complementary role of extending equity in
healthcare to those not covered by formal health insurance. In line will this, it is the duty of
the government to define their place within the context of the national health financing policy
(Soors et al., 2010; WHO, 2010a).
Despite different paths taken by countries to progress towards UHC, there is point of
congruence that health is an entitlement that should be based on citizenship and or residency
rather than financial capability. Use of government general revenue and or donor funding to
cover people who cannot afford to contribute has been put forward as one of the key priority
action of financing healthcare equitably (WHO, 2010a; Oxfam International, 2013).
Subsidies increase CBHIs capacity to reach the poorest of poor, thereby decreasing health
inequities. Using two national households datasets from Thailand, Prakongsai,
Limwattananon & Tangcharoensathien (2009) studied the equity impact of universal
coverage policy in Thailand. Findings derived from measures of equity in financial
contribution, healthcare utilization and public subsidies, and in assessing the incidence of
107
catastrophic health expenditure and impoverishment revealed that low income households
contribute half of the premiums while the government subsidizes the other half by general tax
revenue through the Ministry of Public Health (Prakongsai et al., 2009). Similarly, Honda et
al., 2016 observed that the New Rural Co-operative Medical Care System in China provides
a government subsidy for smaller rural communities. These results were derived from
deductive and inductive analysis of data collected from actors interviews and focus group
dicussions.
Financial support from the government can also be in form of a re-insurance or solidarity
fund that ensures financial sustainability of schemes. Such fund can serve a dual purpose of
cautioning schemes against expenditure fluctuations in cases of local epidemics and bridging
deficits in small CBHIs (Wang & Pielemeier, 2012). Such solidarity mechanisms have to be
understood and normally agreed upon by CBHIs. Targeted exemption is another avenue for
government financing. Cambodia offers a good case in time where this option has worked at
the community level. Empirical study by Jacobs, Price & Oeun (2007) in Cambodia focusing
on exemptions of user fee and access to healthcare found that community leaders are
effective in determining who should benefit from the exemptions from the government equity
fund. Their assessment was rated as accurate to the extent that those selected were poorer
than those not selected. Data was collected from 199 pairs of patients using a pre-coded
structured questionnaire. Unstructured in- depth interviews were used to authenticate the data
collected using the questionnaires.
In addition, the government could counteract, to some extent, the regressive character of flat
contributions by households in many CBHIs. Of course the latter presupposes that the
taxation system itself is progressive, which is not necessarily guaranteed (WHO, 2010a). No
blueprints exist on how best CBHIs can be integrated in a national policy towards UHC. The
options at hand are path dependent and subject to the specificities of the national context. The
most frequent picture however seems to be that of a fragmented approach, with CBHI as one
of the strategies in a pluralistic environment where the CBHI model coexists with and
hopefully complements other financing modalities targeting specific population groups
(Carrin, 2003; Soors et al., 2010).
108
Effective implementation of health policies that are meant to promote equity in healthcare
requires not only improved design but also monitoring and evaluation. Monitoring generates
results that are useful in modifying ineffective government policies as well as reinforcing the
effective ones. In essence monitoring implementation of health policies enables that
government to move toward results-oriented policies that maximizes the potential of players
in health to achieving the policy objectives. Empirical study of CBHIs in Nepal by Deutsche
Gesellschaft fur Internationale Zusammenarbeit (GIZ) (2012) reveals that supervision and
monitoring mechanisms were non-existence in all the schemes despite having been initiated
by the government. These results were derived from analysis of data collected from a survey
of CBHIs in Nepal using descriptive statistics.
Tabor (2005) argues that it may be impractical for CBHIs to measure that actual impact
particularly on health outcomes due to the high cost of gathering health performance data
from a small group of beneficiaries. Carrin et al (2005) suggests that the government can
monitor each CBHIs basic performance, track progress across various CBHIs over time as
well carrying out comparative analysis along the health financing functions. Monitoring
enables the government to proactively stimulate establishment of CBHIs, detect problems in
existing CBHIs and offer practical solutions to the problems.
In relation to training, the government and or donors can build the capacity of CBHIs
management team through provision of basic skills in accounting, management information
systems, setting up of insurance development plans and negotiating and contracting of
providers and other third parties, preparation of organization structures, statues and
regulations as well as monitoring and evaluation (Tabor, 2005). Moreover outcomes from
monitoring should be used as a natural input into the training activities (Carrin et al., 2005).
The following hypothesis was proposed from empirical literature:
H0 Government stewardship does not have a moderating effect on equity in healthcare in
Kenya.
109
2.5 Chapter Summary
This chapter presents a diagrammatic lay out of the proposed conceptual framework
stipulating the health financing functions as the independent variables – enrolment, mix of
contributions, risk pooling, strategic purchasing and the dependent variable equity in
healthcare. The function of stewardship in healthcare is presented as the moderating variable.
The chapter also reviews the theoretical framework where the theories that provide a
structured analysis for examining and understanding innovative healthcare financing and
equity through CBHIs in Kenya are discussed. Additionally, empirical review of studies
carried out by other researchers on the subject of CBHIs and equity in healthcare within the
health financing functions and other related concepts are presented. The hypotheses that were
to be tested in the study as suggested in chapter one are also highlighted at the end of
subtopic. The next chapter discusses the methodology that was used to examine innovative
healthcare financing and equity through CBHIs in Kenya.
110
CHAPTER THREE
3.0 RESEARCH METHODOLOGY
3.1 Introduction
This chapter covers the research methodology that was used in the study to establish examine
innovative healthcare financing and equity through CBHIs in Kenya. It discusses the research
philosophy, paradigm and design especially with respect to the choice of the design. It also
discusses the population of study, sample and sampling techniques, data collection methods
as well as data analysis and data presentation methods that was employed in the study. The
detailed description of the research procedure is important so that if another researcher
follows it, he or she will be able to reach similar conclusions without difficulty.
3.2 Research Philosophy and Research Paradigm
Saunders, Lewis & Thornhill (2007) defines research philosophy as the way data of a certain
phenomenon should be collected and analyzed. Kalof, Dan & Dietz (2008), Saunders et al.
(2016) identified two main philosophical dimensions to distinguish existing research
paradigms, namely ontology and epistemology. Ontology is concerned with the view of how
one perceives reality. Ontologically, existence of reality can be viewed independent of social
factors and their interpretations, termed which can either be objectivist (Saunders et al.,
2016; Neuman 2011).
Conversely, subjectivist or nominalist believes that reality is dependent on social actors and
assumes that individuals contribute to social phenomena or subjective (Saunders et al., 2016;
Wahyuni, 2012). On the other hand epistemology thinking is concerned with the way to
generate, understand and use the knowledge that is deemed to be acceptable and valid
(Saunders et al., 2016). The present study seeks to examine innovative healthcare financing
and equity through CBHIs in Kenya within the health financing functions which an
observable phenomenon based on credible data, facts. An epistemological philosophical
research was therefore adopted in this study.
This study adopted a positivism paradigm. Taylor et al. (2007); Jonker & Pennink (2010)
defined research paradigm as a broad view of phenomena or a worldview of a set of
111
assumptions regarding how things work which serves as a thinking framework that directs
the behaviour of the researcher. The authors further argued that research paradigm can be
categorized as positivist and interpretive view. Where positivist view is objective and beliefs
that in the existence of one truth while interpretive view is based on the fact that there are
many different realities and truths due to the fact that different individuals have different
perceptions, needs and experiences (Kalof et al., 2008; Saunders et al., 2016).
Positivist holds that if different researchers were to study the same factual problem using
similar statistical methods then they will generate similar results (Saunders, Lewis &
Thornhill, 2012; 2016). This implies that there exists a universal generalisation of
phenomena. Interpretivists view contends that though generalization is possible as in the case
of positivist view, this generalisation may result from social conditioning thus understanding
phenomena need to be framed in the context of relevant law or dynamic of social structures
which creates observable phenomena within the social world (Saunders et al., 2016). In order
to gain knowledge on innovative healthcare financing and equity through CBHIs in Kenya
the researcher used positivism paradigm since it allows the researcher generate knowledge by
testing proposed hypothesis and generation of the phenomena taking into account the social
conditioning (Blaikie, 1993; Chia, 2002).
Saunders et al. (2016) identifies three research approaches, namely deductive, inductive and
abductive. Deductive approach involves development of theory and hypothesis that are
subjected to rigorous research while inductive approach is based on the principle of
developing theories from observed empirical data. In abductive research, the researcher
develops new theories or modifies existing ones through numerical and cognitive reasoning.
In the present study the researcher developed a theoretical and conceptual framework
deductively which were subsequently tested using qualitative and quantitative data.
Additionally the variables were operationalized in a way that enabled facts to be measured
quantitatively and the researcher was independent of variables that were being studied.
112
3.3 Research Design
Cooper & Schindler (2014) defined research design as a general framework that outlines how
the researcher will go about answering the research questions. The author categorizes three
types of research design namely exploratory, descriptive and causal research designs.
According to Robson (2002) exploratory research design seeks new insights, questions and
assesses phenomenon in new light. Under this design the researcher is not compelled to
following a structured process thus the findings of this design are tentative.
Descriptive research design is based on the fact that the researcher narrates how various
events regarding a certain phenomenon occur without interfering with the subjects
understudy. Descriptive studies describe characteristics associated with the subject
population and explain the variables that exist between these variables in order to provide a
picture of a particular phenomenon (Cooper & Schindler, 2014). According to Gill &
Johnson (2002) descriptive surveys are concerned primarily with addressing the particular
characteristics of a specific population of subjects, either at a fixed point in time or at varying
times for comparative purposes. Causal research design is used to establish cause and effect
links between variables. It is used to establish whether X leads to occurrence or influences
variable Y (Creswell, 2014). Creswell (2014) posits that explanatory research design aims at
finding the causal relationship or to establish relationships between variables. Under this
design, the researcher maybe interested in finding the relationship between a particular
dependent variable and one or more independent variables. Based on this research design,
predictions on certain outcomes can be made.
Based on the objectives and specified hypothesis of this study the researcher employed
descriptive and causal research designs. Descriptive research design was used to describe
certain variables of interests such as enrolment, mix of contributions, risk pooling, strategic
purchasing, government stewardship and equity in healthcare. Causal or explanatory design
was used in establishing the causal effects relationships between the independent variables
and the dependent variables. Causal research design was also used to explain the magnitude
of the relationship between equity in healthcare and the independent and variables
(enrolment, mix of pre-payment contributions, risk pooling and government stewardship).
113
3.4 Population
A population is the total collection of elements about which we may wish to make some
inferences (Cooper & Schindler, 2014). It is a group of individuals or people, or items under
consideration for statistical purposes (Collis & Hussey, 2009). Mugenda & Mugenda (2003)
define a target population as a group of individuals to which the researcher would like to
generalize his/her results from. Accordingly, a target population is defined as a universal set
of the study of all members of real or hypothetical set of people, events or objects to which
an investigator wishes to generalize the result. The population of this study was composed of
the 115 CBHIs registered in Kenya by the year 2015. The choice of the 115 CBHIs (see
appendix 3) is based on the data provided by KCBHFA, the umbrella association for CBHIs
in Kenya. The 115 CBHIs had been operation for period of between 1-15 years and were
drawn from four networks; Afya Yetu Initiative, ADS Western, ADS Nyanza and STIPPA.
3.5 Sampling Design
Sampling is a statistical method of selecting adequate units or elements from the study
population. Saunders et al (2016) postulate that a sample should be representative of the
study population for the results to be extrapolated back. The sampling design lay out the
process of drawing the study‘s sample.
3.5.1 Sampling Frame
According to Sanders et al., (2009) a sampling frame is a complete list of all the cases in the
population from which a sample is drawn from. It is a physical representation of the target
population and comprises all the units that are potential members of a sample (Kothari &
Garg, 2014). Lavrakas (2008) defines a sampling frame as a list of the target population from
which the sample is selected and that for descriptive survey designs a sampling frame usually
consists of a finite population. Cases listed in a sampling frame should enable the researcher
to answer the research questions and meet the study objectives. To this end a sampling frame
should be accurate, complete and up to date (Saunders, Lewis & Thornhill, 2009). The
sampling frame for this study was 115 CBHIs registered in Kenya under the umbrella
association of KCBHFA in 2015. Thirty three CBHIs were omitted from the study due to
lack of coherent data and intermittent periods of activity. The ensuing situation reduced the
114
population of interest to 82 CBHIs. Pilot data was collected from three CBHIs, making the
number of CBHIs eligible for data collection in the main study to be 79 CBHIs.
3.5.2 Sampling Technique
In the current research data was collected from all possible cases. A census ensures
representation and unbiased selection of elements especially in a small population (Saunders
et al., 2016). The target population was therefore 79 CBHIs registered by the umbrella body
(KCBHFA) in Kenya that had complete and coherent data.
3.5.3 Sample Size
As mentioned above, the target population for this study was 79 CBHIs that had coherent
data. From the information obtained from the CBHIs, the researcher noted that the CBHIs
had an average of four members of management team. For researcher to make inference
about the CBHIs, responses were sought from four members of each CBHIs management
team. This translated to 316 members of CBHIs management team. The study estimated a
sample based on this target population using Yamane (1967) formula. This formula is
preferred over other formulae due to its simplicity and it is scientific. This formula is
specified as shown:
Where n is the sample size, N is the population and denotes the precision error. Based on
this formula and a sample size of 316 and precision error of 1 percent (0.01) the sample size
was calculated as shown:
The sample size of this study was 306 members of management teams from 79 CBHIs
registered under KCBHFA in 2015.
3.6 Data Collection Methods
To provide both the descriptive and causal picture of innovative healthcare financing and
equity through CBHIs in Kenya within the health financing functions study collected both
primary and secondary data. A structured questionnaire used in collection of primary data on
n=316 / [1+316 (0.012)]= 306 ………………………………………………………………..3.1
115
all constructs in the study- enrolment, mix of contributions, risk pooling, strategic
purchasing, government stewardship and equity in healthcare. Additionally, a secondary data
sheet was used to collected data on three constructs- enrolment, mix of prepaid contributions
and equity in healthcare. The table below shows the type of data and data collection
instrument that was used for each study constructs.
Table 3.1 Type of data and data collection tools
No Variables Indicator variables Type of data Data collection Instruments
1 Enrolment Affordability of
contributions
Unit of membership
Timing of collection
Trust
Primary and
Secondary
Questionnaire
Secondary data sheet
2 Mix of
prepaid
contributions
Mix of prepaid
contributions
Primary and
Secondary
Questionnaire
Secondary data sheet
3 Risk Pooling Social solidarity
Mechanisms for
enhanced risk pooling
Size of pools
Primary Questionnaire
4 Strategic
Purchasing
Contracting
Provider payment
mechanism
Referrals
Waiting period
Primary Questionnaire
5 Government
Stewardship
Design
Training
Monitoring
Co-financing
Primary Questionnaire
6 Equity in
Healthcare
Increased access
Equity in Contributions
Sustainability
Quality of Care
Primary and
Secondary
Questionnaire
Secondary data sheet
116
3.6.1 Questionnaire
A questionnaire was developed by the researcher on the basis of the research questions.
According to Best & Kahn (2007) a questionnaire is used to collect factual information from
a large number of the respondents. They also allow respondents time to think about
questions (Cooper & Schindler, 2014). The questionnaire was mostly structured and the
respondents were provided with instructions to ensure that they understand the questions.
Structured questions reduce data collection time while unstructured questions encourage the
respondent to give in depth responses thereby enhancing the quality of data collected
(Cooper & Schindler, 2014).
Both closed and open ended questions were used to capture the quantitative and qualitative
data. Open-ended questions are easy to construct, permit free responses from the
respondents, stimulates respondents‘ feelings allowing them to present a clearer picture of
the subject at hand. Shortcomings of open-ended questions include; collecting responses that
are difficult to classify and capturing information that is irrelevant to the study objectives.
Open-ended questions were used to capture any information that may have been omitted.
Closed-ended questions enable the respondents to select answers among the stated
alternatives. They require minimal writing and hence they save on time and money (Saunders
et al., 2016). According to Kothari & Garg (2014), closed-ended questions are also easy to
administer, compare and analyze. Closed-ended pose a challenge of limiting the respondent
to researcher‘s choices (Saunders et al., 2016). The closed-ended questions sought responses
in a five point likert-type scale. The responses ranged from (1) Strongly disagree (2) disagree
(3) Neutral (4) Agree (5) Strongly Agree. To measure the degree of social solidarity in risk
pooling the researcher sought responses that ranged from (1) None of the cost (2) Some of
the cost (3) Half of the cost (4) Most of the cost (5) All of the cost 1 = None of the cost, 2 =
Some of the cost, 3= Half of the cost 4 = Most of the cost, 5 = All of the cost. In addition
dichotomous scales questions were used. The questionnaire tested with a few members of the
population for further improvements. This was done in order to enhance its validity and
accuracy of data collected for the study (Dillman, 2007; Sauder et al., 2016).
117
3.6.2 Secondary Data Sheet
Like other organizations, CBHIs store a lot of information. Practically, not all the
information was relevant for the current study. The researcher developed a secondary data
sheet that captured only the information that was deemed relevant for this study. For the
purpose of this study, secondary data was collected on the number of households targeted,
number of households enrolled, type of product, premiums per product and product uptake,
utilization of healthcare, longitudinal data on total premiums collected (from households,
government and donors), healthcare costs reimbursements and administration cost.
3.7 Research Procedures
The researcher developed a questionnaire on the basis of the research questions. The
questionnaire was first piloted on a part of the study population before embarking on the full-
scale research (Welman, Kruger & Mitchell, 2007:12; Kumar, 2011). A pilot study or pilot
test is a small-scale study undertaken to explore areas that need more development and
refinement. Litwin (1995:60) posits that pilot testing also predicts difficulties that may arise
during subsequent data collection, which might otherwise have gone unnoticed. Litwin
(1995), Anderson (2004) & Welman et al. (2007) indicate that a pilot study aids in detecting
possible flaws or errors in the measurement procedures and to identify unclear or
ambiguously formulated questions early enough.
This gives the researcher a chance to correct errors and to redesign problematic parts before
the data collection tool is distributed to the respondents. Anderson (2004:218) posits piloting
as process that ensures that a survey generates valuable data. Saunders et al (2016) propose
that a pilot should be carried out on a minimum number of 10 persons, who are of similar
ability and background to the target population. This is done to obtain an assessment of the
validity of the questions, as well as the likely reliability of the data that was collected.
Saunders et al. (2009) suggest that a pilot should be carried out on a minimum number of 10
persons, who are of similar ability and background to the survey target population. This is
done to obtain an assessment of the validity of the questions, as well as the likely reliability
of the data that was collected.
118
In the current study, the following three steps were undertaken: First, the questionnaire was
circulated to three experts in finance, two of which was expert‘s health financing experts in
health financing. Comments were sought on the questionnaire‘s representativeness and
appropriateness. The researcher also worked closely with her supervisors in improving
reliability and structure of the questionnaire. Their recommendations were used to make the
necessary amendments before pilot testing. Their recommendations were used to amend the
layout contents and instructions. Secondly, the questionnaire was piloted and thirdly, the pilot
test data was coded before preliminary analyses were carried out. Content validity was
achieved in two ways. First, through administering the instrument in English, the language
that was familiar and well understood by all respondents who participated in the study.
Secondly, as mentioned earlier, the pilot study was carried out prior to the main study to test
the content of data collection instrument.
3.7.1 Pilot Study
Reliability and validity of measurements are important for the interpretation and
generalization of research findings. Valid, reliable and comparable measures performance of
CBHIs within the health financing and stewardship functions are critical components of the
evidence base for health policy. In this study, a draft questionnaire was pilot tested on 10
respondents drawn from Ithaeni, Kilome and Enzai CBHIs under the network of BIDII. The
respondents of the pilot study were however not included in the study sample.
The respondents were encouraged to comment on the clarity and relevance of the questions
as well as to make suggestions on the instructions. Ambiguous and vague questions were
identified from difference in understating of individual questions by the respondents while
various insufficiencies in the questionnaire were discovered. The comments and suggestions
put forth by the respondent were also used to improve various aspects of the questionnaire.
Questions that did not pass the validity and reliability tests were dropped from the final
questionnaire. The pilot test also used in testing the adequacy of the analysis tools. The pilot
test data was later coded before conducting preliminary analysis to test for reliability using
Cronbach‘s alpha.
119
3.7.1.1 Pilot Test Results
The study used the Test-Retest/Stability Reliability which compares results from an initial
test with repeated measures later on, the assumption being that if the instrument is reliable
there will be close agreement over repeated tests if the variables being measured remain
unchanged. The Kappa score, specificity, and positive predictive values (PPVs) were also
used to measure reliability and validity, respectively. Reliability was tested using Cronbach‘s
alpha. Cronbach‘s alpha is known as a good measure of reliability. The values of Cronbach‘s
alpha ranges between 0 and 1 where the Cronbach‘s alpha values between 0.8 and 1.00
indicate a considerable reliability, values between 0.70 and 0.80 indicate an acceptable
reliability while values below 0.70 are considered less reliable and unacceptable (Nunnally,
1978). In this study, Cronbach‘s alpha coefficient which is a measure of internal consistency
was used to assess reliability. Reliability indices for the pilot study ranged from 729 to 0.941.
This suggested acceptable levels of internal consistency. This implies that the items included
in measuring different constructs were indicative of the same underlying disposition; equity.
Reliability of the constructs is shown below in table 3.2.
Table 3.2: Reliability Test of Constructs
Endogenous and Exogenous
Constructs
Reliability Cronbach’s
Alpha
Numbe
r of
Items
Commen
ts
Enrolment 0.941 12 Accepted
Pre-payment mix 0.743 5 Accepted
Risk pooling 0.838 15 Accepted
Strategic purchasing 0.729 5 Accepted
Government stewardship 0.874 20 Accepted
Equity in health care 0.922 26 Accepted
Findings indicate that enrolment in CBHIs had a coefficient of 0.941, equity in healthcare
had a coefficient of 0.922, government stewardship had 0.874, risk pooling had 0 0.838, pre-
payment mix had a coefficient of 0.743 and strategic purchasing had an alpha value of
0.729.
120
Table 3.3: Kaiser-Meyer-Olkin and Bartlett's test
Test Value
Kaiser-Meyer-Olkin Measure of Sampling Adequacy 0.654
Bartlett's Test of Sphericity Approx. Chi-Square 327.307
Df. 9
Sign. 0
In addition, this study tested for both convergent and discriminant validity. The Bartlett‘s test
for sphericity and Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was ran.
Results in table 2.3 show that KMO test had a score of 0.654, which was well above 0.50
levels, indicating acceptable degrees of sampling adequacy for the variables (Tabachnick &
Fidell, 2014; Brett, Ted & Andrys, 2010). The results also showed that the Bartlett‘s test of
Sphericity had a Chi-Square value of 327.307 with a significant value of 0.000.
3.7.2 Reliability of the Instruments
According to Mugenda & Mugenda (2003), the accuracy of the data collected largely
depends on the data collection instrument in terms of validity and reliability. A scale or test is
reliable if measurements made under constant conditions are likely to give the same results,
assuming that no changes in the basic characteristics being measured occur. Reliability refers
to the consistency of the scores obtained. Reliability of the scale for the constructs describing
the variables of the study was found to be sufficient because all the items and composite
reliability coefficients were equal to or above 0.6 which is set as the acceptable minimum
(Nunnaly, 1978; Cronbach, 1951). Reliability is a measure of how stable, dependable,
trustworthy and consistent a test is in measuring the same thing each time. In research
measurement scores normally constitute the true component and the error component
(Sekaran & Bougie, 2016).
Accordingly, the reliability is higher when the degree of error in an instrument is lower
(Kumar, 2007). The analysis of reliability is specifically important when there are several
items that measure the same concept or phenomenon (before constructing an index or scale)
so as to minimize errors of single items. Reliability may be measured in terms of stability or
121
consistency. The stability aspect of reliability refers to a comparison of the same measure for
the same sample at two or several points in time, i.e., test-retest whereas internal consistency
reflects homogeneity of the several items comprising a scale (Cooper & Schindler, 2014).
3.7.3 Validity of the Instruments
In the simplest terms, a test can be judged valid if it measures what it is intended to measure.
According to Mugenda & Mugenda (2003) validity is the accuracy and meaningfulness of
inferences based on research results. It comprises of the degree to which results obtained
from the data analysis represent the subject of the study. Content validity that refers to
whether the items measure the substance or subject matter they were intended to measure
was another important aspect addressed (Williams, 2007). It is the extent to which evidence
and theory are in congruence, thus supporting the interpretations of test scores entailed by
proposed uses of tests (Saunders et al., 2016). Construct validity therefore, seeks to ensure
that the test is actually measuring the intended attribute and not any other extraneous
attributes. Validity takes different forms including content, criterion-related and constructs
validity (Creswell, 2014).
In this study, assessment of content validity was accomplished by determining of the degree
to which data collected using a particular instrument represents a specific domain or content
of a particular concept. Mugenda & Mugenda (2003) contend that the usual procedure in
assessing the content validity of a measure is to use a professional or expert in a particular
field. To establish the validity of the research instrument, the researcher sought opinions of
management staffs and experts including the study supervisors. The questionnaire was
validated by discussing it with randomly selected management officials of the CBHIS. Their
views were evaluated and incorporated to enhance content and construct validity of the
questionnaire. Additionally, the researcher employed face validity method to evaluate the
extent to which the researcher believes that the data collection tools are appropriate. The
researcher requested experts in research to review items in the tools that were used. Invalid
items were removed before the research is conducted. Furthermore, the researcher assessed
the responses and non-responses per question to determine if there was any technical
dexterity with the questions asked. After piloting testing and refining the questionnaire, a
122
research expert with the help of two research assistant, administered the refined questionnaire
to the sampled respondents.
3.7.4 Administration of the Instruments
After piloting testing and refining the questionnaire, a research expert with the help of two
research assistant, administered the refined questionnaire to the sampled respondents. A total
of 318 questionnaires were distributed to the respondents. The questionnaires were
accompanied by two letters; an introductory letter from United States International
University detailing the purpose of the study and a letter from the researcher addressed to the
respondents explaining the scope and purpose of the study in addition to introducing the
research assistants. To minimize non-sampling errors emanating from data collection
procedures the research assistants were trained on propriety and correct research procedures.
Kothari & Garg (2014) points out that non-sampling error stem from incorrect data collection
procedures such as ineffective interviewers, bias from interviewers and non-response errors.
Knowledge gained by the researcher during pilot testing enabled the researcher to train the
research assistants. Further, the researcher coordinated the entire data collection process.
Data was collected between 23rd
March, 2016 and 14th
April, 2016. In total, 224 complete
and usable questionnaires were received from the respondents. In addition 79 secondary data
sheets were received from each CBHI. Prior to data analysis, the data was cleaned, grouped
on the constructs before responses of each question were coded. The data was then entered
and analyzed in the software package SPSS Version 20.
3.7.5 Ethical Issues
According to Cooper & Schindler (2014) ethics is standard behaviour that guides our
behaviour and relationship with others. In research, ethics refers to the suitability of a
researcher‘s behaviour in relation to the study subject and those affected by the study. Ethics
therefore touches on every aspect of any research from formulating the research topic,
research questions, literature review, research methodology, analysis and interpretation of the
findings. In effect, the entire research process should be meticulously and morally defensible
to those affected by it (Saunders et al, 2016).
123
Before commencing on the field data collection exercise, the researcher sought approval
through a letter of recognition from USIU, and subsequently obtained a research permit from
NACOSTI. The data collection instrument was developed in such a way that the study
procedures do not cause any harm or emotional distress to the respondents. Due to sensitivity
of some information to be collected, the researcher holds a moral obligation to treat the
information with utmost propriety. The research was based on voluntary participation;
participants were reassured of confidentiality and were not under duress in any way to
answer any questions they feel uncomfortable about.
Participants were fully informed about the procedures involved in the research and their
consent was sought before commencing. The research assistants were required to explain to
the respondent the scope and purpose of the study, and further assured them of
confidentiality. There was no infringement of the respondents and his/her human rights were
also observed. The consent of the respondents was required before the questionnaire was
distributed to them, and they were assured of confidentiality of the information under the
study. Honesty and integrity was maintained throughout the study period. Another factor that
was taken into consideration was that the research process entailed a number of steps which
included the research questions, literature review, research design, data collection procedure
and data analysis and interpretation of the findings. Items in the instruments for data
collection were clear, simple and did not have leading to answers.
3.8 Data Analysis Methods
Ordinarily, the amount of data collected in a study is rather extensive and research questions
and hypotheses cannot be answered by a simple perusal of numeric information. According
to Chadran (2004) the data should be processed and analyzed in an orderly and coherent
manner. Quantitative information is usually analyzed through statistical procedures.
Statistical analyses cover a broad range of techniques including decretive and inferential
statistics. Before processing the responses, the completed questionnaires were checked for
completeness and consistence upon which any additional information needed for clarity was
sought from the respondents. The data was then coded in SPSS ready for analysis. The raw
primary data collected was coded prior to being input into SPSS statistical analysis software.
124
Once coded, the data was then cleaned to ensure accuracy and completeness of the
information obtained. In analyzing the data collected, both descriptive and inferential
statistics was utilized. The quantitative data that was obtained from the questionnaires was
coded and keyed into statistical package of social science (SPSS) analysis software. The
current study employed various data analysis methods to answer the research questions and
meet the objectives of the study. In this sub-section the data analysis method that was used to
analyze the secondary and primary data is discussed.
Analysis of Secondary Data
The quantitative data was first coded before applying descriptive statistics in grouping the
responses into various categories including percentages and frequency distribution to indicate
the variable values and the number of occurrences in terms of frequencies and averages.
Patterns and trends were also derived from the data.
3.8.2 Analysis of Primary Data
The primary data was analyzed using both descriptive and inferential analysis.
Descriptive analysis
The data was analyzed using descriptive techniques descriptive statistical tools (SPSS
Version 20 and MS. Excel 2010) were employed. Descriptive statistics enables a researcher
to compare the values numerically (Saunders et al., 2016). A likert scale was used to rate the
extent of agreement on the given constructs of enrollment strategy for the CBHIs in a scale of
1 to 5 where 1 is the least extent whereas 5 is the maximum indicating the level of
agreement. The results from the collected responses were analyzed based on means and their
standard deviations to show the variability of the individual responses from the overall mean
of the responses per each aspect of enrollment strategy. The mean results are therefore given
on a scale interval where a mean value of up to 1 is an indication of a strong extent of
disagreement; 1.1 – 2.0 is disagree; 2.1 – 3.0 is a moderate extent of agreement (neither agree
nor disagree), 3.1 – 4.0 is agree and a mean value of 4.2 and above is an indication of a
strong extent of agreement. The responses were also analysed in terms of frequencies.
125
3.8.2.2 Correlation Analysis
The relationship between independent variable and the constructs under dependent variable
was established using Karl Pearson‘s coefficient of correlation. Correlation analysis is a
statistical measure of strength of linear relationship between paired variables. Commonly
referred to as coefficient correlation, it is one of the methods used measure correlation
between latent variables. In the current study Pearson correlation coefficient (r) was used to
establish the degree of relationship between sub-constructs of the independent variable -
enrolment (affordability of contributions, unit of membership, timing of collections and trust)
and independent variable construct; mix of contributions, risk pooling and strategic
purchasing and the sub-constructs under the dependent variable - equity in health care
(healthcare access, equity in contributions, quality of care and sustainability). Pearson
correlation coefficient was deemed suitable for this study since the data collected for
independent variables and the constructs under the dependent variable was numerical.
Additionally, one of advantage of SEM PLS is its ability to calculate a predicted score Ŷ (for
each sub-construct or construct) from predictors of a sub-construct or construct. Ŷ is a
composite or a weighted linear combination of the predictors. The predicted score were used
in Pearson correlation coefficient to establish degree of relationship between the dependent
variables sub-constructs or construct and the sub-constructs from dependent variable. The
coefficient denoted by ‘r’ ranges from -1 to +1. Pearson correlation coefficients between -1
and +1 signify a weaker positive and negative relationship while a value of 0 implies a
weaker positive and negative relationship while a value of 0 implies independent relationship
between variables in question (Creswell, 2014 ; Saunders et al., 2016).
Correlation coefficient with a value of +1 means a perfect positive correlation. This shows
that the two variables have a precise relationship where an increase in one variable results to
an increase of the other variable. On the other hand, a value of -1 represents a negative
correlation, signifying a precise relationship between variables, where an increase in one
variable results to decrease of the other variable (Saunders et al., 2016).
126
4.8.2.3 Multicollinearity Test
Multicollinearity is the undesirable situation where the correlations among the independent
variables are strong. In other words, multicollinearity misleadingly bloats the standard errors.
Thus, it makes some variables statistically insignificant while they should be else significant
(Martz, 2013). Tolerance of a respective independent variable is calculated from 1 - R2. A
tolerance with a value close to 1 means there is little multicollinearity, whereas a value close
to 0 suggests that multicollinearity may be a threat (Belsley, Kuh & Welsch, 2004). The
reciprocal of the tolerance is known as Variance Inflation Factor (VIF). Equally, the VIF
measures multicollinearity in the model in such a way that if no two independent variables
are correlated, then all the VIF values will be 1, that is, there is no multicollinearity among
factors. But if VIF value for one of the variables is around or greater than 5, then there is
multicollinearity associated with that variable (Martz, 2013).
3.8.2.4 Structural Equation Modeling
Thirdly, multivariate data analysis was conducted using structural equation modelling (SEM)
partial least square (PLS) approach. In this research, SmartPLS(R)
software, a SEM PLS
software Version 3 was employed to develop the measurement and structural model under
study, test hypothesized relationships between variables and bootstrap (Ringle, Sarstedt,
Schlittgen & Taylor, 2013). SEM uses variates in both the measurement and structural
models, referred to as outer and inner models. Each indicators of a construct acts collectively
to define the construct in a measurement models whereas in the structural model the variates
are related to one another in both interdependence and correlational relationships (Hair et al.,
2010).
SEM belongs to a family of multivariate techniques such as multiple regression analysis,
MANOVA and interdependence techniques such as factor analysis. Unlike other dependence
techniques, SEM has the capability of testing for relationships simultaneously. Additionally,
SEM uses multiple measures for each construct an aspect that allows the estimation
procedure to directly correct for the measurement error. This principal difference between
SEM and interdependence techniques is that it allows the relationships between constructs to
be estimated more accurately in SEM (Hair, Black, Babin & Anderson, 2010; Babin &
127
Svensson, 2012). PLS maximizes the variance explained for all endogenous variables by
working with a block of variables as opposed to latent variables in model estimation. The
variances are generated through a series of ordinary least squares regression. PLS based
equation modeling assumes that the sample distribution reasonably represents the distribution
of the population of interest. Since this method does not assume normality it uses boot
strapping to obtain standard errors during hypothesis testing. Bootstrapping is a non-
parametric statistical technique that is used to drawn statistical inference by estimating the
properties of an estimator. The strength of bootstrapping lies in its ability to draw
conclusions about the attributes of the study population strictly from the sample at hand
instead of relying on unrealistic assumptions about the population being studied (Hair, Ringle
& Sarstedt, 2011).
SEM was appropriate for this study since the researcher had multiple exogenous constructs
which were represented by several measurable variables. Additionally, it permitted automatic
correction of measurement errors among constructs as the simultaneous estimation of
measurement and structural models were being executed (Hair et al., 2010). The
development and analysis of the model was carried out in a six stages decision process as
follows;
Stage 1 Defining individual constructs
Stage 2 Developing and specifying the overall measurement model
Stage 3 Designing a study to produce empirical results
Stage 4 Assessing the measurement model validity
Stage 5 Specifying the structural model
Stage 6 Assessing the structural validity
Stage 1: Defining individual constructs
Good theoretical and structural grounding is critical for establishing causality relationships in
SEM. This is especially important in cross-sectional studies (Hair et al., 2010). A pretest on a
section of the target population was carried out to determine the suitability of the items as
well as reliability and validity of the constructs.
128
Stage 2: Developing the overall measurement model
In this stage, each latent constructs to be used in the model were identified and individual
variable were assigned to the latent constructs. This process was represented in a path
diagram for ease of labeling notation for indicators, constructs and relationships between
them. The error item for each indicator was also specified at this stage. Partial Least Square
Path Modeling (PLS) was used to develop the measurement model. PLS belong to the family
of multivariate data analysis techniques that enables the researcher to simultaneously assess
the measurement of latent constructs and test the hypothesized relationships among the
constructs within the same analysis. The technique achieves this by performing iterative set
of factor analysis and ordinary least squares regressions until the difference in the average r2
of the constructs become non-significant (Gefen, Rigdon & Straub, 2011). Advantages of
PLS technique include its ability to can accommodate both reflective and formative
measurement models as well as its capability of holding many constructs and indicators
without leading to estimation problems (Henseler et al. 2009, p.279-281). According to
Fricker, Kreisler, and Tan (2012) reflective indicators are presupposed to be affected by same
underlying concept; the latent construct. Therefore, a change in the underlying latent
construct manifests in all its indicators. As a result, the reflective indicators should be
correlated. On the contrary, the formative indicators are not assumed to cause variation in
latent constructs and thus they are not presumed to be correlated. The current study used
partial least squares estimation to examine the causal relationship among latent variables
(Hair, Anderson, Tatham & Black, 1998).
Stage 3: Designing a study to produce empirical results
Having specified the basic model, the researcher focused on research design and estimation.
With regard to design the researcher collected ordinal data. The data was analyzed through
correlations due to the required interpretive and statistical issues. Unlike other multivariate
methods of data analysis, SEM is more sensitive to the sample size. In deed some of the
alogarithms applied in SEM are unreliable in small samples. As in any other statistical
method, sample size forms the basis of estimation of the sampling error. Generally, larger
samples generate stable and more likely replicable solutions (Hair et al., 2010). Kline (2011)
advocates for a sample size of 200 or larger. Kenny (2012) posits that the SEM model
129
performs optimally in sample sizes ranging from 200-400 since the main test of model fit is
sample size dependent. The sample size for the study was 318 members of CBHIs
management team. The sample size was thus adequate enough to allow the model to run and
was representative of the population of interest. Bootstrapping determines the distribution of
the empirical sample by resampling form a sample with replacement (Kenny, 2012). In the
current study bootstrapping was used to draw 500 samples with replacement from the
original sample 318.
Stage 4: Assessing the measurement model validity
Having specified the measurement model, collected sufficient data and made critical and
having determined the estimation technique, the next step assessing the measurement model
validity. Hair et al. (2010) suggest four criteria of assessing model fit in PLS, namely;
construct unidimensionality, and construct reliability, convergent validity and discriminant
validity. Construct unidimensionality is used to confirm that the indicators used to measure a
particular latent construct only measures that specific construct. Further the authors put
forward that exploratory factor analysis and or CFA can be used to evaluate this criterion.
In consecutive steps, SmartPLS was used to measure the construct, composite and
convergent reliability and discriminant validity. Construct reliability assess consistency of
the scales in measuring a particular latent construct measures. Construct reliability was
assessed by computing the composite reliability and the Cronbach alpha of the constructs.
The Cronbach alphas were all above the 0.6 threshold as specified for PLS analysis (Hair et
al., 2006). The values ranged from 0.74 and 0.99 which indicated good reliability. Composite
reliability measures were evaluated by using SmartPLS (Hensler et al., 2009). Convergent
validity measures that ability of indicators relevant latent constructs to actually measure a
particular construct. The current research used CFA assess convergent analysis at a statistical
significant level of above 0.5 (Nunnally, 1978). Additionally, average variance extracted
(AVE) was used to measure convergent validity. A 0.5 threshold was adapted indicating that
the latent constructs should account for at least fifty percent of the variance in the items
(Hair, Black, Babin, Anderson & Tatham, 2006). Only items significance levels of each test
were retained for further analysis. Discriminant validity measures the inter-constructs
130
covariances. Forenell Lacker Measure was used to assess the discriminant validity of the
outer model. The Fornell Larker measure compares the AVE to the highest squared
correlation of each construct (Fornell & Bookstein, 1982; Hair et al., 2010). Items that did
not discriminate themselves well were deleted as shown in appendix v.
Stage 5: Specifying the structural model
After development and evaluation of the measurement model as well as validation of the
study measures, the next step involves the specification of the structural to test the
hypothesized relationships between the constructs based on the proposed theoretical model
(Fricker et al., 2012). This task was carried out in two steps. First, the path diagrams
indicating the structural relationships among the latent constructs were constructed. In the
second step involved inclusion of the measurement specifications. The two steps approach
was preferred since it ensures accurate representation of reliability of the indicators, hence
avoiding the interaction of the two models in the integrative model (Hensler et al., 2009).
Stage 6: Assessing the structural validity
In this final stage, the validity of the theoretical measurement model was measured against
the sample data collected. This was achieved by evaluating the path coefficients, t-values,
overall model fit and significance levels for the structural paths to determine the causal
relationships among research constructs as hypothesized in the integrative model.
Bootstrapping was used to measure the strength and direction of the hypothesized
relationship. Initially, the significance testing of the independent variables was conducted
without the mediator. The mediating variable (government stewardship) was then included in
the model and the resultant t- values were generated. The second assessment of model fit
allowed an evaluation of the integrative model fit and individual parameters estimates for the
structural path in the structural regression model. The statistical objective of PLS is to show
high r2 and significant t-values, thus rejecting the null hypothesis of no effect. Parameters
with an absolute t-value greater than 1.65 indicate a significance level of 0.1 (i.e. p<0.1), 1.96
indicate a significance level of 0.05 (i.e. p<0.05), those with an absolute t-value over 2.58
present a significance level of 0.01 (i.e. p <0.01), and those with an absolute t-value over
131
3.26 present a significance level of 0.001 (i.e. p<0.001) (Hair et al., 2010; Fricker, Kreisler &
Tan, 2012). Model measure fit criterion is presented in Table 3.4.
Table 3.4: Measures to Fit PLS Model
Measures Procedure Statistical Criterion
Construct
Unidimensionality
Confirmatory Factor Analysis Factor Loading > .70
Construct Reliability Reliability Analysis Cronbach Alpha >0.6
Convergent Validity Factor Analysis
Composite Reliability
Variance
Factor Loadings > .50
Composite Reliability > .70
Average Variance Extracted > .50
Fornell-Larker Measure AVE> (Highest correlation for factor)2
Discriminant Validity Coefficient of Determination R2 > .19 (weak)
R2 > .33 (moderate)
R2 > .57 (substantial)
Sources: Hair et al., 2006; Hair et al., 2010; Vinzi, Trinchera & Amato, 2010; Fricker et al.,
2012).
3.8.2.5 Path Model representation for Equity in Healthcare
The overall model shown in figure 3.1 was constructed to test the hypothesized relationships.
The latent variable are depicted by the elliptic shapes (e.g., ) while the indicators
measuring the latent constructs are represented inside rectangular frames (e.g. ).
Hypothesized directional relationships of one variable on another are depicted with a line
with a single arrow (e.g., →). For example, the indicators scale items for enrolment include
AFF1, TM3, and TRU2 among others.
132
3.8.1.3.7 Path Model representation for Equity in Healthcare
Figure 3.1 Path model for Equity in Healthcare
133
Table 3.5 Items of measure for study constructs
Exogenous Variable Variable Codes Wording
Indicators
1. Enrolment Affordability AFF 1 We give members a chance to allocate premium among preferred products
AFF 2 Members can pay premium in kind e.g. through farm produce or labour
AFF 3 We give subsidies and exemption of premiums for extremely poor and
vulnerable
AFF4 We encourage members to use savings-linked premium payment mechanisms
such as rotating saving groups (Chamas), MPESA
AFF5 Members can make irregular payments of premium
Unit of UM1 The CBHIs membership is based on coffee, tea, villages or mutual benefit
Membership societies or administrative areas
UM2 The CBHIs have adapted households as unit of membership
UM3 CBHIs encourages members to join when they are healthy
UM4 CBHIs membership is open to poor and vulnerable groups
Timing of TM1 Members pay in a single annual premium /contribution
Collections TM2 Members can pay their premiums in installments
TM3 Premium payments correspond with income from e.g. harvest, sale of
livestock or salary payment
TM4 Premium payments are linked to loans from SACCOs and banks
TM5 Mobile premiums payments are allowed
Trust TRU1 Members interact with the scheme‘s administrative / management team about
their needs, concerns and make suggestions for improvements
TRU2 Members participate in setting of benefit package
TRU3 Members participate in setting the premiums
TRU4 The CBHIs uses existing Chamas, community development projects and
credit schemes as entry points for CBHIs membership
TRU5 Members of the scheme are willing to cover the poorest and vulnerable groups
in the community such as orphans and disabled
134
Table 3.5 Items of measure for study constructs (Continued)
Exogenous Variable Variable Codes Wording
Indicators
Pre-payments Mix of MC1 Members‘ contributions are adequate in meeting the cost of the set benefit
Contributions package
MC2 NHIF covers costs of services not covered by the CBHIs
MC3 The CBHIs receives financial support from donor(s)
MC4 The poor and vulnerable members of the CBHIs are covered through
government and or donor subsidies
Risk Pooling Enhanced ERP1 Members of CBHIs comes from a wide range of social economic background
Risk Pooling ERP2 CBHIs targets a large geographical /administrative area
and Social ERP3 Community members are encouraged to join CBHIs when they are healthy
Solidarity ERP4 We have a waiting period before one can benefit from CBHI
ERP5 CBHIs has other branches in other geographical or administrative areas
ERP6 CBHIs reinsures its risks
ERP7 CBHIs has a partnered with the county / national government and or NHIF
ERP8 CBHIs merged with other CBHIs to form a network or a federation
ERP9 Members of the CBHIs have expressed the opinion that if they would not need
healthcare themselves, at least they had done something good for the
community by contributing to the insurance fund
ERP9 Members of the CBHIs are willing to contribute to pay for health care services
used by the sick
ERP10 Members of the CBHIs are willing to contribute to pay for health care services
used by the poor
Strategic Strategic SP1 We have signed a contract with all our health services providers
Purchasing Purchasing SP2 We only select providers who are accredited by NHIF
SP3 The providers must provide health services according to conditions put
forward by the CBHIs
SP4 We allocate resources based on population needs
SP5 CBHIs has been successful in negotiating agreeable terms and contract with
\service providers – in terms of service quality, fee, and reduction in
unnecessary services/prescription (moral hazard)
135
Table 3.5 Items of measure for study constructs (Continued)
Exogenous Variable Variable Codes Wording
Indicators
Government Design The Government recommends
Stewardship AD1 Startup of CBHIs when a minimum percentage of population in enrolled
AD2 A waiting period
AD3 Enrollment of households as opposed to individuals
AD4 A flexible premium collection system
AD5 A benefit package that reflect the needs of the target population
AD6 A standard treatment protocols for members of CBHIs
AD7 A standard referral procedure
AD8 Consolidation of CBHIs through a federation or a network
AD9 Creation of a risk equalization fund or a reinsurance mechanism
AD10 Community participation in management and decision making
Monitoring The government
MO1 Tracks the progress of CBHIs through time
MO2 Monitors the basic performance of CBHIs
Training The government organizes trainings on
TR1 Determination of benefit packages
TR2 Determination of contributions
TR3 Collection of contributions
TR4 Claims processing
TR5 Use of Management Information Systems
TR6 Establishment of Health Insurance Development Plans
TR7 Exchange visits
Co- The government and/ or donors
Financing COF1 Partially or fully subsidizes the poorest and vulnerable members of the
community
COF2 Has set a solidarity fund for financing epidemics and deficits of the CBHIs
136
Table 3.5 Items of measure for study constructs (Continued)
Endogenous Variable Codes Wording
Variable Indicators
Equity in Healthcare ACC1 There is distribution of enrolment across income categories
Healthcare Access ACC2 The contracted providers are within the proximity of covered population
ACC3 We cater for transport / accommodation cost related to healthcare utilization
ACC4 The covered population is entitled to similar benefits
ACC5 The number of members seeking health services has increased in the past 12
months
Equity in AMC1 Everyone pays the same amount
Contributions AMC2 Everyone pays an equal amount of their income
AMC3 We allow flexible premium payments
AMC4 We offer allow members to match premium or products to their income
AMC5 We offer premium subsidies
AMC6 We allocate a larger claim budget for low cost products
Quality of QO1 The CBHIs has a standard client compliant management mechanism
Care QO2 Members have complained about long queues before being seen
QO3 Members have complained on availability of health services
QO4 Members have complained about lack of key prescribed medicines
QO5 Members have raised concerns relate to cleanliness
QO6 Members have raised concerns on availability of trained staff in the contracted
health facilities
QO7 The CBHIs have put in place mechanisms to check on patient perceived
quality of care in contracted health facilities on issues concerning waiting
time, availability of staff, services, drugs and supplies
QO8 There are other organization(s) that conduct quality checks in the contracted
health facilities
QO9 These organizations share their findings with the CBHIs
137
Table 3.5 Items of measure for study constructs (Continued)
Endogenous Variable Codes Wording
Variable Indicators
Sustainability The administrative committee has basic skills in
i) Administrative FS1 Setting of contributions
and Managerial FS2 Collection of contributions and compliance
Capability FS3 Determination of the benefit package
FS4 Claim management
FS5 Marketing and communication
FS6 Contracting with providers
FS7 Use of management information systems
FS8 Accounting
ii) Financial FS9 CBHIs have partnered with organizations that assist in collection of premiums
Sustainability FS10 Premiums are not paid on time
FS11 The CBHIs is funded through a mix of contributions from county / national
government /donors and members contributions
FS12 Government and or donors‘ covers health cost for those who cannot afford to
pay premiums
FS13 Chronic conditions are covered by the CBHIs
FS14 The CBHI have put in place mechanisms to check whether the invoices sent
from the health facilities are correct
FS15 There are instances which health facilities tried to overstate the reimbursement
request amount
FS16 CBHIs is part of a network of CBHIs
FS17 We have merged with other CBHIs
FS18 We are in partnership with NHIF
FS19 Besides treatment we finance community prevention, promotion and
rehabilitation activities
138
3.8.3 Hypotheses Testing
The table below provides the study hypothesis and how they were tested.
Table 3.6: Summary of Hypotheses Testing
Hypothesis Analysis Accept/Reject Criteria
H1 H1: Enrolment is related to
equity in healthcare in CBHIs.
Partial Least Squares Analysis
Path coefficient and T values
Degree of Correlation is
Positive or Negative
Accept hypothesis when level of significance, indicated
by T values t values > 1.65 - 0.1 Sig. level
> 1.96 – 0.05
> 2.5 – 0.001
(two tailed)
H2 H1: Mix of contributions is
related to equity in healthcare
in CBHIs
Partial Least Squares Analysis
Path coefficient and T values
Degree of Correlation is
Positive or Negative
Accept hypothesis when level of significance, indicated
by T values t values > 1.65 - 0.1 Sig. level
> 1.96 – 0.05
> 2.5 – 0.001
(two tailed)
H3 H1: Risk pooling is related to
equity in healthcare in CBHIs
Partial Least Squares Analysis
Path coefficient and T values
Degree of Correlation is
Positive or Negative
Accept hypothesis when level of significance, indicated
by T values t values > 1.65 - 0.1 Sig. level
> 1.96 – 0.05
> 2.5 – 0.001
(two tailed)
H4 H1: Strategic purchasing is
related to equity in healthcare
in CBHIs
Partial Least Squares Analysis
Path coefficient and T values
Degree of Correlation is
Positive or Negative
Accept hypothesis when level of significance, indicated
by T values t values > 1.65 - 0.1 Sig. level
> 1.96 – 0.05
> 2.5 – 0.001
(two tailed)
H5 H1: Government stewardship is
related to equity in healthcare
in CBHIs
Partial Least Squares Analysis
Path coefficient and T values
Degree of Correlation is
Positive or Negative
Accept hypothesis when level of significance, indicated
by T values t values > 1.65 - 0.1 Sig. level
> 1.96 – 0.05
> 2.5 – 0.001
(two tailed)
139
3.9 Chapter Summary
This chapter gave a detailed description of the methodology that was used to examine
innovative healthcare financing and equity through CBHIs in Kenya. The chapter starts with
a discussion of various research philosophies and paradigm before underlining the one that
was employed in the current research. The chapter then discusses diverse research designs
before highlighting the research design that was used to empirically examine the
hypothesized relationships between the constructs based on the proposed theoretical model.
The chapter then describes the population and the sampling method that was used to derive
the sample. The chapter also focused on the research procedures that were employed in the
current study from development, piloting and the refinement of the research instruments. In
addition, administration of the research instrument in the main survey, ethical considerations
and hypothesis testing are discussed. The next chapter reports on results and findings of the
data collected and analysis based on the methodology.
140
CHAPTER FOUR
4.0 FINDINGS
4.1 Introduction
This chapter presents an analysis of the data that was collected through the use of structured
questionnaires, interpretation and discussion of the findings. The mode of presentation
employed for the results depends on the appropriateness of the used mode providing ease of
interpretation and understanding of the results to any interested person. Both descriptive and
inferential analysis methods have been employed in the analysis. The results are presented
according to the research objectives and the chapter is organized according to the themes
derived from the research questions.
4.2 General Information
Under the general information, the study examined the response rate, the duration within
which the CBHIs, the targeted number of households to the CBHIs, the number of
households covered by the CBHIs, average enrolment, average increase in healthcare
utilization (hospital and pharmacy visits) and the average mix of contributions in CBHIs. The
results under this section are based on the frequencies and percentages giving the patterns of
the characteristics of the CBHIs studied.
4.2.1.1 Response Rate for the primary data
Figure 4.1 Response Rate
141
The research targeted to collect secondary data from 79 CBHIs. The study achieved a 100%
response rate. For the primary data, a sample of 306 members of 79 CBHIs management
teams was targeted. The study however achieved a response rate of 71% of the targeted
responses.
4.2.1.2 Responses based on Years of Operation
0%
10%
20%
30%
40%
50%
60%
70%
80%
1-5 years 6-10 years 11- 15 years
70.10%
22.10%
7.80%
Figure 4.2 Responses based on Years of Operation
Figure 4.2 shows that respondents came from CBHIs that were in existence for 1-5 years
(70.1% of CBHIs), 6-10 years (22.1% of CBHIs) and 11-15 years (7.8%).
142
4.2.2 Descriptive Statistics – Secondary Data
4.2.2.1 Targeted Households
20, 8.9%
103, 46.0%
86, 38.4%
15, 6.7%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
45.0%
50.0%
0-500 501-1000 1001-2000 More than 2000
Figure 4.3 Households targeted by CBHIs
According to the findings, majority of CBHIs targeted between 501 – 1000 households, 38%
targeted between 1001 – 2000 and 9% targeted between 0 – 500 households whereas 7%
targeted more than 2000 households in their operations.
4.2.2.2 Households Covered by the CBHIs and Average Enrolment
0
50
100
150
200
250
0-500 501-1000 1001-2000 AverageEnrolment
205, 91.5%
5, 2.2%14, 6.3%
169
0-500
501-1000
1001-2000
Average enrolment
Figure 4.4 Households Covered by CBHIs and Average Enrolment
143
With regard to the number of households covered by the CBHIs, majority (91.5%) of these
CBHIs covered up to 500 households, 6% covered between 1001 – 2000 households whereas
2% covered between 501 – 1000 households in their operations.
4.2.2.3 Number of Households Targeted and Those Covered by CBHIs
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
100.0%
0-500 501-1000 1001-2000 More than 2000
8.9%
46.0%
38.4%
6.7%
91.5%
2.2%6.3%
0.0%
Targeted
Covered
Figure 4.5 Relationships between the Number of Households Targeted and Those
Covered by CBHIs
Findings show that majority CBHIs targeting 501-1000, 1001-2000 and more than 2000
households have not been able to enroll the targeted number of households.
144
4.2.2.4 Benefits Package, Premiums and Products Uptake in CBHIs
7336
288654
206
1561
482
1454
120 0 00
1000
2000
3000
4000
5000
6000
7000
8000
Figure 4.6 Benefits Package, Premiums and Products Uptake in CBHIs
The study established that CBHIs offer diverse products with different benefits package.
Afya Yetu Initiative offers six products; two CBHIs only cover and four CBHIs and NHIF
cover. CBHIs only cover offers two types of products; small households cover (Kes 700) and
an expanded households cover (Kes 968.5). The cover offers security for services accessed in
public hospital services. Through its partnership with NHIF, Afya Yetu Initiative offers a
basic NHIF cover (Kes 2300) and an expanded benefits NHIF cover through CBHIs (Kes
2570). Additionally, members of the CBHIs purchase a NHIF that offers security outside
CBHIs. A basic cover for CBHIs outside CBHIs cost Kes 380 and expanded NHIF cover cost
Kes 650. The composite cover allows members to access services in private not-for-profits
hospitals. ADS western, ADS Nyanza and STIPPA offers four types of covers; a general
outpatient cover (Kes 1200), a general inpatient and outpatient cover (Kes 2000), a general
145
inpatient and outpatient cover with ambulance services (Kes 2400) and a general inpatient
and outpatient cover with ambulance services and funeral expenses (Kes 2700).
Majority of households have purchased a small household cover (7336). The uptake for the
composite products is 1561 and 482 households at a cost of Kes 2300 and Kes 2570
respectively. 654 and 206 members of CBHIs have bought NHIF separately for product 9
and 10 respectively. Non-renewal rates are high among NHIF Product with a total of 1680
households failing to renew their covers in 2015. In total 12,101 households are covered
through these products.
4.2.2.5 Methods of Payment: Inpatient and Outpatient services
Figure 4.7 Methods of Payment: Inpatient and Outpatient services
Majority of CBHIs (94.2%) use fee for service method to pay services providers for both
inpatient and outpatient services. Six percent CBHIs employ mixed methods when
purchasing health services from contracted providers.
146
4.2.2.6 Distance to the Nearest Contracted Service Provider
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
Within 5kilometers
5-10 kilometers More than 10kilometers
62.9%
27.7%
9.4%
Figure 4.8 Distance to the Nearest Contracted Service Provider
According to the findings, majority of the CBHIs (62.9%) members live within a proximity
of 5 kilometers to the nearest contracted service provider. Twenty eight percent of the
respondents reported that their members live 5-10 kilometers while 9% reported that that
their members live more than 10 kilometers from a contracted service providers.
4.2.2.7 Average Mix of Contributions in CBHIs
52669
10932
0
10000
20000
30000
40000
50000
60000
AVERAGE CBHIs CONTRIBUTION AVERAGE NHIF CONTRIBUTION
Figure 4.9 Average mix of contributions in CBHIs
147
Findings in figure 4.9 show that the mean NHIF contribution through CBHIs was 52,668.831
while the mean CBHIs contribution was 109,315.08.
4.2.2.8Trends of Total Premiums collected, healthcare cost reimbursements,
administration cost and deficit/surplus in CBHIs between 2010-2015
Figure 4.10 Trends of Total Premiums collected, healthcare cost reimbursements,
administration cost and deficit/surplus in CBHIs between 2010-2015
Findings on trends of total premiums collected, healthcare cost reimbursements,
administration cost and deficit/surplus in CBHIs between year 2010-2015 shows a sharp
decrease in total premiums collected and surplus between year 2010 and 2013 followed by a
slight increase in the next two years. The healthcare cost decrease gradually between year
2010 and 2014 followed by a slight increase in year 2015. However, the administration cost
remained constant throughout the entire period.
148
4.2.3 Descriptive Analysis - Extent Effect of Equity in Healthcare - Healthcare Access
Table 4.1 Descriptive Analysis - Extent Effect of Equity in Healthcare- Healthcare
Access
Healthcare Access
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std. Dev
Membership is
distributed across
income categories
0
0
2
27
71
4.70
.498
The contracted
providers are within
the proximity of
covered population
0
0
1
27
72
4.71
.472
We cater for
transport and or
accommodation cost
related to healthcare
utilization
49
45 6 0 0 2.58 .624
The covered
population is
entitled to similar
benefits
The number of
members seeking
services has
increased in the past
12 months
0
0
0
0
1
0
28
25
71
75
4.70
4.74
.478
.449
Findings in Table 4.1 shows that membership is distributed across income categories in the
CBHIs (mean = 4.70; std. dev. = 0. 498), the contracted providers of the healthcare services
are within the proximity of covered population in the CBHIs (mean = 4.71; std. dev. =
0.472), the covered population is entitled to similar benefits (mean = 4.70; std. dev. = 0.478).
Similarly, number of members seeking services has increased in the past 12 months (mean =
4.74; std. dev. = 0.449). Findings also indicated that some CBHIs cater for transport and or
accommodation cost related to healthcare utilization (mean = 2.58; std. dev. = 0.624).
149
Table 4.2 Descriptive Analysis - Extent Effect of Equity in Healthcare - Equity in
Contributions
Equity in Contributions
SD
(%)
D
(%)
N
(%)
A
(%)
SA
(%) Mean Std. Dev
Everyone pays equal premium 2 51 41 6 0 2.52 .656
Everyone pays an equal amount of
their income 2 51 41 6 0 2.52 .656
Flexible payments are allowed 2 51 41 6 0 2.52 .656
We offer allow members to match
premium to their income 0 0 1 29 70 4.68 .494
We offer premium subsidies 2 51 41 6 0 2.52 .656
We allocate a larger claim budget for
low cost products 0 0 1 35 64 4.63 .511
With regard to equity in contribution, some CBHIs members pay the same amount (mean =
2.52; std. dev. = 0.656), some members pay an equal amount of their income (mean = 2.52;
std. dev. = 0.656), some CBHIs have a flexible payment system (mean = 2.52; std. dev. =
0.656) and some CBHIs subsidize of premium (mean = 2.52; std. dev. = 0.656). It is also
clear that CBHIs allow members to match products with their income (mean = 4.68; std. dev.
= 0.494), allocate a larger claim budget for low cost products (mean = 4.63; std. dev. =
0.511).
150
Table 4.3 Descriptive Analysis – Extent Effect of Equity in Healthcare - Quality of Care
Quality of Care
SD
(%)
D
(%)
N
(%)
A
(%)
SA
(%) M
Std.
Dev
CBHIs has a standard client compliant
management mechanism 0 0 1 34 64 4.63 .511
Members have complained about long queues
before being seen 2 51 41 6 0 2.52 .656
Members have complained on availability of
health services 2 51 41 6 0 2.52 .656
Members have complained about lack of key
prescribed medicines 2 51 41 6 0 2.52 .656
Members have raised concerns relate to
cleanliness 2 51 41 6 0 2.52 .656
Members have raised concerns on availability
of trained staff in the contracted health
facilities
2 51 41 6 0 2.52 .656
Members have raised concerns on referral
system 2 51 41 6 0 2.52 .656
CBHIs have put in place mechanisms to check
on patient perceived quality of care in
contracted health facilities on issues
concerning waiting time, availability of staff,
services, drugs and supplies
0 0 1 32 67 4.65 .505
There are other organization(s) that conduct
quality checks in the contracted health
facilities
0 0 1 28 71 4.70 .489
These organizations share their findings with
the CBHIs 0 0 1 34 64 4.63 .511
Table 4.3 presents the study findings on the quality of the care services provided in the
CBHIs. According to the findings, the CBHIs have put in place a standard complaint
management mechanism as the respondents reported where the mean value indicated that the
respondents strongly agreed with a mean of 4.63 and a standard deviation of 0.511. However,
the respondents neither agreed nor disagreed indicating that in some CBHIs, members have
complained of long queues before being seen. This had a mean of 2.52 and a standard
deviation of 0.656 indicating that the responses given had mixed reactions where some of the
respondents agreed and some disagreed whereas a significant others had a neutral response.
151
From the table also, the respondents indicated that some members to the CBHIs have
complained on lack of healthcare services (mean = 2.52; std. dev. = 0.656), lack of key
prescribed medicines (mean = 2.52; std. dev. = 0.656), members have raised concerns
relating to cleanliness contracted health facilities (mean = 2.52; std. dev. = 0.656), members
have raised concerns on availability of trained staff in the contracted health facilities (mean =
2.52, std dev. = 0.656), members have raised concerns on referral system (mean = 2.52, std
dev. = 0. 656). The respondents however agreed strongly indicating that the CBHIs have put
in place mechanisms to check on patient perceived quality of care in contracted health
facilities on issues concerning waiting time availability of staff, services, drugs and supplies
(Mean = 4.65, std. dev. = 0.505). There are other organization(s) that conduct quality checks
in the contracted health facilities. This is according to the responses given where the
respondents strongly agreed to this aspect with a mean of 4.70 and a standard deviation of
0.489. Further, findings illustrate that, the organizations share their findings with the CBHIs
as indicated by a mean of 4.63 and a standard deviation of 0.511 showing that the
respondents agreed to this aspect.
Table 4.4 Descriptive Analysis - Extent Effect of Equity in Healthcare - Sustainability
(Administrative and Managerial Capability)
Administrative and
Managerial Capability
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Setting of contributions 0 0 1 32 67 4.66 .503
Collection of
contributions and
compliance
0 0 1 34 65 4.63 .510
Determination of the
benefit package 0 0 1 29 70 4.68 .494
Claim management 0 13 24 63 0 3.48 .745
Marketing and
communication 0 17 23 60 0 3.41 .798
Contracting providers 0 0 40 57 0 4.54 .559
MIS 3 54 38 5 0 2.45 .661
Accounting 0 0 25 25 50 4.25 .835
152
Table 4.4 shows that the administration committee has basic skills in setting contributions
(mean = 4.66; std. dev. = 0.503) , collection of contributions (mean = 4.63; std. dev. = 0.510).
The study also established the existence of other skills in CBHIs including collection of
contributions (mean = 4.63; std. dev. = 0.510), determination of benefits package (mean =
4.68; std. dev. = 0.494), basic skills in Claim management (mean = 3.48; std. dev. = 0.745),
basic skills in Marketing and communication (mean = 3.41; std. dev. = 0.798), basic skills in
contracting health services providers (mean = 4.54; std. dev. = 0.559), basic skills in
recruiting and retaining core staff (mean = 4.25; std. dev. = 0.835). Some CBHIs‘
management team however lacked basic skills in Accounting as indicated by the mean
response of 2.45 and a standard deviation of 0.661.
153
Table 4.5 Descriptive Analysis - Extent Effect of Equity in Healthcare – Sustainability (Financial Sustainability)
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std. Deviation
We have partnered with organizations that
assist in collection of premiums 0 0 1 29 70 4.68 .494
Premiums are not paid on time 0 0 3 40 57 4.54 .559
The CBHIs is funded through a mix of
contributions from county / national
government / donors and members
contributions
2 51 41 6 0 2.52 .656
Government and or donors‘ covers health
cost for those who cannot afford to pay
premiums
2 51 41 6 0 2.52 .656
Chronic conditions are covered by the
CBHIs 0 0 2 26 72 4.70 .496
The CBHI have put in place mechanisms to
check whether the invoices sent from the
health facilities are correct
0 0 17 24 59 4.42 .765
There are instances which health facilities
tried to overstate the reimbursement
request amount
2 47 46 4 0 2.55 .633
The CBHIs is part of a network of CBHIs 0 0 2 31 67 4.65 .514
We have merged with other CBHIs 13 53 25 4 6 2.38 .958
We are in partnership with NHIF 0 0 1 26 73 4.72 .468
Besides treatment we finance community
prevention, promotion and rehabilitation
activities
0 0 2 26 72 4.70 .505
154
According to the findings in Table 4.5 CBHIs worked in partnership with organizations that
assist them in collection of premiums (mean = 4.68; std. dev. = 0.494), CBHIs members do
not pay premiums on time (mean = 4.54; std. dev. = 0.559), CBHIs covers members with
chronic conditions (mean = 4.70; std. dev. = 0.496), CBHIs have put in place mechanisms to
check whether the invoices sent from the health facilities are correct (mean =4.42; std. dev. =
0.765), CBHIs studied are part of a network (mean = 4.65; std. dev. = 0.514), CBHIs were in
partnership with NHIF (mean = 4.72; std. dev. = 0.468), CBHIs finances preventive, promote
and rehabilitative health services (mean = 4.70; std. dev. = 0.505).
Findings also indicate that some CBHIs were funded through a mix of contributions from the
County or National government or donors and members contributions (mean = 2.52; std. dev.
= 0.656), government or donors covers health cost of those who cannot afford to premiums
through some CBHIs (mean = 2.52; std. dev. = 0.656), contracted health facilities has in
some instances tried to overstate the claims reimbursements (mean =2.55; std. dev. = 0.633),
not all the CBHIs had merged with other CBHIs as the mean response suggested (2.38).
155
Table 4.6 Cronbach’s Alpha Coefficients, AVE and KMO values for Equity in
Healthcare (Healthcare Access, Equity in Contributions, Quality of Care and
Sustainability)
Equity in
Health
Care
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Healthcare
access
0.833 ACC1
.551
0.786
.725
67.19% None
ACC2 .707
.848
ACC3 .693
.846
ACC4 .712
.852
QOC 0.961 QOC1 .956 0.722 .981 92.79%
QOC8 .930
.969
QOC9 .868
.939
AMC 0.953 AMC1 .947 0.812 .976 88.52%
AMC2 .945
.973
AMC3 .905
.952
AMC6 .767
.858
S 0.909 FS1 .848 0.784 .917 73.84%
FS2 .785
.862
FS5 .740
.838
FS8 .702
.795
FS10 .792 .880
The results presented in Table 4.6 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. These findings imply that the items of measure were measuring what they were initially
set out to measure, and therefore the data was maintained for further analysis.
4.3 Effect of Enrolment on Equity in Healthcare
This sub-section presents the results of enrolment based on the primary data. Firstly, the
descriptive statistics are discussed. Secondly, Pearson‘s coefficient correlations between the
indicators of enrolment and the indicators of equity in healthcare are presented. Thirdly, the
Cronbach‘s Alpha Coefficients, AVE and KMO values for enrolment and the test for
hypothesized relationship between enrolment and equity in healthcare are presented.
156
4.3.1 Descriptive Analysis - Extent Effect of Enrolment Indicators on Indicators of
Equity in Healthcare
Table 4.7 Extent Effect of Affordability on Equity in Healthcare
Affordability of
Premiums
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
We give members a
chance to allocate
premium among
preferred products
0 0 1 34 64 4.63 .511
Members can pay
premium in kind-
through farm produce or
labour
4 41 55 0 0 2.52 .568
We give subsidies and
exemption of premiums
for extremely poor and
vulnerable
0.4 7 43 34 16 1.99 .847
We encourage members
to use savings-linked
premium payment
mechanisms such as
rotating saving groups
(Chamas), mobile
money transfer
0 0 1 34 64 4.63 .511
Members can make
irregular payments of
premium
1 8 35 52 4 2.50 .752
Findings on premiums affordability in CBHIs indicate that members are given a chance to
allocate premium among preferred products. This is as indicated by a mean of 4.63 for
agreed and a standard deviation of 0.511. Further, CBHIs permit members to use savings-
linked premium payment mechanisms as indicated by a mean of 4.63 for agree and a
standard deviation of 0.511. Findings as well indicate that not all the CBHIs allowed
members can pay premium in kind or in work (Mean = 2.52; Std. Dev. = 0.568), allow
members to make irregular installments payments (Mean = 2.50; Std. Dev. = 0.752).
Findings further illustrated that CBHIs do not give subsidies and exemption of premiums for
extremely poor and vulnerable (Mean = 1.99; Std. Dev. = 0.784).
157
Table 4.8 Extent Effect of Unit of Membership on Equity in Healthcare
Unit of Membership
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
The CBHIs
membership is based
on coffee, tea,
villages or mutual
benefit societies or
administrative areas
0 0 0.4 24 76 4.75 .444
The CBHIs have
adapted households
as unit of membership
0 0 0.4 24 76 4.75 .444
CBHIs encourages
members to join when
they are healthy
0 0 1.3 35 64 4.63 .511
CBHIs membership is
open to poor and
vulnerable groups
0 0 25 25 50 4.26 .830
Findings as presented in Table 4.8 shows that, Coffee, tea, villages, cooperatives or mutual
benefit societies are basis of CBHIs membership. This is as indicated by a mean of 4.75
which is in the interval of 4.1 – 5.0. CBHIs have also adopted households as unit of
membership (mean =4.75; std. dev =0.444). Similarly, there evidence that CBHIs encourages
members to join when they are healthy (mean =4.63; std. dev =0.511). CBHIs membership
is open to poor and vulnerable groups (mean =4.26; std. dev =0.830).
158
Table 4.9 Extent Effect of Timing of Collections on Equity in Healthcare
Timing of collections
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Members pay in a
single annual
premium
/contribution
0 0 0 29 70 4.70 .470
Members can pay
their premiums in
installments
4 54 38 4 0 2.45 .661
Premium payments
correspond with
income from e.g.
harvest, sale of
livestock or salary
payment
0 0 0 28 72 4.72 .451
Premium payments
are linked to loans
from SACCOs and
banks
4 54 38 4 0 2.45 .661
With regard to the timing of collection, the table illustrate that the members of the CBHIs
pay in a single annual premium or contribution. This is according to the mean response
obtained (4.70) which is in the interval of 3.0 – 3.9 for agreement and a standard deviation of
0.470. Premium payments in CBHIs correspond with cash inflows from harvest or livestock
sale or salary payment. This is according to the mean obtained of 4.72 and a standard
deviation of 0.451.
Further, some CBHIs allow members to pay their premiums in installments (Mean =2.45;
Std. Dev. = .661). Similarly, some of the CBHIs had their premium payments linked to loans
from SACCOs and banks (Mean =2.45).
159
Table 4.10 Extent Effect of Trust on Equity in Healthcare
Trust: Members or
Membership
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Interact with scheme‘s
management team and
service providers
0 26 74 4.74 .439 .439 .439
Participate in setting
of benefit package 1 34 64 4.63 .511 .511 .511
Participate in setting
the premiums 0 23 77 4.77 .420 .420 .420
Is based on existing
Chamas, development
projects and credit
schemes
2 34 64 4.62 .522 .522 .522
Are willing to cover
the poorest and
vulnerable
0 16 84 4.83 .384 .384 .384
Findings on trust in CBHIs indicate existence of strong social networks in CBHIs. Members
interact with the scheme‘s administrative/ management team about their needs, concerns and
suggestions for improvements. This is as indicated by a mean of 4.74 for agreed and a
standard deviation of 0.439. Members participate in setting of benefit package (Mean = 4.63;
std. dev. = 0.511). Further, Members of the CBHIs participate in setting the premiums as
indicated by a mean of 4.77 and a standard deviation of 0.420. From the table also, the study
findings show that the CBHIs uses existing Chamas, community development projects and
credit schemes as entry points for CBHIs membership. This is evidenced by a mean of 4.62
and a standard deviation of 0.522. It is also evident that the members are willing to cover the
poorest and vulnerable in the community such as orphans and disabled as the findings
illustrate with a mean of 4.83 and a standard deviation of 0.384.
160
4.3.2 Correlation between Enrollment Indicators and Equity in HealthCare Indicators
for CBHIs
This section presents the correlations results between enrolment indicators and indicators of
equity in healthcare in CBHIs.
4.3.2.1 Correlations between Affordability and Indicators Equity in HealthCare
Table 4.11 Correlation between Affordability and Healthcare Access
Healthcare Access
Affordability Pearson Correlation .726**
Sig. (2-tailed) .000
N 224
The findings in table 4.11 show the output of the Pearson correlation coefficient indicates a
statistically significant strong positive relationship between affordability and healthcare
access. From the table, it is clear that the significance level is 0.000 (p = 0.000), therefore,
there is a statistically significant difference in the affordability and healthcare access.
Table 4.12 Correlation between Affordability and Equity in Contributions
Equity in Contributions
Affordability Pearson Correlation .933**
Sig. (2-tailed) .000
N 224
The findings in table 4.12 show the output of the Pearson correlation coefficient demonstrate
a statistically significant strong positive relationship between affordability and equity in
contributions. From the table, it is clear that the significance level is 0.000 (p = .000),
therefore, there is a statistically significant difference in the affordability and equity in
contributions.
Table 4.13 Correlation between Affordability and Quality of Care
Quality of Care
Affordability Pearson Correlation .936**
Sig. (2-tailed) .000
N 224
The findings in table 4.13 show the output of the Pearson correlation coefficient demonstrate
a statistically significant strong positive relationship between affordability and quality of
161
care. From the table, it is clear that the significance level is 0.000 (p = .000), therefore, there
is a statistically significant difference in the affordability and quality of care.
Table 4.14 Correlation between Affordability and Sustainability
Sustainability
Affordability Pearson Correlation .878**
Sig. (2-tailed) .000
N 224
Table 4.14 shows that the Pearson correlation coefficient between affordability and
sustainability was 0.878 with a p value of 0.000, suggesting that Affordability has a positive
and significant influence on sustainability at 5%.
4.3.2.2 Correlations between Timing of Collections and Indicators of Equity in
HealthCare
Table 4.15 Correlation between Timing of Collections and Healthcare Access
Healthcare Access
Timing of Collections Pearson Correlation .749**
Sig. (2-tailed) .000
N 224
Table 4.15 presents the Pearson correlation coefficient results on the relationship between
timing of collections and healthcare access. Timing of collections and healthcare access has a
statistically significant strong relationship with a significance value of 0.000.
Table 4.16 Correlation between Timing of Collections and Equity in Contributions
Equity in Contributions
Timing of Collections Pearson Correlation .677**
Sig. (2-tailed) .000
N 224
162
The results shown in table 4.16 indicate that the Pearson correlation coefficient between the
timing of collections and equity in contributions was 0.677 with a p-value of 0.000. This
indicates a statistically significant strong (r = .677, p < .01) (Table 4.8).
Table 4.17 Correlation between Timing of Collections and Quality of Care
Quality of Care
Timing of Collections Pearson Correlation .749**
Sig. (2-tailed) .000
N 224
Table 4.17 presents the correlation results between timing of collections and quality of care.
Pearson correlation coefficient demonstrates a statistically significant strong positive
relationship between timing of collections and quality of care with a significance level of
0.000.
Table 4.18 Correlation between Timings of Collections and Sustainability
Sustainability
Timing of Collections Pearson Correlation .664**
Sig. (2-tailed) .000
N 224
The findings in table 4.18 show the output of the Pearson correlation coefficient show a
statistically significant strong positive relationship between timing of collections and
sustainability. The findings are illustrated by a significance value of 0.000.
4.3.2.3 Correlations between Trust and Indicators of Equity in HealthCare
Table 4.19 Correlation between Trust and Healthcare Access
Healthcare Access
Trust Pearson Correlation .771**
Sig. (2-tailed) .000
N 224
163
Pearson correlation coefficient between the trust and healthcare access shows there is
statistically significant difference (r = .771, p < .01) (Table 4.19). Therefore, trust in CBHIs
has a statistically significant difference effect on healthcare access.
Table 4.20 Correlation between Trust and Equity in Contributions
Equity in Contributions
Trust Pearson Correlation .839**
Sig. (2-tailed) .000
N 224
Table 4.20 presents the Pearson correlation coefficient results on the relationship between
trust and equity in contributions. The findings are illustrated by a significance value of 0.000.
Trust and equity in contributions has a statistically significant strong relationship.
Table 4.21 Correlation between Trust and Quality of Care
Quality of Care
Trust Pearson Correlation .872**
Sig. (2-tailed) .000
N 224
The findings in table 4.21 show the output of the Pearson correlation coefficient shows a
statistically significant strong positive relationship between trust and quality of care. From
the findings, it is clear that the significance level is 0.000 (p = .000), therefore, there is a
statistically significant difference in the trust and quality of care in CBHIs.
Table 4.22 Correlation between Trust and Sustainability
Sustainability
Trust Pearson Correlation .797**
Sig. (2-tailed) .000
N 224
The Pearson correlation coefficient in Table 4.22 shows insignificant relationships between
the trust and sustainability (r =.797, p<.01). Thus, trust in CBHIs has a statistically
significant influence on their sustainability.
164
4.3.3 SEM analysis on Enrolment and Equity in Healthcare
4.3.3.1 Cronbach’s Alpha Coefficients, AVE and KMO values for Enrolment
Table 4.23 Cronbach’s Alpha Coefficients, AVE and KMO values for Enrolment
2nd
order
constr
uct
First order
constructs
Cronbac
h’s
alpha
Item
Item
total
correlati
on
KM
O
PCA
compon
ent
loading
varian
ce
extract
ed
Items
deleted
En
rolm
ent
Affordabilit
y 0.964 AF1 0.931 0.763 0.983 96.57%
AF2,
AF3,
AF4
AF4 0.931
0.983
Membership 0.977 MT1 0.956 0.500 0.989 97.78% MT2 ,
MT3
MT4 0.956
0.989
Timing of
collections
0.939 TM1 0.886 0.500 0.971 94.29% TM1,
TM2 ,
TM4 TM3 0.886
0.971
Trust 0.934 TRU
1 0.770 0.766 0.864 71.45% None
TRU
2 0.850
0.907
TRU
3 0.764
0.859
TRU
4 0.825
0.892
TRU
5 0.561 0.685
The results presented in Table 4.23 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. The factors with low standardized regression weights were subsequently deleted. These
findings imply that the items of measure were measuring what they were initially set out to
measure, and therefore the data was maintained for further analysis.
165
4.3.2.2 Hypothesized effect of enrolment in CBHIs on equity in healthcare
Figure 4.11 Path coefficients for effect of enrolment in CBHIs on equity in health care
Figure 4.12 t-values for effect of enrolment in CBHIs on equity in healthcare
166
Table 4.24 Path Coefficients (Mean, STDEV, t-value)
Original
Sample (O)
Sample
Mean (M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Enrolment ->
Equity 0.908386 0.910345 0.017342 52.381868 0.0000
H0: Enrolment in CBHIs is not related to equity in healthcare.
H1: Enrolment in CBHIs is related to equity in health care.
Enrolment in CBHIs had a positive statistically and significant effect on equity in healthcare
at the 0.05 level of significance (β=0.908, t-value=52.382 >1.96, p<0.05) as indicated in
figure 4.11, figure 4.12 and table 4.24. The null hypothesis is therefore rejected and the
alternative Hypothesis H1 that stated that Enrolment in CBHIs is related to equity in health
care is supported. Results thus reveal that, when enrolment increases by 1 unit, Equity in
health care increases by 0.908 units. Figure 4.11 shows that enrolment had a coefficient r2
mean of 0.825 showing the proportion of variation in dependent variable explained by the
SEM model. r2 indicates that 82.5% of the variations in equity in health care can be
accounted for by enrolment in CBHIs.
4.4 Effect of Mix of Contributions on Equity in Healthcare
This section presents the results of mix of contributions based on the primary data. Firstly,
the descriptive statistics are discussed. Secondly, Pearson‘s coefficient correlations between
mix of contributions and the indicators of equity in healthcare constructs are presented.
Thirdly, the Cronbach‘s Alpha Coefficients, AVE and KMO values for mix of contributions
and the test for hypothesized relationship between mix of contributions and equity in
healthcare are presented.
167
4.4.1 Descriptive Analysis - Extent Effect of Mix of Contributions in CBHIs on Equity
in Healthcare
Table 4.25 Descriptive Analysis - Extent Effect of Mix of Contributions in CBHIs on
Equity in Healthcare
Mix of Contributions
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Members‘ contributions
are adequate in meeting
the cost of the set benefit
package
4 54 38 4 0 2.45 .661
NHIF covers costs of
services not covered by
the CBHIs
0 0 3 30 67 4.64 .533
The CBHIs receives
financial support from
donor(s)
0 36 58 5 0 2.69 .592
The poor and vulnerable
members of the CBHIs
are covered through
government subsidies
0 38 56 5 0 2.67 .597
Table 4.25 gives the study findings on the mix of contributions. From the table, NHIF covers
costs of services not covered by the CBHIs for the members as indicated by the mean of 4.64
and a standard deviation of 0.533. However, some of the CBHIs receives financial support
from donor(s) whereas others had no such connections and donor support (mean = 2.69; std.
dev. = 0.592), Members‘ contributions are not adequate in meeting the cost of the set benefit
package (mean = 2.45; std. dev. = 0.661) and the poor and vulnerable members of the CBHIs
are covered through government subsidies (mean = 2.69; std. dev. = 0.597).
168
4.4.2 Correlation between Mix of Contributions in CBHIs and Equity in Healthcare
Indicators
This section presents the correlations results on the mix of contributions and indicators of
equity in healthcare.
4.4.1.1 Correlation between Mix of Contributions and Healthcare Access
Table 4.26 Correlation between Mix of Contributions and Healthcare Access
Healthcare Access
Mix of Contributions Pearson Correlation -.035
Sig. (2-tailed) .599
N 224
According to the findings in Table 4.26, Pearson correlation coefficient between the mix
of contributions and healthcare access did not show a statistically significant relationship
(r = -.035, p>.05). The values are insignificant suggesting that there is no relationship
between mix of contributions and healthcare access.
4.4.2.2 Correlation between Mix of Contributions and Equity in Contributions
Table 4.27 Correlation between Mix of Contributions and Equity in Contributions
Equity in Contributions
Mix of Contributions Pearson Correlation .
Sig. (2-tailed) .
N 224
Mix of contributions was not able to converge to form equity in contributions (Table 4.27).
These findings indicate that the relationship between mix of contributions and equity of
contributions is not statistically different from zero.
169
4.4.1.3 Correlation between Mix of Contributions and Quality of Care
Table 4.28 Correlation between Mix of Contributions and Quality of Care
Quality of Care
Mix of Contributions Pearson Correlation -.114
Sig. (2-tailed) .088
N 224
The Pearson correlation coefficient results in Table 4.28 shows insignificant relationships
between the mix of contributions and the quality of healthcare (r = -.114, p>.05). The
values are insignificant suggesting that there is no relationship between mix of
contributions and quality of care.
4.4.1.4 Correlation between Mix of Contributions and Sustainability
Table 4.29 Correlation between Mix of Contributions and Sustainability
Sustainability
Mix of Contributions Pearson Correlation -.127
Sig. (2-tailed) .058
N 224
Pearson correlation coefficient in Table 4.29 shows insignificant relationships between the
mix of contributions and sustainability (r = -.127, p>.05). Thus, the mix of contributions
in CBHIs does not significantly influence sustainability.
4.4.2 SEM results for Mix of Contributions and Equity in Healthcare
4.4.2.1 Cronbach’s Alpha Coefficients, AVE and KMO values for Mix of Contributions
Table 4.30 Cronbach’s Alpha Coefficients, AVE and KMO values for Mix of
Contributions
First order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Mix of
Contributions
0.714 MC2 0.523 0.500 0.715 51.11% MC1,MC3
MC4 0.523 0.715
170
The results presented in Table 4.30 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. The factors with low standardized regression weights were subsequently deleted. These
findings imply that the items of measure were measuring what they were initially set out to
measure, and therefore the data was maintained for further analysis.
4.4.2.2 Hypothesized effect of Mix of Contributions in CBHIs on equity in healthcare
Figure 4.13 Path coefficients for effect of mix of contributions on equity in healthcare in
CBHIs
171
Figure 4.14 t-values for effect of mix of contributions on equity in healthcare in CBHIs
Table 4.31 Path Coefficients (Mean, STDEV, t-values)
Original
Sample
(O)
Sample
Mean (M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Mix of
Contributions ->
Equity
-0.117995 -0.098878 0.112944 1.044718 0.29666
H0: Mix of contributions in CBHIs is not related to equity in health care.
H1: Mix of contributions in CBHIs is related to equity in health care.
Mix of contributions in CBHIs had a negative and insignificant effect on equity in healthcare
at the 0.05 level of significance (β=-0.118, t-value=1.045<1.96, p>0.05) as indicated in figure
4.13 and 4.14 and table 4.31. The null hypothesis is therefore not rejected.
4.5 Effect of Risk Pooling in CBHIs on Equity in Healthcare Indicators
This sub-section presents the results of risk pooling based on the primary data. Firstly, the
descriptive statistics are discussed. Secondly, Pearson‘s coefficient correlations between risk
pooling and the indicators of equity in healthcare constructs are presented. Thirdly, the
Cronbach‘s Alpha Coefficients, AVE and KMO values for risk pooling and the test for
hypothesized relationship between risk pooling and equity in healthcare are presented.
172
4.5.1 Descriptive Analysis - Extent Effect of Risk Pooling on Equity in Healthcare
Table 4.32 Extent Effect of Risk Pooling in CBHIs on Equity in Healthcare
Enhanced Risk Pooling
SD (%) D (%) N
(%)
A
(%)
SA
(%) Mean Std. D
Members:
Interact with the scheme‘s
management team about their
needs, concerns and make
suggestions for improvements
0 0 2 21 76 4.74 .487
Participate in setting of benefit
package 0 0 17 16 67 4.49 .775
Participate in setting the
premiums 0 0 1 24 75 4.74 .471
Are willing to cover the
poorest and vulnerable in the
community
CBHIs:
Uses existing Chamas,
community development
projects and credit schemes as
entry points for CBHIs
membership
1
0
38
0
56
0
5
7
0
92
2.67
4.92
.597
.288
Has a partnered with the
county / national government
and or NHIF
Have merged with other
CBHIs to form a network or a
federation
Social Solidarity
Members of the scheme have
expressed the opinion that if
they would not need healthcare
themselves, at least they had
done something good for the
community by contributing to
the insurance fund
0
0
0
0
0
0
0
2
0
0
0
8
9
23
19
38
91
77
81
52
4.90
4.76
4.81
4.41
.313
4.36
.395
.710
173
Table 4.32 Extent Effect of Risk Pooling in CBHIs on Equity in Healthcare (continued)
How much do you think
members of the CBHIs are
willing to contribute to pay
for health care services used
by the
None
of the
cost
(%)
Some
of the
cost
(%)
Half
of the
cost
(%)
Most
of the
cost
(%)
All of
the
cost
(%)
Mean Std.
Dev
Sick 0 0 1 35 64 4.63 .511
Poor 1 30 56 13 0 2.82 .659
According to the findings as illustrated in Table 4.32, the respondents agreed and reported
that the members of CBHIs comes from a wide range of social economic background (Mean
= 4.74; Std. Dev. = 0.487), CBHIs targets a large geographical / administrative area (mean =
4.49; std. dev. = 0.775), community members are encouraged to join CBHIs when they are
healthy (mean = 4.74; std. dev. = 0.471), CBHIs‘ have a waiting period before one can
benefit from insurance (mean = 4.74; std. dev. = 0.471), CBHIs has a risk transfer
mechanism (mean = 4.92; std. dev. = 0.288), CBHIs has partnered with the county/ national
and or NHIF (mean = 4.76; std. dev. = 0.436), CBHIs merged with other CBHIs to form a
network or a federation (mean = 4.81; std. dev. = 0.395). Responses on the aspect that CBHIs
has other branches in other geographical or administrative areas were neutral (mean = 2.67;
std. dev. = 0.597) indicating that some of the CBHIs from different geographical or
administrative regions have merged to form one CBHIs.
With regard to the social solidarity, the study findings showed that the members of the
scheme have expressed the opinion that if they would not need healthcare themselves, at least
they had done something good for the community by contributing to the insurance fund
(mean = 4.41; std. dev. = 0.709). Also, the findings shows that, the respondents agreed that
members of the CBHIs are willing to contribute to pay for healthcare services used by the
sick (mean = 4.63; std. dev. = 0. .511). It is also clear that the respondents neither agreed nor
disagreed that members of the CBHIs are willing to contribute to pay for healthcare services
used by the poor (mean = 2.82; std. dev. = 0.659). The findings show that the respondents
had mixed reactions with some displaying willingness to contribute for health services used
by the poor, other were unwilling while a significant number were neutral on this aspect.
174
4.5.2 Correlation between Risk Pooling in CBHIs and Equity in HealthCare Indicators
This section presents the Pearson correlation and SEM results on the risk pooling and its
effect on indicators of equity in health care.
4.5.2.1 Correlation between Risk Pooling and Healthcare Access
Table 4.33 Correlation between Risk pooling in CBHIs and Healthcare Access
Healthcare Access
Risk Pooling Pearson Correlation .494**
Sig. (2-tailed) .000
N 224
According to the findings in Table 4.33, Pearson correlation coefficient between risk pooling
in CBHIs and healthcare access show a statistically significant but weaker relationship (r
=.494, p < .000).
4.5.2.2 Correlation between Risk Pooling and Equity in Contributions
Table 4.34 Correlation between Risk Pooling and Equity in Contributions
Equity in Contributions
Risk Pooling Pearson Correlation .
Sig. (2-tailed) .
N 224
Risk pooling in CBHIs was not able to converge to form equity in contributions (Table 4.34).
These findings indicate that the relationship between risk pooling in CBHIs and equity of
contributions is not statistically different from zero.
175
4.5.2.3 Correlation between Risk Pooling and Quality of Care
Table 4.35 Correlation between Risk Pooling and Quality of Care
Quality of Care
Risk Pooling Pearson Correlation .490**
Sig. (2-tailed) .000
N 224
The Pearson correlation coefficient in Table 4.35 shows a statistically significant but weaker
relationships between the risk pooling and the quality of healthcare (r =.490, p < .000)
4.5.2.4 Correlation between Risk Pooling and Sustainability
Table 4.36 Correlation between Risk Pooling and Sustainability
Sustainability
Risk Pooling Pearson Correlation .497**
Sig. (2-tailed) .000
N 224
Findings of Pearson correlation coefficient shows a statistically significant but weaker
relationships between risk pooling in CBHIs and sustainability (r =.497, p < .000). Thus, the
risk pooling in CBHIs fairly influences sustainability.
4.5.3 SEM analysis for Risk Pooling and Equity in Healthcare
4.5.3.1 Cronbach’s Alpha Coefficients, AVE and KMO values for Risk Pooling
Table 4.37 Cronbach’s Alpha Coefficients, AVE and KMO values for Risk Pooling
First
order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Enhanced
risk
pooling
0.816 ERP1 0.669 0.602 0.79 83.73%
ERP2,
ERP5,
ERP7
ERP3 0.761
0.969
ERP4 0.525
0.993
ERP6 0.586
0.957
ERP8 0.559 0.834
176
The results presented in Table 4.37 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. The factors with low standardized regression weights were subsequently deleted. These
findings imply that the items of measure were measuring what they were initially set out to
measure, and therefore the data was maintained for further analysis.
4.5.3.2 Hypothesized effect of risk pooling in CBHIs on equity in healthcare
Figure 4.15 Path coefficients for effect of Risk pooling in CBHIs on equity in healthcare
Figure 4.16 t-values for effect of Risk pooling in CBHIs on equity in healthcare
177
Table 4.38 Path Coefficients (Mean, STDEV, t-values)
Original
Sample (O)
Sample
Mean (M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Risk Pooling -
> Equity 0.552729 0.561240 0.050984 10.841140 0.0000
H0 Risk pooling in CBHIs is not related to equity in healthcare.
H1: Risk pooling in CBHIs is related equity in healthcare.
Risk pooling in CBHIs had a positive statistically and significant equity in healthcare at the
0.05 level of significance (β=0.553, t-value=10.81 >1.96, p<0.05) as indicated in figures 4.15
and 4.16 and table 4.38. The null hypothesis is therefore rejected and the alternative
Hypothesis H1 that stated that risk pooling in CBHIs is related to equity in health care is
supported. Results thus reveal that, when Risk pooling in CBHIs increases by 1 unit, Equity
in healthcare increases by 0.553 units. Figure 4.15 shows that risk pooling had a coefficient r2
mean of 0.306 showing the proportion of variation in dependent variable explained by the
SEM model. r2 indicates that 30.6% of the variations in equity in healthcare can be accounted
for by Risk pooling in CBHIs.
4.6 Effect of Strategic Purchasing in CBHIs on Equity in Healthcare Indicators
This sub-section presents the results of strategic purchasing based on the primary data.
Firstly, the descriptive statistics are discussed. Secondly, Pearson‘s coefficient correlations
between strategic purchasing and the indicators of equity in healthcare constructs are
presented. Thirdly, the Cronbach‘s Alpha Coefficients, AVE and KMO values for strategic
purchasing and the test for hypothesized relationship between strategic purchasing and equity
in healthcare are presented.
178
4.6.1 Descriptive Analysis - Extent Effect of Strategic Purchasing on Equity in
Healthcare
Table 4.39 Descriptive Analysis - Extent Effect of Strategic Purchasing on Equity in
Healthcare
Strategic
Purchasing: CBHIs
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
%)
Mean Std.
Dev
Signed a contract
with health services
providers
0 0 0 20 80 4.79
.419
Select accredited
providers 0 0 1 31 68 4.67
.501
Providers must
provide services
according to
conditions put
forward by the CBHIs
0 0 1 25 74 4.73
.475
Allocate resources
based on population
needs
Has been successful
in negotiating
agreeable terms and
contract with service
providers
0
0
0
0
0
0
20
20
80 4.79
80 4.79
.419
.420
According to the study findings CBHIs have put in place strategic purchasing mechanisms as
indicated in Table 4.39. The respondents strongly agreed indicating that CBHIs only select
providers who are accredited (mean = 4.79; std. dev. = 0.419), allocation of resources are
done based on population needs (mean = 4.67; std. dev. = 0.501), CBHIs allocate resources
based on population needs (mean = 4.73; std. dev. = 0.475), CBHIs has been successful in
negotiating agreeable terms and contract with service providers in terms of service quality,
fee, and reduction in unnecessary services/prescription (moral hazard) (mean = 4.79; std. dev.
= 0.419).
179
4.6.2 Correlation between Enrollment in CBHIs and Equity in HealthCare Indicators
This section presents the Pearson correlation and SEM results on the strategic purchasing and
its effect on indicators of equity in health care.
4.6.2.1 Correlation between Strategic Purchasing and Healthcare Access
Table 4.40 Correlation between Strategic Purchasing and Healthcare Access
Healthcare Access
Strategic Purchasing Pearson Correlation .785**
Sig. (2-tailed) .000
N 224
Findings in Table 4.40 indicates that Pearson correlation coefficient between strategic
purchasing in CBHIs and healthcare access has statistically significant strong positive
relationship (r =.785, p < .000).
4.6.2.2 Correlation between Strategic Purchasing and Equity in Contributions
Table 4.41 Correlation between Strategic Purchasing and Equity in Contributions
Equity in Contributions
Strategic Purchasing Pearson Correlation .
Sig. (2-tailed) .
N 224
Strategic purchasing in CBHIs was not able to converge to form equity in contributions
(Table 4.41). These findings indicate that the relationship between strategic purchasing in
CBHIs and equity of contributions is not statistically different from zero.
180
4.6.2.3 Correlation between Strategic Purchasing and Quality of Care
Table 4.42 Correlation between Strategic Purchasing and Quality of Care
Quality of Care
Strategic Purchasing Pearson Correlation .903**
Sig. (2-tailed) .000
N 224
The Pearson correlation coefficient results in Table 4.42 indicates that strategic purchasing
and quality of care has a statistically significant relationships (r = .903, p<.01). This shows
that strategic purchasing influences quality of care.
4.6.2.4 Correlation between Strategic Purchasing and Sustainability
Table 4.43 Correlation between Strategic Purchasing and Sustainability
Sustainability
Strategic Purchasing Pearson Correlation .840**
Sig. (2-tailed) .000
N 224
The Pearson correlation coefficient in Table 4.43 shows a statistically significant strong
positive relationships between the strategic purchasing and the sustainability (r =.840, p <
.01). Thus, strategic purchasing in CBHIs influences sustainability.
181
4.6.3 SEM analysis for Strategic Purchasing and Equity in Healthcare
4.6.3.1 Cronbach’s Alpha Coefficients, AVE and KMO values for Strategic Purchasing
Table 4.44 Cronbach’s Alpha Coefficients, AVE and KMO values for Strategic
Purchasing
First order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Strategic
purchasing
0.876 SP1 0.702 0.802 0.832 73.95%
SP5 SP2 0.603
0.777
SP3 0.821
0.91
SP4 0.793 0.892
The results presented in Table 4.44 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. The factors with low standardized regression weights were subsequently deleted. These
findings imply that the items of measure were measuring what they were initially set out to
measure, and therefore the data was maintained for further analysis.
4.6.3.2 Hypothesized effect of strategic purchasing in CBHIs on equity in health care
Figure 4.17 Path coefficients for effect of Strategic purchasing in CBHIs on equity in
healthcare
182
Figure 4.18 t-values for effect of Strategic purchasing in CBHIs on equity in health care
Table 4.45 Path Coefficients (Mean, STDEV, t-values)
Original
Sample (O)
Sample
Mean (M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Purchasing ->
Equity 0.552729 0.561240 0.050984 10.841140 0.0000
H0 Strategic purchasing in CBHIs is not related to equity in healthcare.
H1: Strategic purchasing in CBHIs is related to equity in healthcare.
Strategic purchasing in CBHIs had a positive statistically and significant equity in healthcare
at the 0.05 level of significance (β=0.911, t-value=54.416 >1.96, p<0.05) as indicated in
figures 4.17 and 4.18 and table 4.39. The null hypothesis is therefore rejected and the
alternative Hypothesis H1 that stated that Strategic purchasing in CBHIs is related to equity in
healthcare is supported. Results thus reveal that, when Strategic purchasing in CBHIs
increases by 1 unit, Equity in healthcare increases by 0.830 units. Figure 4.17 shows that
Strategic purchasing had a coefficient r2 mean of 0.830 showing the proportion of variation
in dependent variable explained by the SEM model. r2 indicates that 83% of the variations in
equity in healthcare can be accounted for by strategic purchasing.
183
4.7 Moderating effect of Government Stewardship on Equity in Healthcare
This section presents the descriptive statistics, Cronbach‘s Alpha Coefficients, AVE and
KMO values for government stewardship and the test for hypothesized relationship of
moderating effect of government stewardship and equity in healthcare.
4.7.1 Descriptive Analysis - Respondents on Extent Effect of Government Stewardship
Indicators on Equity in Healthcare
Table 4.46 Descriptive Analysis - Extent Effect of Government Stewardship on Equity
in Healthcare - Advisory Role on Design
Advisory Role on Design
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Startup of CBHIs on
minimum of population
enrolled
0
47 48 6 0 2.58 .601
A waiting period 0 47 47 6 0 2.59 .629
Enrollment of households
as opposed to individuals
0
47
47
6
0
2.59
.621
A flexible premium
collection system
0 44 51 5 0 2.62 .603
A benefit package based on
target population needs
A standard treatment
protocols
A standard referral
procedure
Consolidation of CBHIs
Creation of a risk
equalization fund
Community participation in
management and decision
making
0
0
0
0
0
0
48
49
44
43
44
45
46
45
50
52
50
49
6
6
6
6
6
6
0
0
0
0
0
0
2.29
2.58
2.62
2.64
2.63
2.62
.622
.623
.616
.612
.615
.616
As illustrated in Table 4.46, there is some extent of government influence on the aspect of
minimum percentage of population required to enroll before a CBHIs begins its operations
(mean = 2.58; std. dev. = 0.601). Similarly, there is some extent of government influence on
the awaiting period set in CBHIs (mean = 2.59; std. dev. = 0.629). Further, findings show
that to some extent, the government influences the enrolment of households as opposed to
individuals (mean = 2.59; std. dev. = 0. 621), a flexible premium collection system (mean =
184
2.62; std. dev. = 0.603), a benefit package that reflect the needs of the target population
(mean = 2.59; std. dev. = 0.622), a standard treatment protocols for the members of the
CBHIs (mean = 2.58; std. dev. = 0.623), a standard referral procedure (mean = 2.62; std.
dev. = 0.616), consolidation of CBHIs through a federation or network (mean = 2.64; std.
dev. = 0.612), creation of a risk equalization fund (mean = 2.63; std. dev. = 0.615) as well
that the government encourages community participation in management and decision
making (mean = 2.62; std. dev. = 0.616).
Table 4.47 Descriptive Analysis - Extent Effect of Government Stewardship on Equity
in Healthcare- Monitoring
Monitoring of
CBHIs Activities
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std. Dev
Tracks the
progress of CBHIs
through time
0 47 47 6 0 2.59 .628
Monitors the basic
performance of
CBHIs
0 44 50 6 0 2.62 .609
As illustrated in Table 4.47, the government does not does not track the performance of
CBHIs through time (mean = 1.72; std. dev. = 0.713) nor does it monitor the performance of
the CBHIs (mean = 1.54; std. dev. = 0.774).
185
Table 4.48 Descriptive Analysis - Extent Effect of Government Stewardship on Equity
in Healthcare- Training
Training
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Determination of
benefit packages 0 44 50 5 0 2.62 .609
Determination of
contributions 0 47 47 5 0 2.59 .615
Collection of
contributions 0 49 45 6 0 2.58 .624
Claims processing 0 48 45 6 0 2.58 .631
Use of MIS 0 47 46 6 0 2.58 .630
Establishment of
Health Insurance
Development Plans
0 47 47 6 0 2.60 .620
Educational visits 0 41 53 6 0 2.66 .608
The findings shows that the government does not organize trainings on determination of
benefit packages (mean = 1.73; std. dev. = 0.852), setting of contributions (mean = 1.79; std.
dev. = 0.807), collection of premiums (1.79; std. dev. = 0.862), claims processing (1.57; std.
dev. = 0.590), use of management information system (1.71; std. dev. = 0.842), establishment
of Health Insurance Development Plans (1.75; std. dev. = 0.842). Findings from the table
also illustrate that the government does not organize exchange visits to other CBHIs as
indicated by a mean of 1.73 and a standard deviation of 0.852.
186
Table 4.49 Descriptive Analysis - Extent Effect of Government Stewardship on Equity
in Healthcare- Co-financing
Co- Financing: The
government
Strongly
Disagree
(%)
Disagree
(%)
Neutral
(%)
Agree
(%)
Strongly
Agree
(%)
Mean Std.
Dev
Partially or fully
subsidizes the poorest
and vulnerable
members of the
community through
CBHIs
0 41 54 5 0 2.66 .601
Set a solidarity fund
for financing
epidemics and
deficits of the CBHIs
0 42 52 6 0 2.65 .609
With regard to the co-financing role, the findings show that there is lack of financial support
from government in form of general revenue and or donors support to cover the poorest and
vulnerable through the CBHIs. In addition, government initiated re-insurance and solidarity
funds that reinforce CBHIs financial sustainability are non-existence. The government and
or donors do not partially or fully subsidize the poor and vulnerable members of the
community through the CBHIs. This is illustrated by a mean of 2.66 which is in the interval
for moderate disagreement with a standard deviation of 0.601. The findings also indicated
that the government has not set aside a solidarity fund to caution CBHIs against epidemics
and deficits. This evidenced by a mean of 2.65 and a standard deviation of 0.609. Generally,
there is no much variance of the individual responses from the calculated mean. This implies
that there is absence of government stewardship in co-financing healthcare with CBHIs in
Kenya.
187
4.7.2 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship
4.7.2.1 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship – Design
Table 4.50 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Design
First order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Design
0.876 AD1 0.902 0.956 0.920 90.23% None
AD2 0.948
0.958
AD3 0.954
0.963
AD4 0.920
0.935
AD5 0.960
0.968
AD6 0.966
0.973
AD7 0.938
0.950
AD8 0.922
0.937
AD9 0.932
0.945
AD10 0.936 0.949
The results presented in Table 4.50 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. These findings imply that the items of measure were measuring what they were initially
set out to measure, and therefore the data was maintained for further analysis.
188
Table 4.51 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Monitoring
First order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Monitoring 0.958 MO1 0.919 0.5 0.980 95.97% None
MO2 0.919 0.980
The results presented in Table 4.51 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. These findings imply that the items of measure were measuring what they were initially
set out to measure, and therefore the data was maintained for further analysis.
Table 4.52 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Training
First order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Training
0.986 TR1 0.934 0.913 0.952 92.19% None
TR2 0.959
0.970
TR3 0.962
0.973
TR4 0.945
0.959
TR5 0.951
0.964
TR6 0.967
0.976
TR7 0.901 0.926
The results presented in Table 4.52 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. These findings imply that the items of measure were measuring what they were initially
set out to measure, and therefore the data was maintained for further analysis.
189
Table 4.53 Cronbach’s Alpha Coefficients, AVE and KMO values for Government
Stewardship - Co-financing
First
order
constructs
Cronbach’s
alpha Item
Item total
correlation KMO
PCA
component
loading
variance
extracted
Items
deleted
Co-
financing
0.903 COF1 0.823 0.5 0.955 91.13% None
COF2 0.823 0.955
The results presented in Table 4.53 showed Cronbach‘s alpha coefficients of above the 0.7
threshold for all first order constructs, total item correlations of above 0.3, AVE of above
65%, KMO values greater than 0.5 and satisfactory principal component loadings of above
0.50. These findings imply that the items of measure were measuring what they were initially
set out to measure, and therefore the data was maintained for further analysis.
4.7.3 Multicollinearity Test
Table 4.54 Multicollinearity Test
Collinearity Statistics
Tolerance VIF
Enrolment .630 1.588
Government .763 1.311
Mix of contributions .551 1.815
Purchasing .814 1.228
Risk pooling .715 1.399
The table above indicates the test results for multicollinearity, using both the VIF and
tolerance. With VIF values being less than 5, it was concluded that there was no presence of
multicollinearity in this study. The VIF shows us how much the variance of the coefficient
estimate is being inflated by multicollinearity.
190
4.8 Overall Model
4.8.1 Reliability
Details of construct reliability are presented in Table 4.54.
Table 4.55 Construct reliability
Construct Composite
Reliability
Cronbach's
Alpha
Enrolment 0.961604 0.954142
Equity 0.966859 0.960305
Government 0.993834 0.993534
Mix of Contributions 0.750234 0.746829
Purchasing 0.914737 0.874918
Risk Pooling 0.870987 0.823401
4.8.2 Convergent Validity
The CFA results of item loadings and their respective t-values are reported in Table 4.56.
The items were significantly loaded on the proposed factors with loading higher than 0.5.
The AVE of all the constructs were above the 0.5 threshold indicating that the measurement
scales exhibited adequate measurement validity.
Table 4.56 Convergent Validity of outer model
Construct
Original
Sample
(O)
Sample
Mean
(M)
Standard
Deviation
(STDEV)
Standard
Error
(STERR)
t-
value AVE
Equity
0.786
ACC4 <- Equity 0.806 0.809 0.076 0.076 10.669
AMC1 <- Equity 0.956 0.955 0.015 0.015 61.957
AMC6 <- Equity 0.815 0.815 0.056 0.056 14.529
FS1 <- Equity 0.935 0.932 0.023 0.023 39.835
FS10 <- Equity 0.878 0.875 0.055 0.055 16.042
FS5 <- Equity 0.839 0.837 0.060 0.060 13.879
QOC8 <- Equity 0.950 0.947 0.017 0.017 56.184
QOC9 <- Equity 0.896 0.893 0.037 0.037 24.267
Government
0.885
AD1 <- Government 0.913 0.884 0.132 0.132 6.900
AD10 <-
Government 0.952 0.911 0.170 0.170 5.598
AD2 <- Government
191
0.944
0.910
0.147
0.147
6.410
AD3 <- Government 0.958 0.922 0.129 0.129 7.408
AD4 <- Government 0.927 0.889 0.137 0.137 6.763
AD5 <- Government 0.964 0.927 0.161 0.161 6.007
AD6 <- Government 0.968 0.931 0.161 0.161 6.002
AD7 <- Government 0.936 0.899 0.156 0.156 6.012
AD8 <- Government 0.929 0.887 0.164 0.164 5.666
AD9 <- Government 0.941 0.904 0.139 0.139 6.759
COF1 <-
Government 0.892 0.858 0.107 0.107 8.304
COF2 <-
Government 0.911 0.874 0.129 0.129 7.078
Risk pooling
0.578
ERP1 <- Risk
pooling 0.804 0.801 0.056 0.056 14.312
ERP3 <- Risk
Pooling 0.881 0.878 0.041 0.041 21.255
ERP4 <- Risk
Pooling 0.625 0.596 0.127 0.127 4.935
ERP6 <- Risk
Pooling 0.684 0.663 0.112 0.112 6.121
ERP8 <- Risk
Pooling 0.779 0.787 0.052 0.052 14.961
Mix of
Contributions 0.552
MC1 <- Mix of Cont 0.913 0.860 0.186 0.186 4.912
MC2 <- Mix of Cont 0.881 0.844 0.191 0.196 4.618
MO1 <- Government 0.937 0.901 0.146 0.146 6.418
MO2 <- Government 0.938 0.897 0.152 0.152 6.173
TR1 <- Government 0.938 0.897 0.166 0.166 5.662
TR2 <- Government 0.961 0.921 0.162 0.162 5.948
TR3 <- Government 0.973 0.935 0.160 0.160 6.079
TR4 <- Government 0.950 0.914 0.143 0.143 6.660
TR5 <- Government 0.946 0.910 0.147 0.147 6.442
TR6 <- Government 0.968 0.931 0.154 0.154 6.286
TR7 <- Government 0.902 0.863 0.175 0.175 5.154
Purchasing
0.729
SP1 <- Purchasing 0.826 0.822 0.062 0.062 13.380
SP2 <- Purchasing 0.772 0.773 0.069 0.069 11.186
SP3 <- Purchasing 0.916 0.916 0.024 0.024 37.914
SP4 <- Purchasing
Table 4.56 Convergent Validity of the outer Model
192
0.894
0.897
0.033
0.033
26.847
Enrolment
0.738
AFF4 <- Enrolment 0.916 0.920 0.036 0.036 25.181
TM1 <- Enrolment 0.859 0.860 0.042 0.042 20.347
TM3 <- Enrolment 0.869 0.865 0.045 0.045 19.301
TRU1 <- Enrolment 0.853 0.852 0.048 0.048 17.894
TRU2 <- Enrolment 0.918 0.920 0.027 0.027 34.393
TRU3 <- Enrolment 0.834 0.831 0.046 0.046 18.041
TRU4 <- Enrolment 0.919 0.922 0.023 0.023 40.455
TRU5 <- Enrolment 0.581 0.573 0.086 0.086 6.754
AFF1 <- Enrolment 0.931 0.936 0.017 0.017 56.313
4.8.3 Discriminant validity
As indicated in Table 4.57 all the constructs in the model met this criteria indicating that
discriminant validity is supported.
Table 4.57 Measures of Discriminant Validity
Construct Fornell Larker Measure (AVE >
highest correlation2)
Enrolment 0.7383>0.6685
Equity 0.78558>0.66851
Government 0.88479>0.003921
Mix of Contributions 0.552417>0.091431
Purchasing 0.729302>0.658589
Risk Pooling 0.577872>0.369683
Table 4.58 Latent Variable
Correlations / Correlation matrix
of constructs
Enrolment Equity Government
Mix of
Contributio
ns
Purchasing Risk
Pooling
Enrolment 1.00000
Equity 0.81762 1.00000
Government -0.06262 -0.06052 1.00000
Mix of
Contributions -0.07215 -0.11802 0.30238 1.00000
Purchasing 0.88472 0.81154 -0.06741 -0.08386 1.00000
Risk Pooling 0.60802 0.55225 -0.03748 -0.08032 0.52640 1.0000
0
Table 4.56 Convergent Validity of the outer Model
193
4.8.4 Structural Model Estimation and Hypothesis Testing
Figure 4.19, 4.20, 4.21 and 4.22 presents the paths coefficients and t-statistics for the overall
model without moderation and the moderated overall model.
194
4.8.5 Optimum model without Moderation
Figure 4.19 Path coefficients for the optimum model without moderation
195
Figure 4.20 t- values for the optimum model without moderation
196
Table 4.59 Path Coefficients (Mean, STDEV, t-Values)
Original
Sample
(O)
Sample
Mean
(M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Enrolment -> Equity 0.473319 0.476809 0.147603 3.206705 0.001429
Purchasing -> Equity 0.495651 0.493281 0.148596 3.335560 0.000915
Figure 4.19 shows that the endogenous latent variable Equity in Healthcare had a coefficient
r2 mean of 0.882 implying that the out the four exogenous variables only two exogenous
variables; Enrolment and Strategic Purchasing explain 88.2% of variation in Equity in
Healthcare. Enrolment account for 47.3% of variation in Equity in Healthcare while Strategic
Purchasing account for 49.5% of variation Equity in Healthcare. Figure 4.20 suggests that
the hypothesized paths between Enrolment and Equity in Healthcare (β=3.207) and Strategic
Purchasing and Equity in Healthcare (β=3.336 are significant at 0.05 level of significance.
197
4.8.10 Optimum model with Moderation
Figure 4.21 Path coefficients for the optimum moderated model
198
Figure 4.22 t-values for the optimum moderated model
199
Table 4.60 Path Coefficients (Mean, STDEV, t-values)
Original
Sample (O)
Sample
Mean
(M)
Standard
Deviation
(STDEV)
T Statistics
(|O/STDEV|) P Values
Enrolment*Gov ->
Equity -0.054218 -0.05422 0.105034 0.516191 0.60595
purchasing*Gov ->
Equity -0.139900 -0.01541 0.135515 1.032357 0.302405
Figure 4.21 shows that the endogenous latent variable Equity in Healthcare had a coefficient
r2 mean of 0.898 implying that the two exogenous variables, Enrolment and Strategic
Purchasing explain 89.8% of variation in Equity in Healthcare. This represents an
improvement in the variation explained compared to when the moderating variable,
government stewardship was excluded (r2
increases by 7%).
Under moderation, Enrolment account for 53.2% of variation in Equity in Healthcare while
Strategic Purchasing account for 32.5% of variation Equity in Healthcare. This presents a
slight improvement in variation explained by Enrolment (r2
increases by 5.9%) while the
variation of Equity in Healthcare accounted by Strategic Purchasing decreases by 17%.
Figure 4.22 suggests that the hypothesized paths between Enrolment and Equity in
Healthcare (β=4.596) and Strategic Purchasing and Equity in Healthcare (β=2.591) are
significant at 0.05 level of significance.
4.9 Chapter Summary
This chapter reported the descriptive and inferential analysis of the results from the main
study. First, the chapter presents the descriptive analysis, followed by the diagnostic test and
inferential statistics. Correlation analysis was used to establish the degree of relationship
between sub-constructs of the independent variables and the sub-constructs of the dependent
variable. A six step approach to SEM was applied using SmartPLS software version 3.0.
SmartPLS was employed to develop the measurement and structural model under study, test
hypothesized relationships between variables and bootstrap. The results shows enrolment in
CBHIs influences and equity in healthcare positively, mix of contributions in CBHIs
influences equity in healthcare negatively, risk pooling in CBHIs influences equity in
healthcare positively, strategic purchasing in CBHIs influences equity in healthcare
200
positively at the 0.05 level of significance. The optimum model shows that out of the four
exogenous variables only two of them; enrolment and strategic purchasing influences equity
in healthcare. When the optimum model was moderated by government stewardship,
enrolment in CBHIs improves while strategic purchasing decreases significantly. The next
chapter provides a discussion of the findings, conclusions and recommendations.
201
CHAPTER FIVE
5.0 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
5.1 Introduction
This chapter provides a discussion of the findings reported in chapter four. The findings are
presented into four sections. It also presents the conclusions drawn from the study as well as
the suggestions on how to improve the performance of CBHIs as an alternative mechanism of
extending equity in healthcare and UHC for excluded groups. The chapter concludes with
suggestions of further research.
5.2 Summary of Findings
The purpose of the study was to examine innovative healthcare financing and equity through
CBHIs in Kenya. To accomplish this, the study investigated the effect of the health financing
functions enrolment, mix of contributions, risk pooling, strategic purchasing and the
moderating effects of government stewardship in CBHIs on equity in healthcare and drew
inferences from 82 CBHIs where responses were sought from four members of each CBHIs
management team. A sample size of 318 was determined using Yamane (1967) formula. The
study used survey sampling to select the CBHIs that offer pre-paid community health
financing. A structured questionnaire was used to collect primary data while a secondary data
sheet was used to capture longitudinal data on enrolment, mix of contributions and equity in
healthcare (number of members accessing healthcare and financial sustainability). The
questionnaire was structured and mostly had closed-ended questions that sought responses in
a five point likert-type scale. Comments and recommendations of health financing experts
and supervisors on representativeness and appropriateness of the survey questionnaire were
sought before the pilot study was carried out. A draft questionnaire was pilot tested on 10
respondents drawn from three CBHIs under the network of BIDII. Ethical research
procedures were observed throughout the entire research process. The data was analyzed
using SPSS version 20, MS. Excel, Karl Pearson‘s coefficient of correlation and SmartPLS
version 3.
202
Findings based on the secondary data show that 91.5% of CBHIs covered up to 500
households; average enrolment across the studied CBHIs was 169 households; majority of
CBHIs have not been able to enroll the targeted households; most of the enrolled households
(7336) had purchased the cheapest cover which cost Kes.700 per year which targets small
households while 1680 dropped from NHIF covers. Further, the study found that majority of
CBHIs uses fee for service method to pay services providers for both inpatient and outpatient
service whilst 62.9% of the covered population lives with a proximity of 5 kilometers to the
nearest contracted service provider. With regard to the mix of contribution, the study
established that contributions from members through CBHIs only product and CBHIs and
NHIF products are the only sources of CBHIs funds. Finding on the trends of total premiums
collected, healthcare cost reimbursements, administration cost and deficit or surplus in
CBHIs between 2010-2015 shows an up and down fluctuation of the total premiums
collected, healthcare cost reimbursements and deficit or surplus while the administration cost
remained constant throughout the entire period.
Findings based on primary data show that respondents came from CBHIs that have been in
operation shows that respondents came from CBHIs that were in existence for 1-5 years
(70.1% of CBHIs), 6-10 years (22.1% of CBHIs) and 11-15 years (7.8%). The study also
established that enrolment and strategic purchasing in CBHIs significantly influent equity in
healthcare in Kenya. The two independent variables had a coefficient of 0.882, indicating
that 88.2% of the variations in equity in healthcare can be accounted for by two independent
variables (enrolment and strategic purchasing in CBHIs). When moderated by government
stewardship, the variation of equity in healthcare that is accounted for by enrolment and
strategic purchasing increases to 89.8%. There was a positive statistically and significant
relationship between enrolment in CBHIs and equity in healthcare (β=0.908, t-value=52,382
>1.96, p>0.05). Pearson correlation coefficient demonstrate a statistically significant strong
positive relationship between the sub-constructs of enrollment and the sub-constructs under
equity in healthcare including; affordability and healthcare access (r=0.726, p=0.000),
affordability and equity in contributions (r=0.933, p=0.000); affordability and quality of care
(r=0.936, p=0.000); affordability and sustainability (r=0.878, p=0.000); timing of collections
and healthcare access (r=0.749, p=0.000), timing of collections and equity in contributions
203
(r=0.677, p=0.000); timing of collections and quality of care (r=0.749, p=0.000) and timing
of collections and sustainability (r=0.664, p=0.000); trust and healthcare access (r=0.779,
p=0.000), trust and equity in contributions (r=0.839, p=0.000); trust and quality of care
(r=0.872, p=0.000) and trust and sustainability (r=0.797, p=0.000).
Mix of contributions in CBHIs had a negative and insignificant effect on equity in healthcare
(β=-0.118, t-value=1.045<1.96, p>0.05). Similarly, Pearson correlation coefficient shows an
insignificant relationships between the mix of contributions and three of the sub-constructs
under equity in healthcare including healthcare access (r=-.035, p>.05), quality of care (r=-
.114, p>.05) and sustainability (r=-.127, p>.05). Mix of contributions and equity in
contributions did not show any relationship. Risk pooling in CBHIs had a positive
statistically and significant effect on equity in healthcare (β=0.553, t-value=10.841>1.96,
p<0.05). Pearson correlation coefficient shows a statistically significant but weaker
relationships between risk pooling in CBHIs and three sub-constructs under equity in
healthcare including healthcare access (r=.494, p<.000), quality of care (r=.490, p<.000) and
sustainability (r=.497, p<.000). Risk pooling and equity in contributions did not show any
relationship. Strategic purchasing in CBHIs had a positive statistically and significant effect
on equity in healthcare at the 0.05 level of significance (β=0.830, t-value=54.416 >1.96,
p<0.05). Pearson correlation coefficient shows a statistically significant strong positive
relationship strategic purchasing in CBHIs and three of the sub-constructs under equity in
healthcare including healthcare access (r=.785, p<.000), quality of care (r=.903, p<.000) and
sustainability (r=.840, p<.000). Strategic purchasing and equity in contributions did not show
any relationship.
5.3 Discussion of Results
This sub-section presents an in-depth discussion of the finding from the data analysis and a
comparison with previous empirical studies related to the research objectives.
5.3.1 Enrolment and Equity in Healthcare
The study established that there was a positive and significant relationship between
enrolment in CBHIs and equity in healthcare at 5% level of confidence. Enrolment was a
204
second order latent construct whose antecedents were affordability of contributions, unit of
membership, timing of collections and trust. Carrin et al. (2005) used these antecedents to
study the contribution of CBHIs to performance of health systems in developing countries.
These precedents were also applied in an analytical framework developed by Mathauer &
Carrin (2010). The framework uses the health financing functions to analyze the performance
of a health financing system.
This finding is consistent with the findings of Carrin et al. (2005) & McCord et al. (2012)
that established that high membership rates increase equity in healthcare in CBHIs. Findings
from this study show that the average enrolment in the studied CBHIs was 169 households
with majority of CBHIs enrolling up to 500 households. This finding is in congruence with
WHO (2010) that voluntary schemes attract fewer members. As a result of changes in NHIF
premiums prices in 2015, 1680 households dropped the cover. Dercon et al. (2012) concur
that low income household‘s display high price elasticity due to low and irregular income, a
factor that influences demand for health insurance. CBHIs can make premiums affordable in
various ways including allowing members to allocate their spare income among preferred
products, allowing premium payment in kind, installmental payments, saving linked
payments system and premium subsidies for the poorest segment (Carrin et al., 2005; Dror,
2007; Aggarwal, 2010; ILO, 2013). The study found that together the 82 CBHIs that were
studied offers 10 different products. Members are allowed to allocate their income among the
preferred products. Majority of members have purchased a small household at Kes 700; the
cover offers services in public hospitals only. The high uptake levels shows that the cover
responds to the households‘ ability to pay since it is the cheapest CBHIs only cover. The
CBHIs also encourage members to use savings-linked premium payment mechanisms. The
study however found that the schemes do not allow installmental payments, in-kind
payments for premium are not allowed and subsidies for poor people through CBHIs are non-
existent.
Recruiting members from villages, pre-existing mutual and development groups makes it
easy for CBHIs to enroll more members. Similarly, adequate membership rates are easily
achieved when CBHIs use households as a unit of membership. Such enrolment strategies
205
allow CBHIs to extend membership beyond those who would join the scheme voluntarily
(Atim, 1998), and thus mitigate adverse selection problem. The finding of this study supports
earlier findings of various researchers by Atim (1998), Desmet at al., (1999), Carrin (2003) &
Carrin et al. (2005) that found that most CBHIs recruit their members from mutual benefits
societies besides using family as a unit of membership. To enhance enrolment through
inclusion of all members of the excluded segment, schemes need to ensure that members of
the uncovered segments are included. This study found that CBHIs membership is open to
the poorest and vulnerable groups while majority of members (63%) live with proximity of 5
kilometers from contracted healthcare providers. Result on the influence of geographical
proximity on enrolment is supported by findings of Carrin et al. (2005). Carrin et al. (2005)
found that physical proximity in China Rural Co-operative was suggestive of the low
enrolment rates and small risk pools given that enrolment was reliant on trust among
community members.
Finding of this study also show that CBHIs allow members to pay premium during cash
inflows. This finding is consistent with the findings of Criel (1998); Criel (2002); De Allegri
et al. (2006); ILO (2012); Chen et al. (2012) & Matul et al., (2013) that found that
households express a will to be allowed match cash inflows with premium payment periods.
This study also found that members are required to pay premium in a single payment. This
finding is inconsistent with the findings by Basaza et al. (2007) that found that a Ugandan
mutual greatly increased its membership by spreading its premium payments over the year.
Trust is a critical determinant of uptake in CBHIs given their voluntary nature of enrolment.
Chen et al. (2012) puts forward that trust in communities can manifest in three ways, trust
among community members, trust in health providers contracted by the scheme, and trust in
the management team and CBHI scheme. This trust is fostered through involvement of
members in setting of premiums and benefits package. Frequent interactions of members,
management team and service providers also offer forums for members to raise concerns and
make suggestions for improvements on the services offered by the providers. The study also
established members interact with the scheme management team during annual general
meetings and other scheduled meeting where they air their views, concerns and give
206
feedback on issues concerning CBHIs. Additionally, that there exist high levels trusts among
members since CBHIs draw their membership from mutual benefits societies which already
enjoy high degree of reciprocity and/or mutually beneficial support. Members are actively
involved in determination of benefits packages and setting of premiums. These results is
supported by finding by Chen et al. (2012) that found that trust in CBHI management team
and in the scheme itself has a positive impact on enrolment decisions. Further, the perception
of fairness and transparency in schemes is better positioned to nature trust relationships with
the community (Chen et al., 2012).
5.3.2 Mix of Contributions and Equity in Healthcare
The WHO report of 2010 postulates prepayments as the most efficient and equitable method
of raising funds for healthcare. Hypothecated, general and payroll taxes, insurance or a
combination of two have been hailed as the most progressive ways of funding healthcare for
universal coverage (Doetinchem, 2010; Durairaj & Evans, 2010). Like other countries that
have espoused UHC, Kenya faces immense challenges of trading off and balancing
competing demands as it moves along stages towards realization and sustenance of equity in
healthcare (Carrin et al., 2007). Majority of its population works is the informal sector. This
presents practical difficulties in collecting tax and health insurance contributions due to lack
of institutional capacity to collect taxes (WHO, 2010a; KNBS, 2016). As a result, only
17.1% of Kenyans are covered while only 2.9% of the poorest quintile is covered (MoH,
2014).
Lack of institutional capacity to collect payroll taxes in LIMCs means that majority of those
who work in the informal sector have to pay for health services at the point of use (WHO,
2010a). As a result, high levels OOP push about 1.48 million Kenyans below the poverty line
(Xu et al., 2003; Chuma & Maina, 2012). In Kenya, cross-subsidization necessary for
effective risk pooling is weakened by high levels of disintegration in health financing (WHO,
2010a; Chuma & Okungu, 2011; MoH, 2015).
The study finding revealed that membership contributions are the only source prepayments in
CBHIs. Members pre-pay through both CBHIs and NHIF. In addition, the current study
207
found that majority of CBHIs (88.3%) covered up to 500 households. This hampers their
ability to mobilize adequate resources required for achieving equity in healthcare. This
finding agree with Tabor (2005) & Schieber et al. (2012) that found that CBHIs face a
challenge of mobilizing sufficient resources attributable to their small size and the
contributory capacity of the target population (Tabor, 2005; Schieber et al., 2012). Finding
on the trends of total premiums collected, healthcare cost reimbursements, administration
cost and deficit or surplus in CBHIs between 2010-2015 shows an up and down fluctuation
of the total premiums collected, healthcare cost reimbursements and deficit or surplus while
the administration cost remained constant throughout the entire period. The low premiums
collected in 2015 compared to 2010 collection can be attributed to low renewal rates for the
composite CBHIs and NHIF products following the increase of NHIF rates for the informal
sector.
Subsidization of the poor and exemption of the poorest and socially excluded groups is
critical for reduction of disparities healthcare access and financial risk protection particularly
among these segments of the population. Donor funding channeled through pre-payment and
pooling structures such as CBHIs reduces fragmentation and duplication of Official
Development Assistance (ODA) and other forms of international aid efforts (WHO, 2010a).
The study found that the poor and vulnerable are covered through government and or donor
funding. The funding is however not channeled through CBHIs which means that the finding
is at odds with the recommendations of WHO (2010) on creation of health equity funds that
exempt the poorest and vulnerable group and subsidize the poor. Further, the finding also
disagrees with recent practices from Rwanda and Ghana where health equity funds have
ensured rapid inclusion of the poorest and vulnerable groups (Durairaj et al., 2010; WHO,
2010a).
5.3.3 Risk Pooling and Equity in Healthcare
Risk pooling is one of the fundamental principles of insurance. Long-term viability of a risk
pool is an important factor to attaining equity in healthcare. Financial sustainability is
contingent on the size and cross section of insured risks (WHO, 2010a). The study finding
established that risk pooling in CBHIs had a positive statistically and significant equity in
208
healthcare at the 0.05 level of significance. This implies that there exists higher risk
equalization in CBHIs in Kenya that enable them to spread risks across the members. This
finding agree with the recommendations of Wang & Pielemeier (2012) that larger risk pools
have higher risk equalization mechanisms that enable them to spread risk, lower transaction
cost and offer accurate and stable premiums. The study also found that all the CBHIs studied
are part of network. Carrin et al. (2005) recommends creation of larger risk pools by
encouraging merging of the small pools.
Results from the current study indicate risk pooling in CBHIs and healthcare access have a
statistically significant and weak relationship. This finding concurs with recommendations by
James & Savedoff (2010) that establishment of risk pooling measures enhances healthcare
access particularly to the poor. This result is also in congruence with Ndiaye et al. (2007) &
Chuma et al. (2013) on existence of predominantly small size of risk pool in individual
CBHIs which explains the weak relationship between risk pooling and healthcare access. The
empirical results also revealed that risk pooling had statistically significant but weaker
relationships between the risk pooling and the quality of healthcare. This result concurs with
Spaan et al. (2012) findings on the relationship between health insurance and quality of care
in sub-Saharan Africa. A systematic review by Spaan et al. (2012) found that there was a
weak positive effect of both social health insurance and CBHI on quality of care.
As far as risk pooling in CBHIs and their sustainability is concerned, the study established
that there is a weak positive relationship between the two constructs. Carrin (2011) &
Mebratie et al. (2013) concur that risk transfer mechanisms through reinsurance enhance
viability of CBHIs particularly small pools typical of CBHIs. A study by Private Sector
Innovation Programme for Health (PSP4H) (2014) in Kenya revealed that the limited
resource base from the community impacts CBHIs ability to raise adequate resources. For
that reason the pools are intrinsically small hampering broader risk spreading across the
population. The small size of schemes and resource constraints puts the viability of the
CBHIs at risks (Chen et al, 2012; Mebratie et al., 2013). This explains the significant but
weak relationship between risk pooling and sustainability of CBHIs in Kenya. With regard to
mix of contributions and equity in healthcare, empirical results revealed there is no
209
relationship between the two constructs. The results are contrary to Doetinchem (2010) and
Durairaj & Evans (2010) that posit that hypothecated, general and payroll taxes, insurance or
a combination of two have been hailed as the most progressive way of funding an equitable
healthcare system.
5.3.4 Strategic Purchasing and Equity in Healthcare
The healthcare financing function of purchasing involves a series of decisions which
includes; the active identification of the members health needs, their preferences and values;
the search for the best health services taking into consideration the members needs and
priorities in the health sector; searching for service providers to purchase from taking into
consideration the quality, efficiency and equity as well as determining the best payment
methods and contractual arrangements (WHO 2000; Basaza, et al., 2010). The study found
that strategic purchasing in CBHIs had a positive statistically and significant equity in
healthcare in Kenya. This implies that CBHIs identify members‘ health needs and search for
the most cost effective interventions in realization equity in healthcare. This result is
supported by finding of Baeza et al. (2002) in the ILO study. Baeza et al. (2002) found that
67 CBHIs practiced some form of strategic purchasing with some schemes embracing even
greater roles in strategic purchasing. Lagomarsino & Kundra (2008) and Christian Aid
(2015) recommend that from inception CBHIs should define a benefit package before
employing strategic models in purchasing the services.
The results also indicated that strategic purchasing in CBHIs had a statistically significant
and positive relationship with healthcare access. Hendricks et al. (2011) & Christian Aid
(2015) found that CBHIs in Kwara and Lagos state increase access to a set of health services
through contracts with public and private hospitals. Various previous researches reported
distance as a negative predictor of healthcare access. Geographical access measured using
time and distance required to access care was found to be a barrier to healthcare access in
Rwanda (Schneider & Diop, 2004), Kenya (Chuma & Okungu, 2011), ILO (2012), Ghana,
Burkina Faso, and Mali, Burkina Faso (Parmar et al., 2013). This barrier arises when
patients cannot reach a health facility due to long distance to health facility, huge
transportation charges and lost wages.
210
Quality of care offered to members of an insurance scheme is a critical determinant of health
outcomes and overall patient satisfaction; and in turn it influences access to care.
Disappointments with health services outcomes resulting from comparison between the
original expectations and the reality between the services offered reflect poor quality of care
(Criel & Waelkens, 2003). Result from current research also suggests that strategic
purchasing in CBHIs has a strong positive relationship quality of care. This implies that
strategic purchasing In CBHIs improves the quality of care accessed by CBHIs members.
Previous studies reported that CBHIs improve quality of care by empowering enrollees
through fostering dialogue between communities and health care providers on patient
perceived quality of care (Criel & Waelkens, 2003). Introduction of contractual arrangements
contingent on quality standards was also found to improve quality of care accessed by CBHIs
members (Tipke et al., 2008 & Robyn et al., 2013). A recent systematic review concluded
that there was a weak positive effect of both social health insurance and CBHI on quality of
care in Sub Saharan Africa (Spaan et al., 2012).
The finding from current study also indicates that strategic purchasing in CBHIs and equity
of contributions are not related. The result is at odds with the finding by Munge et al. (2016)
that when strategic purchasing is well designed and executed promotes equity. Further,
research finding established that strategic purchasing in CBHIs has strong positive
relationship sustainability. This result disagrees with the finding of Jacobs et al. (2008) that
most of CBHIs management teams do not negotiate with service providers on price and
quality of health services. As a result this reduces the attractiveness of the schemes,
predisposing them to failure.
5.3.5 Moderating effect of Government Stewardship on Equity in Healthcare
Alvarez-Rosete et al. (2013) describes stewardship as a broader overarching accountability
over the performance of the entire health system and eventually over the health of the whole
population. The distinguishing and conceptually useful facet of stewardship lies in its ability
to allocate ultimate responsibility for the health of the entire population. Beyond the formal
health structures government stewardship is often hypothesized as a critical determinant of
211
successful and sustainable health financing in community based structures such as CBHIs
(Preker & Carrin, 2004).The current study used four measures to estimate government
stewardship. These measures include the government‘s role in design of CBHIs, monitoring
CBHIs related activities, as a trainer and as co-financier in CBHIs.
The results of the present study established that government‘s role in design of CBHIs,
monitoring CBHIs related activities, as a trainer and as co-financier in CBHIs in Kenya in
non-existence. This implies that like most of CBHIs in Africa, CBHIs in Kenya were created
in response to extensive exclusion of certain population categories and have survived
regardless of a vacuum of government stewardship. They continue reducing differentials in
health access and financial risk protection for the excluded groups in Kenya, therefore
playing a critical role of extending equity in healthcare in Kenya. Their penetration level
however remains low with majority of them covering up to 500 households. Lack a right mix
of contributions and sufficient funds to subsidize and exempt the precluded segment of
population results to exclusion of the poorest and vulnerable segment of the community.
This finding is in congruence with Carrin et al. (2005) that views stewardship as critical to
encouraging enrolment across different income categories. Similarly, Mladovsky &
Mossialos (2006) views government stewardship as critical to the success of schemes as a
strategy for achieving its equity and UHC objectives. Pauly et al (2006) who advocates for
minimal government regulation citing an increase of cream skimming and adverse selection
in present of government subsidies.
With regard to the role of the government in design in CBHIs there lacks government
guidance on design issues ranging from minimum enrolment, risk management, community
participation and operating procedures. The finding is contrary to recommendation by Tabor
(2005) and Wang & Pielemeier (2012) on the critical role of government in design of CBHIs,
particularity at the start up and early operational phases. Soors (2010) & WHO (2010)
postulates that it is the duty of the government to define the role of CBHIs in realization of
equity goals within the context of the national health financing policy.
212
Effective implementation of health policies that are meant to promote equity in healthcare
requires not only improvement in design but also monitoring and evaluation. Monitoring
generates results that are useful in modifying ineffective government policies as well as
reinforcing the effective ones. Finding from current research shows that the government does
not track the performance of CBHIs through time. This finding agrees with GIZ (2012)
finding in Nepal. A review of CBHIs in Nepal by GIZ established that supervision and
monitoring mechanisms were non-existence in all the schemes despite having been initiated
by the government. Tabor (2005) argues that it may be impractical for CBHIs to measure that
actual impact particularly on health outcomes due to the high cost of gathering health
performance data from a small group of beneficiaries. This finding however disagrees with
Carrin et al. (2005) views that monitoring enables the government to proactively stimulate
establishment of CBHIs, detect problems in existing CBHIs and offer practical solutions to
the problems. The authors recommend that the government should monitor each CBHIs basic
performance, track progress across various CBHIs over time in addition to carrying out
comparative analysis along the health financing functions.
As far as training is concerned the empirical results do not support Tabor (2005) & Carrin et
al. (2005) recommendation that the government and or donors should build the capacity of
CBHIs management team through provision of basic skills in accounting, management
information systems, setting up of insurance development plans and negotiating and
contracting of providers and other third parties, preparation of organization structures, statues
and regulations as well as monitoring and evaluation. Further, Carrin et al. (2005)
recommend that the outcome from monitoring should be used as an input into the training
activities. Finding on the co-financing role contrasts with WHO report of 2010
recommendation on co-financing being a key priority action for financing healthcare
equitably. Oxfam International (2013) put forward that subsidies increase CBHIs capacity to
reach the poorest of poor, thereby decreasing health inequities.
213
5.4 Conclusions
5.4.1 Enrolment and Equity in Healthcare
The results show that there was a positive and significant relationship between enrolment in
CBHIs and equity in healthcare in Kenya. This implies that the enrolment strategies put in
place by CBHIs stimulate willingness to pay among those excluded were effective in
achievement of equity in healthcare. An effective enrolment strategy in CBHIs should focus
on affordability of premiums, draw their membership from existing mutual groups, allow
members to pay premium when household liquidity is high and exploit social capital that is
intrinsic in target communities. The amount of premium set should be proportionate the
households ability to pay while increase in premium rates results to decrease demand for an
insurance cover. Majority of CBHIs have not met their enrolment targets; this can be
attributed to the voluntary nature of these schemes and lack of subsidies for the poor and
vulnerable groups.
The CBHIs management team also gives members a chance to allocate their income among
the range of product offering. To increase enrolment CBHIs recruit their members from
existing structures in the community namely; households, villages, cooperatives or mutual
benefit societies. Such pre-existing communal networks are rich in social capital whose one
of the components is trust. Social capital is a critical factor when considering CBHIs since its
constituents namely; trust, cooperation and reciprocity facilitate collective action thus
encouraging enrolment. This enables the CBHIs to extend their membership beyond those
who can join the scheme voluntarily. Having a membership that is open to the poor is a
reflection reciprocity norms and social solidarity that have long existed in informal risk
sharing mechanisms. The reciprocity norms inform the values and attitudes of CBHIs
members with regard collective action of address barriers to access to healthcare and sharing
of health care costs. Trust in CBHIs is reinforced by taking into consideration members‘
preferences in setting of benefits packages and premiums. Annual general meetings and other
scheduled meeting offers a platform for members to air their views, complaints and give
feedback. In conclusion, the strengths of CBHIs in stimulating enrolment of precluded
groups‘ lies in their focus on pre-existing social capital and in their ability to stimulate
214
willingness to pre-pay for healthcare through community involvement. By so doing, the
schemes play a critical complementary role of propelling the country towards UHC.
5.4.2 Mix of Contributions and Equity in Healthcare
The negative relationship between mix of contributions and equity in healthcare in CBHIs in
Kenya suggest presence of insufficient amount prepayments and lack of the right
combination of prepayments necessary for realization and sustenance of equity in healthcare
through CBHIs particularly for the excluded segments of the population. The current mix of
contributions in CBHIs is composed of CBHIs and NHIF premium contributions only. The
composite CBHIs and NHIF products covers services that are not covered by the CBHIs
cover. Although the study did not establish existence of any form of direct government and
or donor subsidies through CBHIs, findings shows that poor and vulnerable members of
CBHIs benefit from some form of subsidies from government or donor funding. This can be
explained by the fact that government offers free healthcare at level 2 and 3 facilities that is
funded by HSSF. Although it is fraught with shortcomings, the waiver system also caters for
all or part of healthcare costs in level 4 and 5 facilities. Absence of contributions from
government and or donors to subsidize or exempt the poor and vulnerable through CBHIs
hampers extension of equity in healthcare to these groups. The study then concludes that
current mix of contributions composed of only CBHIs and NHIF the members‘ contributions
is not adequate enough to allow for reallocation of resources for subsidization or exemption
of poor and vulnerable groups. The current mix does not offer an optimal mix of funds
necessary for increased access to care and financial risk protection for precluded groups.
5.4.3 Risk Pooling and Equity in Healthcare
The positive and significant relationship between risk pooling and equity in healthcare
implies that the risk pools in CBHIs in Kenya are significant and come from a cross section
of risks. Despite the existence of small risk pools in each scheme, CBHIs have solidified the
risk pools by encouraging the small pools to merge into bigger ones under a network. The 79
CBHIs that were studied were under 4 networks with the largest network having 67 CBHIs.
The larger risk pools have higher risk equalization mechanisms that enable them to cross-
subsidize, spread risk, lower transaction cost and offer accurate and stable premiums. The
215
varied socioeconomic backgrounds from which the members are drawn from permits
reallocation of pooled funds from wealthier households to poorer households. Encouraging
members to enroll when they are healthy and having a waiting period reduces adverse
selection by ensuring that the scheme does not attract a disproportionate high number of
riskier members. Risk spreading mechanisms within CBHIs in a network offsets risk
variations between CBHIs. Similarly, risk transfer measures through reinsurance enhance
viability of CBHIs particularly small pools atypical of CBHIs. The study concludes that
CBHIs have embraced enhanced risk pooling mechanisms that enhance reallocation of
resources for efficiency and equity gains. The extent of risk pooling is however hampered by
the limited resource base from the community that dictates the amount of premiums.
5.4.4 Strategic Purchasing and Equity in Healthcare
The positive and significant relationship between strategic purchasing and equity in
healthcare implies that CBHIs is responsive to members‘ healthcare needs and searches for
the most cost effective interventions in realization equity in healthcare. CBHI have signed
formal contracts with service providers which imply that the benefits package, quality of care
and payments methods are well spelt out. Contracting accredited service providers acts a gate
keeping measure on issues related to quality of care delivered to CBHIs members. Allocation
of resources based population needs ensures that the neediest members of the CBHIs access
the services they need without exposure to financial ruin. This study concludes that CBHIs
takes an actives role in searching and implementation of purchasing of health services, a
practice that ensures that members access quality care, reallocates resources based on
members needs and responds to the way health services are delivered and in effect it results
to equity in healthcare.
5.4.5 Moderating effect of Government Stewardship on Equity in Healthcare
The finding shows that government stewardship in CBHIs had positive effect on the
relationship between enrolment and strategic purchasing and equity in healthcare in Kenya.
This implies that the government despite the absence of governments role as a steward in the
design, training, monitoring and co-financing of CBHIs other existing regal and regulatory
framework in the insurance industry (particularly in micro-insurance) may be influencing
216
how the health financing functions are executed in CBHIs. This is evidenced by the role
complementary role that CBHIs play in extension of coverage, the networks that they have
established among themselves and with NHIF. The studied CBHIs have enrolled 12,101
households; they have partnered with NHIF to diversify the mix of contributions; they have
employed cost effective methods of purchasing health services and have merged to form
networks for enhanced risk pooling. These practices have increased access to healthcare;
quality of care and have ensured equity in contributions. Absence of government
stewardship has however hampered the size and role played by CBHIs in extending equity in
healthcare. For instance, the sizes of risk pool are small while absence of subsidies and or
exemptions have resulted to exclusion of the poorest and socially excluded segments of the
community. This hampers their efforts of CBHIs in extending the equity goals. Lack of
specific legal and regulatory framework means that their place within the context of the
national health financing policy is not well defined.
5.5 Recommendations
Recommendations are divided into sections; suggestions for improvement and suggestions
for further research. Suggestions for improvement are based study findings. Adoption of
these findings will contribute towards policy debate and dialogue by providing a nuanced
view of CBHIs with an aim clarifying the role played by CBHIs in realization of UHC and
how the interactions between them and the broader health financing system. Suggestions for
further research propose gaps that have not been addressed by the current research and can be
taken up for further research.
5.5.1 Suggestions for Improvements
5.5.1.1 Enrolment and Equity in Healthcare
Based on the number of household enrolled in CBHIs, it is clear that CBHIs contribute
towards reducing differences in healthcare access and enhancing financial risk protection.
This study also found that members are required to pay premium in a single annual payment.
Although single annual payments ease collection premiums, they may reduce enrolment
particularly for households that does not receive lump sum incomes; suggesting that CBHIs
217
management team in consultation with members explore the possibility of spreading
premium payments on a need basis.
5.5.1.2 Mix of Contributions and Equity in Healthcare
Presence of the right mix of contributions for financing healthcare is the responsibility of
government. Mix of contribution in CBHIs had a negative and insignificant influence on
equity in healthcare in Kenya. This means that the proportion and mix of contributions in
CBHIs does not favour their efforts of extending equity in healthcare to excluded segment of
the Kenyan population. Based on the mix of contributions, government and or donors
contributions are conspicuously absent based on the mix of contributions. To increase the
amount of pooled funds in CBHIs, government and policy makers should ensure that
necessary legislation is put in place to allow channeling of some general government
revenue, sin taxes and or donor funding through CBHIs for greater risk equalization. The
funds can be consolidated into an equity fund for subsidizing CBHIs premium for the poor
and exemption of the poorest and vulnerable households, a measure that that will further
enhance equity.
5.5.1.3 Risk Pooling and Equity in Healthcare
The small risk pools in individual CBHIs weaken cross subsidization. To strengthen risk
equalization CBHIs have consolidated into networks. Risk pooling in CBHIs therefore had a
positive statistically and significant influence on equity in healthcare. Government and policy
makers should however enhance broader risk pooling among various health insurers.
Additionally, risk pooling can be improved through promotion of sector wide approach
(SWAp) where all the resources are combined.
5.5.1.4 Strategic Purchasing and Equity in Healthcare
Strategic purchasing in CBHIs had a positive statistically and significant influence on equity
in healthcare in Kenya. However, strategic purchasing in CBHIs and equity of contributions
are not related. This means the optimal performance of this function is hampered by lack of
reallocation of resources and lack greater subsidization in the country due to the vacuum in
government stewardship.
218
5.5.1.5 Moderating influence of Government Stewardship on Equity in Healthcare
Government stewardship in CBHIs had a negative and insignificant influence on equity in
healthcare in Kenya. Governments through the ministry of health have the ultimate
responsibility for ensuring all segments of the population obtain services they need without
suffering financial ruin associated with their utilization. This implies it‘s the government
responsibility to guide the operation of CBHIs as a complementary health financing
mechanism lies in inclusion of the poor and vulnerable groups. This study recommends that
the government should define the place of CBHIs within the context of the national health
financing policy. This will require enacting the necessary legal and regulatory framework to
guide CBHIs administrative and fiscal structures. A clear regulatory framework supporting
pro-poor financing mechanisms can help improve equity in contributions and improve
efficiency and equity through resources reallocation.
5.5.2 Suggestions for Further Research
This research draws attention to the innovative healthcare financing and equity through
CBHIs in Kenya. Inclusion of the precluded population is critical for realization of UHC and
equity in healthcare. An in depth research should be carried out to establish why CBHIs have
failed to meet their enrolment targets despite the existence of strong social capital.
The study found that there was a sharp decline of total premiums collected, healthcare cost
reimbursements, administration cost and deficit or surplus in 2013. Further research should
therefore aim at establishing the reason for the sharp decline in 2013. Efficiency in execution
of the health financing functions is critical for preserving the meagre resources for realization
of equity goals. A similar study using different data method such as data envelopment
analysis can be used to measure how efficiently each health financing function is carried for
realization of equity in healthcare. Further, a study on the progressive effect of NHIF
coverage on renewal rates in CBHIs would offer insight on the expected effects on CBHIs as
NHIF expands its benefits beyond those offered by CBHIs.
219
REFERENCES
Adebayo, E. F., Uthman, O.A., Wiysonge, C.S., Stern, E, A., Lamont, K.T., & Ataguba, J.E.
(2015). A systematic review of factors that affect uptake of community-based health
insurance in low-income and middleincome countries. BMC Health Services Research ,
15:543.
Aggarwal, A. (2010). Impact evaluation of India‘s ‗Yeshasvini‘ community-based health
insurance programme. Health Economics , 19 (1), 5-35.
Ahmed, S. R., & Mesbah, F. S. (2015). Catastrophic and Impoverishing Effects of Out-of-
Pocket Health Expenditure: New Evidence from Egypt. American Journal of
Economics , 5(5): 526-533.
Ahmed, S., Hoque, M.E., Sarker, A.R., Sultana, M., Islam, Z., Gazi, R., & Khan, J.A.M.
(2016). Willingness-to-Pay for Community-Based Health Insurance among Informal
Workers in Urban Bangladesh. PLoS ONE , 11(2).
Akazili , J., Gyapong, J., & McIntyre, D. (2011). Who pays for health care in Ghana? . Int J
Equity Health , 10:1–13.
Alesina, A., & La Ferrara, E. (2002). Who trusts others? Journal of Public Economics ,
85(2), 207–234.
Alvarez-Rosete, A ., Hawkins, B., & Parkhurst, J. (2013). Health System Stewardship and
Evidence Informed Health Policy. London School of Hygiene and Tropical Medicine.
Alvarez-Rosete, A. (2008). The role of the state in public health. In: Heggenhougen K
(ed).International Encyclopedia of Public Health , 211-18.
Anderson, V. (2004). Research Methods in Human Resource Management . London: CIPD
Andriani, L. (2013). Social Capital: a Road Map of Theoretical Frameworks and
Empirical Limitations. London: Department of Management Birkbeck University of
London. ISSN 2052-0581.
Aradhna, A. (2011). Achieving Equity in Health through Community-Based Health
Insurance: India's Experience with a Large Cbhi Programme. The Journal of
Development Studies , 47.11: 1657.
Armstrong, J. L. (1997). Stewardship and public service. Ottawa: Canadian Public Service
Commission. Discussion paper.
220
Aryeetey, G., Jehu-Appiah, C., Agyepong, I., Spaan, E., & Baltussen, R. (2012). Household
perceptions and their implications for enrolment in the national health insurance
scheme in Ghana. Health Policy and Planning , 17(1), pp.43-51.
Asadul, I., & Pushkar, M. (2012). Health Shocks and Consumption Smoothing in Rural
Households: Does Microcredit Have a Role to Play? Journal of Development
Economics , 97.2: 232-43.
Asante, A., Price, J., Hayen, A., Jan, S., & Wiseman, V. (2016). Equity in Health Care
Financing in Low- and Middle-Income Countries: A Systematic Review of Evidence
from Studies Using Benefit and Financing Incidence Analyses. PLOS ONE , 10.1371.
Asfaw, A., & von Braun, J. (2004). Can community health insurance schemes shield the poor
against the downside health effects of economic reforms? The case of rural Ethiopia.
Health Policy , 70: p. 97-108.
Ataguba, J.E. (2008). Community Health Insurance Scheme as a viable option for rural
population in Nigeria. Oxford: University of Oxford. p. 1–11.
Ataguba, J.E., Ichoku, H.E., & Fonta, W. M. (2008). Estimating the willingness to pay for
community healthcare insurance in rural Nigeria. Laval: Poverty and Economic Policy
Research Network.
Atim, C. (2000). Contribution financière des mutuelles de santé au financement, à la
fourniture et à l’accès aux soins de santé : Synthèse de travaux de recherche menés
dans neuf pays d’Afrique de l’Ouest et du Centre. STEP: Bureau International du
Travail.
Atim, C. (1998). Contribution of Mutual Health Organizations to Financing, Delivery, and
Access to Health Care : Synthesis of Research in Nine West and Central African
Countries, PHR Technical Report No. 18.
Atim, C. (2011). Ghana NHIS Design and Experience Lessons for Nigeria with special
reference to funding sources of financing.
Babatunde, R. O., Oyedeji, O. A., Omoniwa, , A. E., & Adenuga, A. H . (2016). Willingness-
to-pay for community based health insurance by farming households: a case study of
Hygeia community health plan in Kwara state, Nigeria . Trakia Journal of Sciences ,
No 3, pp 281-286.
221
Babin, B., & Svensson, G. (2012). Structural equation modelling in social science research:
Issues of validity and reliability in the research process. European Business Review ,
24(4), 320-330.
Baeza, C., Montenegro, F., & Núñez, M. (2002). Extending Social Protection in Health
through Community Based Health Organizations: Evidence and Challenges.”
Discussion Paper. Universitas Program—Strategies and Tools against Social
Exclusion and Poverty (STEP). Geneva: International Labor Organization (ILO).
Baker, L.C., Bundorf, M. K., & Kessler, D. P. (2014). Vertical Intergration: Hospital
Owership of physician practices is associated with higher prices and spending. Health
Affairs , 33 (5), 756-763.
Baltussen, R., & Ye, Y. (2006). Quality of care of modern health services as perceived by
users and non-users in Burkina Faso. Int J Qual Health Care , 18:30-34.
Barnum, H., Kutzin, J., & Saxenian, H. (1995). Incentive and Provider Payment Methods.
International Journal of Health Planning and Management , 10: 23–45.
Basaza, R., Criel, B., Van der Stuyft, P., Basaza, R., Criel, B., & Van der Stuyft, P. (2007).
Low enrollment in Ugandan Community Health Insurance schemes: underlying causes
and policy implications. BMC Health Services Research , 2007. 7: p. 105.
Belsley, D. A. (2004). Regression Diagnostics: Identifying Influential Data and Sources of
Collinearity, 1st ed. New York: John Willey & Sons, Inc.
Bennett, S. (2004). The role of community-based health insurance within the health care
financing system. Health Policy and Planning , 19, 147-158.
Bennett, S., Creese, A., & Monash, R. (1998). Health Insurance Schemes for Poeple outside
Formal Sector Employment. ARA Paper No.16. Geneva: World Health Organization.
Berwick, D. M. (2004). Lessons from developing nations on improving healthcare. BMJ ,
328:1124-1129.
Best, J. W., & Khan, J. V. (2007). Research in Education. New Delhi: Prentice Hall.
Bidgood, E. (2013). Health Systems: Germany. Civitas.
Bidgood, E. (2013). Healthcare Systems: France. Civitas.
Bird, R. M. (2015). Tobacco and Alcohol Excise Taxes for Improving Public Health and
Revenue Outcomes: Marrying Sin and Virtue? Policy Research Working Paper 7500.
Washington DC: World Bank.
222
Bitrán, R., & Giedion, U. (2003). Waivers and exemptions for health services in developing
countries. World Bank: Social Protection Discussion Paper Series; No. 0308.
Blaikie, N. (1993). Approaches to Social Enquiry (1st ed.). Cambridge: Polity Press
.Bourdieu, P. (1986). The Forms of Capital pp. 241-258 in Handbook of Theory and
Research for the Sociology of Education, edited by J.G. Richardson. Westport, CT:
Greenwood Press.
Bowman, W. (2011 ). Financial Capacity and Sustainability of Ordinary Nonprofits.
Nonprofit Management and Leadership , Vol. 22, No. 1, Fall 2011, pp. 37–51.
Brett, W., Ted, B., & Andrys, O. (2010). Explatory Factor analysis: A five step guide for
novices. Journal of Emergency Primary Health Care , 8 (3).
Brinkerhoff, D. (2003). Accountability and Health Systems: Overview, Framework, and
Strategies. Bethesda, MD: The Partners for Health Reformplus Project, Abt Associates
Inc.
Broomberg, J., & Price, M. (1990). The Impact of the Fee-for-Service Reimbursement
System on the Utilization of Health Services. South Africa Medical Journal , 78(4):
130–32.
Buigut, S., Ettarh, R., & Amendah, D.D. (2015). Catastrophic health expenditure and its
determinants in Kenya slum communities. International Journal for Equity in Health 5,
14:46.
Camp, W. G. (2001). Formulating and evaluating theoretical frameworks for career and
technical education research. Journal of Vocational Educational Research , 26(1), 27-
39.
Carrin G., Desmet, M., & Basaza, R. (2001). Social health insurance development in low-
income developing countries: new roles for government and non-profit health
insurance organizations. Ch.10 in Scheil-Adlung X. Building Social Security: The
challenge of privatization . London: Transaction Publishers.
Carrin, G. (2003). Community based Health Insurance Schemes in Developing Countries:
facts, problems and perspectives. Geneva: WHO.
Carrin, G. (2011). Health Financing for the Developing World: Supporting Countries' Search
for Viable Systems. ASP Pub Press.
223
Carrin, G., et al. (2007). Health financing reform in Kenya – assessing the social health
insurance proposal. S Afr Med J 2007 , 97: 130-135.
Carrin, G., Mathauer, I., Xu, K., & Evans, D. B . (2008). Universal coverage of health
services: tailoring its implementation. Bull World Health Organ , 86:857-63.
Carrin, G., Ron, A., & Yang, H et al. (1999). The reform of the rural coooperative medical
system in the People's Republic of China: Interim experience in 14 pilot counties.
Social Science and Medicine , 48:961-972.
Carrin, G., Waelkens, M., & Criel, B. (2005). Community-based health insurance in
developing countries: a study of its contribution to the performance of health financing
syste. Tropical Medicine and International Health, volume 10 no 8 pp 799–811.
Catherine, R., & Salmen, L. (2000). Understanding Social Capital Agricultural Extension in
Mali: Trust and Social Cohesion. Washington, D.C: The World Bank.
Central Bank of Kenya & Financial Services Deepening Kenya. (2009). FinAccess National
Survey 2009: Dynamics of Kenya's changing financial landscape. Niarobi: FSD Kenya.
Chandran, E. (2004). Research Methods: A quantitaive Approach . Nairobi : Daystar
University.
Chankova, S., Sulzbach, S., & Diop, F. (2008). Impact of mutual health organizations:
evidence from West Africa. Health Policy Plan , 23(4): p. 264-276.
Chatham House. (2014). Shared Responsibilities for Health A Coherent Global Framework
for Health Financing. Final Report of the Centre on Global Health Security Working
Group on Health Financing: Chatham House Report.
Chen, K. Z., Liu, Y., Hill, R., Xiao, C., & Liu, L. (2012). Can we relax liquidity constraints
on microinsurance demand? Piloting an innovative design for agricultural insurance.
Technical report, Microinsurance Innovation Facility. Geneva: ILO.
Chen, K., Daukste, R., Przybyl, A., & Fechter, N. (2012). Trust, Community-Based Health
Insurance and Enrollment Rates. Financing Practicum, pp. 1-32.
Chia, R. (2002). The Production of Management Knowledge: Philosophical Underpinnings
of Research Design", in Partington, D. (ed.). Essential Skills for Management Research,
1st ed. SAGE Publications Ltd., London , pp. 1-45.
Christian Aid . (2015). A Review of Community-Based Health Insurance Schemes: Lessons
from Nigeria and Ghana. Abuja, Nigeria : Christian Aid.
224
Chuma, J., & Maina, T. (2012). Catastrophic health care spending and impoverishment in
Kenya . BMC Health Services Research , 12: 413.
Chuma, J., & Okungu, V. (2011). Viewing Kenyan health system through an equity lens:
Implications for Universal Health Coverage. International Journal for Equity in Health
, 10:22.
Chuma, J., Mulupi, S.,& McIntyre, D. (2013). Providing Financial Protection and Funding
Health Service Benefits for the Informal Sector: Evidence from sub-Saharan Africa.
Working Paper 2. RESYST.
Churchill, C. (2006). What is insurance for the poor? in C. Churchill (ed.), Protecting the
poor, A microinsurance compendium. Geneva: ILO.
Coleman, S. J. (1988). Social Capital in the Creation of Human Capital. American Journal of
Sociology , Vol. 94 pp. S95-S120.
Collins, D., Quick, J. D., Musau, S. N., Kraushaar, K., & Hussein, I. M. (1996). The fall and
rise of cost sharing in Kenya: the impact of phased implementation. Health Policy Plan
, 11:52–63.
Collis, J., & Hussey, R. (2009). Business Research: A Practical Guide for Undergraduate &
Postgraduate Students. 3rd ed. London: Palgrave Macmillan.
Cooper, D.R.,& Schindler, P.S. (2014). Business Research Methods (12th edn). Boston, MA:
McGraw-Hill Higher Education.
Costa-Font, J., Gemmill. M.,& Rubert, G. (2009). Revisiting the health care luxury goods
hypothesis: aggregation, precision, and publication biases? Health, Econometrics and
Data Group (HEDG) Working Papers 09/02. Department of economics, University of
York.
Creese, A., & Bennett, S. (1997). Rural Risk-Sharing Strategies. In: Schieber, G.(ed),
Innovations in Health Care Financing Proceedings of a World Bank Conference,
March 10-11. Wasnington, DC : World Bank.
Creswell, J. W. (2014). Research Design: Qualitative, Quantitative and Mixed Methods. Los
Angeles, CA: SAGE Publications Inc.
Creswell, J.W. (2009). Research Design: Qualitative, Quantitative, and Mixed Methods
Approaches, 3rd edn. Thousand Oaks, California: Sage Publications.
225
Criel B., Noumou Barry, A., & von Roenne, F. (2002). Le Projet PRIMA en Guinée
Conkary-une expérience d’organisation de mutuelles de santé en Afrique rurale.
Brussels: Medicus Mundi Belgium.
Criel, B. (1998). District-based health insurance in sub-Saharan Africa, Part II : Case-
studies. Anvers : Studies in Health Services Organisation & Policy.
Criel, B., & Van Dormael, M. (1999). Mutual health organizations in Africa and social health
insurance systems: Will European history repeat itself? Tropical Medicine and
International Health , 4, 155–159.
Criel, B., Van Dormael, M., Lefevre, P., Menase, U. & Van Lerberghe, W. (1998). Voluntary
Health Insurance in Bwamanda, Democratic Republic of Congo. An exploration of its
meaning to the community. Tropical Medicine and International Health, vol.3, nr.8,
pp.640-653.
Criel, B., & Waelkens, M.P. (2003). Declining subscriptions to the Maliando Mutual Health
Organisation in Guinea-Conakry (West Africa): what is wrong? Social Science
Medicine , 57 (7), 1205-1219.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests . Psychometrika,
16, 297-334.
Davies, P., & Carrin, G. (2001). Risk Pooling - Necessary but not sufficient . Bulletin of the
World Health Organization , 79 (7).
De Allegri, M., Sanon, M., & Sauerborn, R. (2006b). To enrol or not enrol?: A qualitative
investigation of demand for health insurance in rural West Africa. Social Science
Medicine , 62 (6), 1520-1527.
De Allegri, M., Sauerborn, R., Kouyaté, B., & Flessa, S. (2009). Community health
insurance in sub-Saharan Africa: what operational difficulties hamper its successful
development? . Tropical Medicine and International Health , 14(5), 586-596.
Debebe, Z. Y., Kempen, L. V., & Hoop, T. D. (2012). A perverse' net' effect? Health
insurance and ex-ante moral hazard in Ghana. The Hague: Institute of social studies.
Defourny, J., & Failon, J. (2008). Community-Based Health Insurance Schemesin Sub-
Saharan Africa: Which Factors Really Influence Enrolment? Working Paper. Centre
for Social Economy.
226
Dercon, S., Gunning, J. W., Zeitlin, A., Cerrone, C., & Lombardin, S. (2012). ‘Health
insurance participation: Experimental evidence from Kenya. ILO Microinsurance
Innovation Facility Research Paper No. 10. Geneva: ILO.
DeRock, D., Knowles, J., Wittenberg, T., Raney, L., & Cordova, P. (1996). Rural health
services at Seguridad Social Campesino facilities: Analyses of facility and households
surveys. Technical No. 13. Bethesda, MD. Health financning and Sustainability Project,
Abt Associates Inc.
Desmet, M., Chowdhury, A.Q., & Islam, M.K. (1999). The potential for social mobilisation
in Bangladesh: the organisation and functioning of two health insurance schemes.
Social Science & Medicine , 48(7): p. 925-938.
Deutsche Gesellschaft für Internationale Zusammenarbeit (GIT). (2012). Review of
Community-based Health Insurance Initiatives in Nepal. Health Sector Support
Programme.
Dhillon, R.S., Bonds, M.H., Fraden, M., Ndahiro, D.,& Ruxin, J. (2012). The impact of
reducing financial barriers on utilization of primary health care facility in Rwanda.
Glob Public Health , 7(1): 71-86.
Dillman, D. (2007). Mail and internet surveys: The tailored design method (2nd ed.). New
York: Wiley.
Diop, F. (2005). Determinants of Financial Stability of Mutual Health Organizations in the
Thies Region of Senegal: Household Survey Component. Bethesda, MD: Abt
Associates Inc.
Diop, P.F., Sulzbach, S., & Chankova, S. (2006). The Impact of Mutual Health
Organizations on Social Inclusion, Access to Health Care, and Household Income
Protection: Evidence from Ghana,Senegal, and Mali. Bethesda, MD; The Partners for
Health Reform plus Project. Abt Associates,Inc.
Doetinchem, O. (2010). Hypothecation of tax revenue for health. World Health Report
(2010). Geneva: WHO.
Doherty, J., Kirigia, D., Ichoku, H., McIntyre, D., Hanson, K., & Chuma, J. (2014).
Increasing the fiscal space for health: The experience of Kenya, Nigeria and Lagos
State (Nigeria). RESYST.
227
Donaldson, L. (1985). In Defence of Organization Theory, A Reply to the Critics. Cambridge:
Cambridge University Press.
Donaldson, L., & Davis, J. H. (1991). Stewardship Theory or Agency Theory: CEO
governance and shareholder returns. Australian Journal of Management , 16: 49-65.
Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts,
evidence, and implications. Academy of Management Review , 20, 65–91.
Donfouet, H. P. P., Mahieu, P., & Malin, E. (2013). Using respondents‘ uncertainty scores to
mitigate hypothetical bias in community-based health insurance studies. Eur J Health
Econ , 14:277–85.
Donfouet, H. P. P., Makaudze, E. M., Mahieu P. A., & Malin, E. (2011). Willingness to pay
for a community-based prepayment scheme in rural Cameroon. Int J Health Care
Finance Econ , 11:209–20.
Donfouet, H.P., & Mahieu, P. (2012). Community-based health insurance and social capital:
a review. Health Economic Review , Volume 2, Number 1, Page 1.
Dong, H., De Allegri, M., Gnawali, D., Souares, A., & Sauerborn, R. (2009). Dropout
analysis of community-based health insurance membership at Nouna, Burkina Faso.
Health Policy , Vol. 92, pp. 174–179.
Dong, H., Kouyate, B., Cairns, J., & Sauerborn, R. (2004). Differential willingness of
household heads to pay community-based health insurance premia for themselves and
other household members. Health Policy Plan , 19:120–6.
Drechsler, D., & Jütting, J. P. (2010). Six Regions, one story. In Global marketplace for
private health insurance: strength in numbers. Edited by Preker, A. S., Sweifel, P., &
Schellekens, O. P. Washington DC: World Bank.
Dror, D., & Jacquier, C. (1999). Micro-insurance: extending health insurance to the
excluded. International Social Security Review , Vol. 52. 1/99.
Dror, D. M., Radermacher, R., & Koren, R. (2007). Willingness to Pay for Health Insurance
Among Rural and Poor Persons: Field Evidence from Seven Micro Health Insurance
Units in India. Health Policy , (82):1.
Dror, I. (2007). Demystifying Micro Health Insurance Package Design - Choosing
Healthplans All Together (Chat). Microfinance Insights , 4:17-19.
228
Dror, D. M., & Armstrong, J. (2006). Do micro health insurance units need capital or
reinsurance? A simulated exercise to examine different alternatives . Geneva : The
International Association for the Study of Insurance Economics.
Durairaj, V., & Evans, D. B. (2010). Fiscal space for health in resource-poor countries.
World Health Report (2010) Background Paper, 41. Geneva: WHO.
Durairaj, V., D‘Almeida, S., & Kirigia, J. (2010). Ghana’s Approach to Social Health
Protection. Background paper for the 2010 World Health Report. GENEVA : WHO.
Dutta, A., & Hongoro, C. (2013). Scaling Up National Health Insurance in Nigeria:
Learning from Case Studies of India, Colombia, and Thailand. Washington, D.C:
Futures Group, Health Policy Project.
Eriksson, M. (2011). Social capital and health—Implications for health promotion. Global
Health Action , 4: 5611.
Fairbank, A. (2003). Sources of Financial Instability of Community-Based Health Insurance
Schemes: How Could Social Reinsurance Help? Technical Report No. 024. Bethesda,
MD: The Partners for Health Reformplus Project, Abt Associates Inc.
Fay, M. (2005). The Urban Poor in Latin America. Washington. D.C : World Bank.
Fernandes, A et al. (2009). Health financing systems review of Rwanda- options for universal
coverage. World Health Organization and Ministry of Health, Republic of Rwanda.
Field, J. (2008). Social Capital. New York: Routledge.
Fornell, C., & Bookstein, F. L. (1982). Two structural equation models: LISREL and PLS
applied to consumer exit-voice theory. Journal of Marketing Research , 19, 440-452.
Franco, L.M., Diop, F.P., Burgert, C., Kelley, A.G., Makinen, M., & Siampara, C.H.T.
(2008). Effects of mutual health organizations on use of priority health care services in
urban and rural Mali: A case-control study. in Bulletin of the World Health
Organization , Vol. 86, No.11, pp. 830–838.
Fricker, S., Kreisler, C., & Tan, L. (2012). An exploration of the application of PLS path
modeling approach to creating a summary index of respondent burden. JSM , Section
on Survey Research Methods.
Fronstin, P. (2008). Sources of Health Insurance and Characteristics of the Uninsured:
Analysis of the March 2008 Current Population Survey. ERBI issue Brief , No. 321.
229
Fuchs, V. R. (1996). What every philosopher should know about health economics.
Proceedings of the American Philosophical Society , 140(2): 186-96.
Galárraga, O., Saso-Rubí, S. G., Salinas- Rodríguez, A., & Sesma-Vázquez, S. (2010).
Health insurance for the poor: impact on catastrophic and out-of-pocket health
expenditures in Mexico. The European Journal of Health Economics , 11(5), 437-447.
Gefen, D., Rigdon, E. E., & Straub, D. W. (2011). Editor's Comment: An update and
Extension to SEM Guidelines for Administrative and Social Science Research . MIS
Quartely , 35 (2), PP. iii -xiv.
Giedion, U., Alfonso, E. A., & Díaz, Y. (2013). The impact of universal coverage schemes in
the developing world: A review of existing evidence . Washing DC: World Bank.
Gill, J., & Johnson, P . (2002). Research Methods for Managers (3rd edn). London: Sage.
Gilson, L. (2003). Trust and development of health care as a social institution. Social Science
and Medicine , 56:1453-1468.
Gilson, L., Kalyalya, D., Kuchler, F., Lake, S., Oranga, H., & Ouendo, M. (2001). Strategies
for promoting equity: experience with community financing in three African countries.
Health Policy, 58: 37-67.
Gnawali, D. P., Pokhrel, S., Sanon, M., De Allegri, M., Dong, H., Sauerborn, R. (2009). The
effect of community-based health insurance on the utilization of modern health
facilities: evidence from Burkina Faso. Health Policy , 90 (2-3): 214-22.
Goudge, J., Akazili, J., Ataguba, J., Kuwawenaruwa, A., Borghi, J., Harris, B., & Mills, A.
(2012). Social solidarity and willingness to tolerate risk- and income-related cross-
subsidies within health insurance: experiences from Ghana, Tanzania and South
Africa. Wits: Centre for Health Policy, School of Public Health, University of the
Witwatersrand.
Government of Kenya. (2010). 2009 Kenya Population and Housing Census, Volume II.
Nairobi: Ministry of State for Planning, National Development and Vision 2030; and
Kenya National Bureau of Statistics.
Grant, C., & Osanloo, A. (2014). Understanding, selecting, and integrating a theoretical
framework in dissertation research: Creating the blueprint for your ―house‖.
Administrative issues journal: connecting education, practice, and research, Volume 4,
issue 2.
230
Gustafsson-Wright, E., & Schellekens, O. (2013). Achieving universal health coverage in
Nigeria one state at a time: A public private partnership community-based health
insurance model. Working Paper 2. Brooke Shearer Working Paper Series.
Hafner, T., & Shiffman, J. (2012). The emergence of global attention to health systems
strengthening. Health Policy and Planning, 28: 41-50.
Hair, J. F. Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate Data
Analysis (5th Ed.). Upper Saddle River, NJ: Prentice Hall.
Hair, J. F., Black, W. C., Babin. B. J., & Anderson, R. E . (2010). Multivariate Data Analysis
(5th Ed.). New Jersey: Pearson Prentice Hall.
Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). Pls: indeed a silver bullet. Journal of
Marketing Theory and Practice , 19(2), 139-151.
Hair, Jr., J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006).
Multivariate Data Analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.
Heller, P. (2006). The Prospect of Creating ‗Fiscal Space‘ for the Health Sector. Health
Policy and Planning, 21(2): 75-79.
Hendricks, M., & et.al. (2011). Cardivascular disease prevention in rural Nigeria in the
context of a community based health insurance scheme: Quality improvement
Cardiovascular care Kwara-I(QUICK-I). BMC Public Health , 11(86), 1471-2458.
Henseler, J., Ringle, C., & Sinkovics, R. (2009). The Use of Partial Least Squares Path
Modeling in International Marketing. Advances in International Marketing , 20, 277-
319.
Hill, C. W. L,. & Jones, T. M. (1992). Stakeholder-agency theory. The Journal of
Management Studies , 29 (2), 131–155.
Honda, A., McIntyre, D., Hanson, K., & Tangcharoensathien, V . Eds. (2016). Strategic
Purchasing in China, Indonesia and the Philippines. Geneva: WHO.
Hsiao, W. C. (2004). Experience of community health financing in the Asian region. In A.
S.Preker & G. Carrin (Eds.), Health financing for poor people: resource mobilisation
and risk sharing (pp.119-155). Washington: World Bank.
Hsiao, W.C . (2001). Unmet health needs of two billion: Is community financing a solution?
Preliminary draft for the Commision on Macroeconomics and Health. WHO.
231
Hsiao, W.C., & Shaw, R. P. (2007). Social Health Insurance for developing countries.
Washing DC: World Bank.
Huzuma, J. (2011). Coexistence of performance based financing (PBF) and Community
Based Health Insurance (CBHI): Rwanda experience.
ILO & STEP - GTZ. (2006). The role of micro- insurance as a tool to face risks in the
context of social protection. Geneva: ILO/STEP - GTZ.
ILO. (2012). Protecting the poor: A microinsurance compendium, Volume II. Craig, C &
Matul, M (eds). Geneva: Munich Re Foundation,Micro Insurance Network, ILO.
ILO. (2014). Universal Health Protection: Progress to date and the way forward. Paper 10.
geneva: ILO, Social Protection Department.
International Labour Organisation. (2013). Community-based health insurance/ health
microinsurance premiums: What are the advantages of CBHI schemes? Geneva: ILO.
International Monetary Fund. (2012). Regional Economic Outlook: Sub-Saharan Africa.
Sustaining growth amid global uncertainty . Washington DC.
Jacobs, B., Bigdeli, M., van Pelt, M., Ir, P., Salze, C., & Criel, B. (2008). Bridging
community-based health insurance and social protection for health care – a step in the
direction of universal coverage? Tropical Medicine and International Health , 13(2),
140-143.
Jacobs, B., Price, N. L., & Oeun, S. (2007). Do exemptions from user fees mean free access
to health services? A case stu- dy from a rural Cambodian hospital. Tropical medicine
& international health , 12:1391-1401.
Jakab, M., & Chrishnan, C. (2001). Community involvement in Health Care Financing: A
survey of literature on impacts strengths and weaknesses. Washington DC: World
Bank.
James, C., & Savedoff, W. (2010). Risk pooling and redistribution in health care: an
empirical analysis of attitudes towards solidarity. World health report 2010
background paper, no. 5. Geneva: WHO.
Jonker, J., & Pennink, B. (2010). The essence of research methodology: a concise guide for
master and phD students in management science. London: Springer Science and
Business Media.
232
Jost, T.S . (2008). Health care access in the United States: conflicting concepts of justice and
little solidiarity. Medical Law , 27:605-16.
Jütting, J. (2000). Do mutual health insurance schemes improve theh access to healthcare?
Preliminary results from household survey in rural Senegal. International conference:
Health systems financing in low-income African and Asian Countries.
Jütting, J. P. (2004). Do community based health insurance schemes improve poor people's
access to health care? Evidence from Rural Senegal. World Development , 3 (2); 273-
288.
Jütting, J.P. (2004). Do Community-based Health Insurance Schemes Improve Poor People‘s
Access to Health Care? Evidence From Rural Senegal. World Development , 32(2): p.
273-288.
Kalk, A., Groos, N., Karasi, J. C., & Girrbach, E . (2010). Health systems strengthening
through insurance subsidies: the GFATM experience in Rwanda. Tropical Medicine
and International Health , 15(1):94-7.
Kamau, A., & Holst, J. (2008). National Hospital Insurance Fund: An evealuation Report.
Nairobi: MoH/GTZ Health Sector Programme Kenya.
Kenny, D.A. (2012). Measuring Model Fit. Retrieved electronically November 5:
http://www.davidakenny.net/cm/fit.htm, 2016.
Kenya national bureau of statistics. (2010). 2009 Census Report.
Kimani, J. K., Ettarh, R., Kyobutun, C., Mberu, B., & Kanyiva, K. (2012). Determinants for
participation in a public health insurance among residents of urban slums in Nairobi,
Kenya: Results of a cross-sectional survey. BMC Health Services Research , 12(66),
12-66.
Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd Edition).
New York: The Guilford Press.
KNBS. (2010). Economic Survey 2010.
KNBS. (2016). Economic Survey 2016.
KNBS. (2007). Kenya Household Integrated Budget Survey 2005/06.
Kothari, C. R. & Garg, G. (2014). Research Methodology: Methods and Techniques (3rd ed).
New Delhi: New Age International Publishers.
233
Krishna, A., & Shrader, E. (1999). Social Capital Assessment Tool. Conference on Social
Capital and Poverty Reduction. Washington DC: The World Bank.
Kruk, M.E., Goldmann, E., & Galea, S. (2009). Borrowing and Selling to Pay for Health
Care in Low- and Middle-Income Countries. Health Affairs , 28:1056-66.
Kumar, R. (2011). Research Methodology: A Step by Step Guide for Begginers . New Delhi:
Sage Publications.
Kutzin, J. (2001). A descriptive framework for country-level analysis of health care financing
arrangements. Health Policy , 56:171-204.
Kutzin, J. (2008). Health Financing Policy: A Guide for Decision-makers. Health Financing
Policy Papers.
Kutzin, J. (2010). Reforms in the pooling of funds . In C. C. Joseph Kutzin, Implementing
Health Financing Reform, Lessons from countries in transition. Copenhagen, Denmark:
WHO: European Observatory on Health Systems and Policies.
Lagomarsino, G., & Kundra, S. S. (2008). Risk Pooling: Challenges and Opportunities.
Technical partner paper 6. The Rockefeller Foundation.
Lagomarsino, G., Garabrant, A., Adyas, A., Muga, E., & Otoo, N . (2012). Moving towards
universal health coverage: health insurance reforms in nine developing countries in
Africa and Asia. Lancet , 380: 933-43.
Langenbrunner, J., & Somanathan, A. (2011). Financing Health Care in East Asia and
Pacific. Washington, DC: World Bank.
Lavrakas, P. J. ed. (2008). Encyclopedia of Survey Research Methods. California: Sage
Publications.
Leftley, R. (2005). Technical assistance for the promotion of microinsurance: The
experience of Opportunity International, CGAP Working Group on Microinsurance,
Good and Bad Practices Case Study, No. 11. Geneva: ILO.
LeRoy, P., & Holtz, J. (2012). Third-party payment mechanisms in health microinsurance. In
C. Churchill and M. Matul (Eds.), protecting the poor: a microinsurance compendium
volume II (pp. 132-155). Geneva : International Labour Organisation (ILO).
Levesque, J., Harris, M.F., & Russell, G. (2013). Patient-centred access to health care:
conceptualising access at the interface of health systems and populations. International
Journal for Equity in Health , 12:18.
234
Lin, N. (2001). Social Capital: A theory of social structure and action. Cambridge , UK:
Cambridge.
Litwin, M. S . (1995). How to measure survey reliability and validity: The Survey
Kit(Volume 7). Thousand Oaks, CA: Sage Publications.
Luse, A., Mennecke, B., & Townsend, A. (2012). Selecting a research topic: A framework
for doctoral students. International Journal of Doctoral Studies , 7: 143-152.
Lysaght, Z. (2011). Epistemological and paradigramatic ecumenism in 'Pasteur's Quadrant.'
Tales from doctoral research. Official conference proceedings of the third Asian
Conference on Education in Osaka, Japan. Osaka.
Makaka, A. (2012). Role of government stewardship and leadership in SHP. Regional
Conference on Social Health Protection in the East African Community, Kigali. Kigali.
Maleika, M., & Kuriakose, A.T. (2008). Microinsurance: Extending Pro-Poor Risk
Management through the Social Fund Platform. Washington, DC: World Bank.
Marsh, D., & Sharman, J. C . (2009). Policy Diffusion and Policy Transfer. Policy Studies ,
30 (3): 269–288.
Martz, E. (2013). Enough is Enough! Handling Multicollinearity in Regression Analysis.
Mathauer, I., & Carrin, G. (2010). The role of institutional design and organizational
practice for health financing performance and universal coverage. Discussion Paper
No. 5. Geneva: WHO.
Mathauer, I., Schmidt, J., & Wenyaa, M. (2007). Extending social health insurance to the
informal sector in Kenya. An assessment of factors affecting demand. The International
Journal of Health Planning and Management , 23(1): 51-68.
Matul, M., Tatin-Jaleran, C., & Kelly, E. (2011). Improving client value from
microinsurance: Insights from India, Kenya, and the Philippines. Microinsurance
Paper No. 12, Microinsurance Innovation Facility. Geneva: ILO.
McCord, M. J., Steinmann, R., Tatin-Jaleran, C., Ingram, M., & Mateo, M. (2012). Making
finance work for Africa: the landscape of microinsurance in africa 2012. Munich Re
Foundation, GIZ-Program Promoting Financial Sector Dialogue in Africa, African
Development Bank Group, Microinsurance Network and Microinsurance Innovation
Facility: ILO.
235
McCord, M., Osinde, S. (2003). Reducing Vulnerability: The supply of Health Insurance in
East Africa. A synthesis Report. Nairobi: MicroSave Africa.
McIntyre, D., Thiede, M., Dahlgren, G., & Whitehead. (2006). What are the economic
consequences for households of illness and paying for health care in low- and
middleincome country contexts? Social Science & Medicine , 62(4), 858-865.
McIntyre, D.,& Meheus, F. (2014). Fiscal Space for Domestic Funding of Health and Other
Social Services. Working Group on Financing Paper 5. London: CHATHAM HOUSE.
McPake B. (1993). Community financing of health care in Africa: An eveluation of the
Bamako Initiative.
Mebratie, A. D., Sparrow, R., Alemu, G., & Bedi, A.S. (2013). Community-Based Health
Insurance Schemes A Systematic Review Working Paper No. 568 . International
Institute of Social Studies.
Meng, Q., & Xu Ling. (2014). Monitoring and evaluating progress towards universal health
coverage in China. PLoS Med , 11(9).
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded source
book (2nd ed.). Newbury Park, CA : Sage.
Mills, A., Ally, M., Goudge, J., Gyapong, J., & Mtei, G. (2012). Progress towards universal
coverage: the health systems of Ghana, South Africa and Tanzania. Health Policy and
Planning , 27 (Suppl. 1), i4-i12.
Ministry for Medical Services. (2011). Market Assessment of Private Prepaid Schemes in
Kenya. Niarobi: MOMS.
Ministry of Health. (2015). Kenya National Health Accounts 2012/13. Nairobi: Ministry of
Health.
Ministry of Medical Services and Ministry of Public Health and Sanitaion. (2012). Kenya
Health Policy 2012-2030. GoK.
Mladovsky, P. (2014). Social capital and enrolment in community-based health insurance in
Senegal. London : London School of Economics and Political Science Doctor of
Philosophy.
Mladovsky, P., & Mossialos, E. (2006). A conceptual framework for community-based health
insurance in low-income countries: social capital and economic development. London:
LSE Health, The London School of Economics and Political Science.
236
MoH. (2014). 2013 Kenya Household Health Expenditure and Utilisation Survey. Nairobi:
MoH.
Moreno-Serra, R., & Smith, P. C. (2012). Does progress to universal coverage improve
population health? Lancet , 380: 917-23.
Morestin, F., & Ridde, V. (2009). How can the poor be better integrated into health
insurance programs in Africa? An overview of possible strategies. Université de
Montréal.
Msuya, J.M., Jütting, J.P., & Asfaw, A. (2008). Impact of Community Health Funds on the
Access to Health Care: Empirical Evidence from Rural Tanzania. International Journal
of Public Administration , 30(8): p. 813-833.
Mugenda, O. M., & Mugenda, A. G. (2003). Research Methods: Quantitative and
Qualitative Approaches . ACTS press.
Muiya, B. M., & Kamau, A. (2013). Universal health care in Kenya: Opportunities and
challenges for the informal sector workers. International Journal of Education and
Research , 1(11):1-10.
Munge, K., Mulupi, S., & Chuma, J. (2016). Strategic purchasing for universal health
coverage: A critical assessment Community-Based Health Insurance Schemes in
Kenya. RESYST.
Munguti, R. (2010). New hospital fund rates blocked. Daily Nation , 1 october.
Musau, S. N. (1999). Community-based health Insurance: Experiences and Lessons learned
from East and Southern Africa. Technical report no. 34. Bethesda, MD. Partnerships
for health reforms, Abt associates Inc.
Nahapiet, J., Gratton, L., & Rocha, H. O. (2005). Knowledge and relationships: When
cooperation is the norm. European Management Review , 45: 1-12.
Ndiaye, P., Soors, W., & Criel, B. (2007). Editorial: a view from beneath: community health
insurance in Africa. Tropical Medicine and International Health , 12(2), 157-161.
Neuman, W.L. (2011). Basics of Social Research: Qualitative and Quantitative Approaches
(7th Ed). Bacon, Boston: Pearson Education.
Nunnally, C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. OAU. (2001).
Abuja declaration on HIV/AIDS, tuberculosis and other related infectious diseases.
Addis Ababa: Organization of African Unity.
237
OECD-DAC secretariat simulation of DAC members‘ net ODA volumes in 2008 and 2010.
(2010). OECD.
Okech, C.T. (2013). Review of on-going efforts in the implementation of the Health Care
Financing Functions and Proposal for feasible Financing Options. Unpblished report.
Oxfam International. (2008). Health Insurance in low-income countries. Joint NGO Briefing
Paper.
Oxfam International. (2013). Universal Health Coverage: Why health insurance scheme are
leaving the poor behind. Oxford : Oxfam GB.
Ozawa, s., & Walker, D.G. (2009). Trust in the context of Community Based Health
Insurance Schemes in Cambodia: Villagers trust in health insurers . Advances in Health
Economics and Health Services Research , 12:107-32.
Panda, P., Chakraborty, A., Dror, D. M., & Bedi, A.S. (2013). Enrolment in community-
based health insurance schemes in rural Bihar and Uttar Pradesh, India. Health Policy
and Planning , 1-15.
Parmar, D., Allegri, M.D., Savadogo, G. & Sauerborn, R. (2013). Do community-based
health insurance schemes fulfil the promise of equity? A study from Burkina Faso.
Health Policy and Planning , 29(1), pp. 76-84.
Pauly, M. V., Zweifel, P., Scheffler, R. M., Preker, A. S., & Bassett, M. (2006). Private
health insurance in developing countries. Health Aff (Millwood) , 25, 369-79.
Picazo, O. (2014b). Seven remaining policy and programmatic tasks to achieve effective
UHC coverage in the Philippines. Manila: WHO Country Office in the Philippines.
Poletti, T., Balabanova, D., Ghazaryan, O., Kamal-Yanni, M., Kocharyan, H., Arakelyan, K.,
& Hakobyan, M. (2007). Options for scaling up community-based health insurance for
rural communities in Armenia. Health Systems Development Programme. London
School of Hygiene and Tropical Medicine: London.
Prakongsai, P., Limwattananon, S., & Tangcharoensathien, V. (2009). The equity impact of
the universal coverage policy: lessons from Thailand. Advances in Health Economics
and Health Services Research , 21:57-81.
Preker, A. S., & Carrin, G. (2004). Health Financing for Poor People: Resource
Mobilization and Risk Sharing. Washington. D. C: World Bank.
238
Preker, A. S., Liu, X., Velenyi, E. V., & Baris, E. Eds. (2007). Public Ends, Private Means:
Strategic Purchasing of Health Services. Washington D C: World Bank.
Preker, A.S. (2002). Health Care Financing for Rural and Low-Income Populations: The
Role of Communities in Resource Mobilization and Risk Sharing. Background Reports.
Commission on Macro-Economics and Health.
Private Sector Innovation Programme for Health (PSP4H). (2014). Comparing Kenya’s
Private Health Markets with neighbouring markets; A Focus on Health Financing.
NAIROBI : PSP4H.
Putnam, R. D . (2000). Bowling Alone: the collapse and revival of American community.
New York: Simon & Schuster.
Ranson, M. K., Sinha, T., Chatterjee, M., Gandhi, F., Jayswal, R., Patel, F., Morris, S. S., &
Mills, A. J. (2007). Equitable utilisation of Indian community based health insurance
scheme among its rural membership: cluster randomised controlled trial. BMJ (Clinical
research ed) , 334 (7607). p. 1309.
Rashad, A.S ., & Sharaf, M. F. (2015). Catastrophic Economic Consequences of Healthcare
Payments: Effects on Poverty Estimates in Egypt, Jordan, and Palestine. Economies , 3,
216-234.
Ridde, V., Haddad, S., Nikiema, B., Oudraogo, M., Kafando, Y., & Bicaba, A. (2010). Low
coverage but few inclusion errors in Burkina Faso: a community-based targeting
approach to exempt the indigent from user fees. BMC Public Health , 10(631), 2-8.
Ringle, C. M., Sarstedt, M., & Schlittgen, R., Taylor, C. R. (2013). PLS path modeling and
evolutionary segmentation. Journal of Business Research , 66 (9), 1318-1324.
Robert, A., Bereson, P, J. , & Harlan, M. K. (2013). Achieving the potential Health Care
Performance Measures . Washington: Urban Institute.
Robson, C. (2002). Real World Research. (2nd ed.). Oxford: Blackwell.
Robyn, P.J., Sauerborn, R., Bärnighausen, T. (2013). Provider payment in community-based
health insurance schemes in developing countries: a systematic review. Health Policy
Plan , 28(2):111-122.
Rogers, E. (1962). Diffusion of Innovations, 5th Edition. Simon and Schuster.
Rogers, E. (2003). Diffusion of Innovations, 5th Edition. Simon and Schuster.
239
Ron, A. (1999). NGOs in community health insurance schemes: examples from Guatemala
and the Philippines. Social Science & Medicine , 48, 939-50.
Saksena, P., Xu, K ., & Durairaj, V. (2010). The drivers of catastrophic expenditure:
outpatient services, hospitalization or medicines? World Health Report (2010)
Background Paper, No 21 . Geneva: World Health Organization.
Salomon, J.A., Tandon, A., & Murray, C.J.L. (2004). Comparability of self rated health:
cross sectional multicountry survey using anchoring vignettes. BMJ , 328(7434):258-
61.
Saunders, M., Lewis, P., & Thornhill. (2009). Research Methods for Business Students (5th
ed). Harlow: Prentice Hall.
Saunders, M., Lewis, P., & Thornhill, A. (2012). Research Methods for Business Students
(6th Ed). Essex, England: Pearson Education Ltd.
Saunders, M., Lewis, P., & Thornhill, A. (2016). Research Methods for Business Students
(7th ed). Person Education Limited.
Saunders, M., Lewis, P., & Thornhill, A. (2007). Research Methods for Business Students,
(4th ed,.). Harlow: Prentice Hall Financial Times.
Schieber, G., Cashin, C., Saleh, K., & Lavado, R. (2012). Health Financing in Ghana.
Washington DC: The World Bank.
Schneider, P. (2005). Trust in micro-health insurance: an exploratory study in Rwanda. Soc
Sci Med , 61, 1430-8.
Schneider, P. (2004). Why Should the Poor Insure? Theories of Decision-Making in the
Context of Health Insurance. Health Policy and Planning , 19 (6):349- 355.
Schneider, P., & Diop, F. (2004). Community-based health insurance in Rwanda. In: A.S.
Preker and G. Carrin (eds): Health financing for poor people: Resource mobilization
and risk sharing. pp. 251–74. Washington DC: World Bank.
Sekaran, U., & Bougie, R. (2016). Research Methods for Business (7th ed). West Sussex,
UK: Wiley.
Sinha,T., Ranson, M.K., Chatterjee, M., Acharya, A., & Mills, A.J. (2005). Barriers to
accessing benefits in a community-based insurance scheme: lessons learnt from SEWA
Insurance, Gujarat. Oxford University Press.
240
Smit, N., & Mpedi, L. (2010). Social protection for developing countries: Can social
insurance be more relevant for those working in the informal economy? Law,
Democracy & development.
Smith, P C., & Witter, S. N. (2004). Risk Pooling in Health Care Financing:The Implications
for Health System Performance. HNP Discussion Paper. Washington, DC: World
Bank.
Smith, P. C. (2008). Formula funding of health services: learning from experience in some
developed countries. Geneva: WHO.
Soeters, R., & Griffiths, F. (2003). Improving government health services through contract
management: A case of Cambodia . Health Policy and Planning , 18: 74-83.
Soors, W., Devadasan, N., Durairaj, V., & Criel, B. (2010). Community Health Insurance
and Universal Coverage: Multiple paths, many rivers to cross. World Health Report
(2010) Background Paper, 48. GENEVA: WHO.
Spaan, E., Mathijssen, J., Tromp, N., McBain, F., ten Have, A., & Baltussen, R. (2012). The
impact of health insurance in Africa and Asia: a systematic review. Bulletin of the
World Health Organisation , 90(9),685-692.
Speech by H.E. Hon. Uhuru Kenyatta, C.G.H., President and Commander-in-Chief of the
Defence Forces of the Republic of Kenya During the Madaraka Day Celebrations‖
(Nyayo National Stadium). (2013, June 1). Retrieved October 12, 2015, from
http://www.statehousekenya.go.ke/.
Starfield, B. (2011). The hidden inequity in health care. International Journal for Equity in
Health , 10:15.
Stenberg, K., Elovainio, R., Chisholm, D., Fuhr, D., Perucic, A., Rekve, D., & Yurekli, A.
(2010). Responding to the challenge of resource mobilization- mechanisms for raising
additional domestic resources for health World Health Report (2010) Background
Paper, No 13. Geneva : WHO.
Su, TT., KouyatЙ, B., & Flessa, S . (2006). Catastrophic household expenditure for health
care in a lowincome society: A study from Nouna District, Burkina Faso. Bull World
Health Organ , 84(1):21-7.
Tabachnick, B. G., & Fidell, L. S. (2013). Using Multivariate Statistics (6th ed.). Harlow,
England : Pearson Education.
241
Tabor, S. R. (2005). Community-based health insurance and social protection policy.
Washington DC: World Bank.
Takakura, M. (2011). Does social trust at school affect students‘ smoking and drinking
behaviour in Japan? . Social Science and Medicine , 72(2), 299-306.
Tangcharoensathien, V., Mills, A., & Palu, T. (2015). Accelerating health equity: the key role
of universal health coverage in the Sustainable Development Goals. BMC Medicine ,
13:101.
Tao, Y., Kizito, H., Qinpei, Z., & Xiaoni. (2014). Methods of measuring horizotal equity in
health resource allocation: a comparative study. Health Economics , 4:10.
Tibandebage, P., & Mackintosh, M. (2005). The Market Shaping of Charges, Trust and
Abuse: Health Care Transactions in Tanzania . Social Science & Medicine , 61(7):
1385-95.
Tipke, M., Diallo, S., Coulibaly, B., Storzinger, D., Hoppe-Tichy, T., Sie A.,& Muller O.
(2008). Substandard anti-malarial drugs in Burkina Faso. Malar J , 7:95.
Travis, P., Egger, D., Davies, P., & Mechbal, A. (2002). Towards better
stewardship:concepts and critical issues. Geneva: WHO.
Tundui, C., & Macha, R. (2014). Social capital and willingness to pay for community based
health insurance: An empirical evidence fro rural Tanzania. Journal of Finance and
Economics , Vol 2, Issue 4, 50-67 ISSN 2291 4951.
Usoroh, E. E. (2012). Achieving universal health coverage in Nigeria: The Nigeria Health
Insurance Scheme as a tool. 48th International Course in Health Development .
Amsterdam: Royal Tropical Institute.
Van Leeuwen, E . (2008). Changing moral experiences in European health care. Medical
Law , 27:57-604.
Van Slyke, M. (2006). Agents or Stewards: Using Theory to Understand the Government-
Nonprofit Social Service Contracting Relationship. Journal of Public Administration
Research and Theory , 17: 157-187.
Veillard, J. H. M., Brown, A. D., Baris, E., Permanand, G., Klazinga, N.S. (2011). Health
system stewardship of National health Ministries in the WHO European region:
Concepts, functions and assessment framework. Health Policy , 103: 191-9.
242
Vinzi, V., Trinchera, L., & Amato, S. (2010). PLS Path Modeling: From Foundations to
Recent Developments and Open Issues for Model Assessment and Improvement.
Chapter 2 Handbook of Partial Least Squares. Springer Handbooks of Computational
Statistics.
Wahyuni, D. (2012 ). The Research Design Maze: Understanding Paradigms, Cases,
Methods and Methodologies. Journal of Applied Management Accounting Research ,
Vol. 10, No. 1, pp. 69-80.
Wang, H., & Pielemeier, N. (2012). Community-Based Health Insurance: An Evolutionary
Approach to Achieving Universal Coverage in Low-Income Countries. Journal of Life
Sciences , 6: 320-29.
Welman, C., Kruger, F., & Mitchell, B . (2007). Research Methodology (3rd). Cape Town:
Oxford University Press.
Whetten, D. A. (1989). What Constitutes a Theoretical Contribution? . Academy of
Management Review , Vol. 14, No. 4, 490-495.
WHO South Africa. (2015). Risk pooling in health financing: Rationale and principles.
Working paper. WHO.
Williams, C. (2007). Research Methods. Journal of Business & Economics Research , 5 (7).
Witter, S., & Garshong, B. (2009). Something old or something new? Social health insurance
in Ghana . BMC International Health and Human Rights , 9:20.
Wodtke, G, T., Elwert, F.,& Harding, D.J. (2012). Poor families, Poor neighbourhoods: How
family poverty intesifies the impact of concentrated disadvantage. Report 12-776.
Population Studies Report.
Woolcock, M. (1998). Social Capital and Economic Development: Toward a Theoretical
Synthesis and Policy Framework. In Theory and Society , 27, 151-208.
World Bank. (2012). World Development Report 2013: Jobs. Washington, DC: World Bank.
World Bank . (2014). Philippines: Delivering Successful Sin Tax Reform. Geneva : World
Bank.
World Bank. (2015). World Development Indicators database. Washington DC: World Bank.
World Health Organization. (1948). Constitution of the World Health Organization. Geneva:
WHO.
243
World Health Organization. (1978). Declaration of Alma-Ata:Ata: international conference
on primary health care. Geneva: WHO.
World Health Organization. (1981). Global Strategy for Health for All by the Year 2000.
Geneva: WHO.
World Health Organization. (2000). The World Health Report 2000 - Health systems:
improving performance. Geneva: WHO.
World Health Organization. (2001). Macroeconomics and health: Investing in health for
economic development. Geneva: WHO.
World Health Organization. (2001). Report on WHO meeting of experts on the
StewardshipFunction in Health Systems. Geneva: WHO.
World Health Organization. (2005). Social health insurance: Sustainable health financing,
universal coverage and social health insurance In: 58th World Health Assembly.
WHO.
World Health Organization. (2006). The World Health Report 2006 - Working together for
health. Geneva: WHO.
World Health Organization . (2007). Everybody's Buisness - Strengthening Helath Systems to
improve Health outcomes. Geneva: WHO.
World Health Organization. (2008). The World Health report 2008: primary health care now
more than ever. Geneva : WHO.
World Health Organization. (2009). Taskforce on Innovative International Financing for
Health systems, Working Group 2 Report. Geneva: WHO.
World Health Organization. (2010a). Health Systems Financing - The path to universal
coverage. Geneva: WHO.
World Health Organization. (2010b). Constraints to scaling up the health Millennium
Development Goals: costing and financial gap analysis. Geneva: WHO.
World Health Organization. (2012). Intersectoral Governance for Health in All Policies:
Structures, actions and experiences. European observatory on health systems and
policies. Copenhagen: WHO.
World Health Organization. (2013). The state of health financing in the African Region.
Brazzavile: WHO, Region Office of Africa.
244
World Health Organization. (2014a). WHO African Region Expediture Atlas. Brazzaville:
WHO Regional Office for Africa.
World Health Organization. (2014b). Health Systems Governance for Universal Health
Coverage: Action Plan . Geneva : WHO.
World Health Organization. (2014c). Fiscal space, public finance management and health
financing policy: a collaborative agenda. Montreux, Switzerland: WHO.
World Health Organization. (2015a). Tracking Universal Health Coverage: First Global
Monitoring Report. Geneva: WHO.
World Health Organization. (2015b). Do health expenditures meet health needs? Geneva:
WHO.
World Health Organization. (2015c). Positioning health in the post-2015 development
agenda. Geneva: WHO.
World Health Organization. (2015d). Anchoring universal health coverage in the right to
health: What difference would it make? Policy brief. Geneva: WHO.
World Health Organization. (2015e). World Health Statistics. Geneva: WHO.
World Health Organization. (2015f). External AID for health remains insufficient in low
income countries. WHO/HA Policy Highlight No.3. Geneva: WHO.
World Health Organization. (2016). World Health Statistics: Monitoring health for the
SDGs. Geneva: WHO.
World Health Organization- GTZ. (2004). National Social Health Insurance Financial
projections and future bilateral / multilateral cooperation. sixth joint WHO/GTZ
mission report. Nairobi: WHO - GTZ.
Xu, K., Evans, D.B., Carrin, G., Agilar-Rivera, A.M, Musgrove, P., & Evans, T . (2007).
Protecting households from catastrophic health spending (Millwood). Health Affairs ,
26:972-983.
Xu, K., Evans, D.B., Kawabata., K., Zeramdini, R., Klavus, J., & Murray, C.J. (2003).
Household catastrophic health expenditure: a multicountry analysis. Lancet ,
362(9378): 111-117.
Xu, K., Saksena, P., & Evans, D.B. (2010). Health financing and access to effective
interventions. World Health Report (2010) Background Paper, 8. Geneva: World
Health Organization.
245
Yamane, T . (1967). Statistics: An Introductory Analysis, (2nd Ed.). New York: Harper and
Row.
246
APPENDICES
APPENDIX 1: Approval Letter from USIU-A
247
APPENDIX 2: APPROVAL LETTER FROM NACOSTI
248
APPENDIX 3: NACOSTI RESEARCH CLEARANCE PERMIT
249
APPENDIX 4: QUESTIONNAIRE
Instructions
In section I – II of this questionnaire you have been provided with one type of question:
1. Tick (√) the one which closely matches your opinion.
SECTION ONE: GENERAL QUESTIONS
1.1. What is the name of your CBHIs? …………………………………..
1.2. How many years has the CBHIs been operating? ........................................
1-5 years ( ) 6-10 years ( ) 11-15 years ( ) > 15 years ( )
SECTION TWO: HEALTH FINANCING FUNCTIONS
2.1 Enrolment
Indicate with a tick (√) the statement that best describes the enrolment strategy for the
CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
2.1.1 Affordability 1 2 3 4 5
We give members a chance to allocate premium among preferred products
Members can pay premium in kind e.g. through farm produce or labour
We give subsidies and exemption of premiums for extremely poor and
vulnerable
We encourage members to use savings-linked premium payment
mechanisms such as rotating saving groups (Chamas), MPESA
Members can make irregular payments of premium
2.1.2 Unit of membership
The CBHIs membership is based on coffee, tea, villages or mutual benefit
societies or administrative areas
The CBHIs have adapted households as unit of membership
CBHIs encourages members to join when they are healthy
CBHIs membership is open to poor and vulnerable groups
Any other, please indicate
2.1.3 Timing of collection
250
Members pay in a single annual premium /contribution
Members can pay their premiums in installments
Premium payments correspond with income from e.g. harvest, sale of
livestock or salary payment
Premium payments are linked to loans from SACCOs and banks
Mobile premiums payments are allowed
Any other, please indicate
2.1.4 Trust
Members interact with the scheme‘s administrative / management team and
providers about their needs, concerns and make suggestions for
improvements
Members participate in setting of benefit package
Members participate in setting the premiums
The CBHIs uses existing Chamas, community development projects and
credit schemes as entry points for CBHIs membership
Member of the scheme are willing to cover the poorest and vulnerable
group in the community such as orphans and disabled
Any other, please indicate
2.2 Mix of Contributions
Indicate with a tick (√) the statement that best describes the adequacy and mix of
contributions in the CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
1 2 3 4 5
Members‘ contributions are adequate in meeting the cost of the set benefit
package
NHIF covers costs of services not covered by the CBHIs
The CBHIs receives financial support from donor(s)
The poor and vulnerable members of the CBHIs are covered through
government subsidies
251
Any other, please indicate
2.3 Indicate with a tick (√) the statement that best describes the risk pooling strategies that
are employed by the CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
RISK POOLING 1 2 3 4 5
i) Enhanced risk pooling
Members of CBHIs comes from a wide range of social economic
background
The scheme targets a large geographical /administrative area
Community members are encouraged to join CBHIs when they are healthy
We have a waiting period before one can benefit from CBHI
The CBHIs has other branches in other geographical or administrative areas
CBHIs reinsures its risks
The CBHIs has a partnered with the county / national government and or
NHIF
The CBHIs merged with other CBHIs to form a network or a federation
Any other, please indicate
ii a) Social solidarity
Members of the scheme have expressed the opinion that if they would not
need health care themselves, at least they had done something good for the
community by contributing to the insurance fund
Any other, please indicate
252
Indicate with a tick (√) the statement that best describes the risk pooling strategies that are
employed by the CBHIs. Tick your choice in the appropriate answer box.
1 = None of the cost, 2 = Some of the cost, 3= Half of the cost 4 = Most of the cost, 5 = All
of the cost
ib) Social solidarity
How much do you think members of the CBHIs are willing to contribute to
pay for health care services used by the sick?
How much do you think members of the CBHIs are willing to contribute to
pay for health care services used by the poor?
Any other, please indicate
2.4 Indicate with a tick (√) the statement that best describes the strategic purchasing activities
undertaken by the CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
STRATEGIC PURCHASING 1 2 3 4 5
We have signed a contract with all our health services providers
We only select providers who are accredited by NHIF
The providers must provide health services according to conditions put
forward by the CBHIs
We allocate resources based on population needs
The CBHIs has been successful in negotiating agreeable terms and contract
with service providers – in terms of service quality, fee, and reduction in
unnecessary services/prescription (moral hazard)
Any other, please indicate
253
2.5 Indicate with a tick (√) the statement that best describes the role played the government
and or donors‘ financial sustainability of your company. Tick your choice in the appropriate
answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
GOVERNMENT ROLE 1 2 3 4 5
A. ADVISER ON DESIGN OF CBHIS: The Government recommends
Startup of CBHIs when a minimum percentage of population in enrolled
A waiting period
Enrollment of households as opposed to individuals
A flexible premium collection system
A benefit package that reflect the needs of the target population
A standard treatment protocols for members of CBHIs
A standard referral procedure
Consolidation of CBHIs through a federation or a network
Creation of a risk equalization fund or a reinsurance mechanism
Community participation in management and decision making
Any other, please indicate
B. MONITORING OF CBHI ACTIVITIES: The government
Tracks the progress of CBHIs through time
Monitors the basic performance of CBHIs
Any other, please indicate
C. TRAINING: The government organizes trainings on
Determination of benefit packages
Determination of contributions
Collection of contributions
254
Claims processing
Use of Management Information Systems
Establishment of Health Insurance Development Plans
Exchange visits
Any other, please indicate
D. CO- FINANCING: The government and/ or donors
Partially or fully subsidizes the poorest and vulnerable members of the
community
Has set a solidarity fund for financing epidemics and deficits of the CBHIs
Any other, please indicate
2.6 EQUITY IN HEALTH CARE
2.6.1 Indicate with a tick (√) the statement that best describes access to health services to
members of your CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
Access 1 2 3 4 5
There is distribution of enrolment across income categories 1
The contracted providers are within the proximity of covered population 2
We cater for transport / accommodation cost related to healthcare utilization 3
The covered population is entitled to similar benefits 4
The number of members seeking services has increased in the past 12
months
5
Any other, please indicate
255
2.6.2 Indicate with a tick (√) the statement that best describes contribution across different
income categories within the CBHIs. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
Equity in Contribution 1 2 3 4 5
Everyone pays the same amount
Everyone pays an equal amount of their income
We allow flexible premium payments
We offer allow members to match premium or products to their income
We offer premium subsidies
We allocate a larger claim budget for low cost products
Any other, please indicate
2.6.3 Indicate with a tick (√) the statement that best describes perceived quality of care from
contracted health care providers. Tick your choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
2.6.3 Quality of care 1 2 3 4 5
The CBHIs has a standard client compliant management mechanism
Members have complained about long queues before being seen
Members have complained on availability of health services
Members have complained about lack of key prescribed medicines
Members have raised concerns relate to cleanliness
Members have raised concerns on availability of trained staff in the
contracted health facilities
The CBHIs have put in place mechanisms to check on patient perceived
quality of care in contracted health facilities on issues concerning waiting
time, availability of staff, services, drugs and supplies
There are other organization(s) that conduct quality checks in the contracted
health facilities
These organizations share their findings with the CBHIs
Any other, please indicate
256
2.6.4.1 Indicate with a tick (√) the statement that best describes the administrative and
managerial capability of the CBHIs management team. Tick your choice in the appropriate
answer box.
1 = Strongly Disagree, 2 = Disagree, 3= Neutral 4 = Agree, 5 = Strongly Agree
1 2 3 4 5
Administrative and managerial capability: The administrative committee
has basic skills in
Setting of contributions
Collection of contributions and compliance
Determination of the benefit package
Claim management
Marketing and communication
Contracting with providers
Management information systems
Accounting
Any other, please indicate
2.6.4.2 Financial Sustainability
Indicate with a tick (√) the statement that best describes the practice of the CBHIs. Tick your
choice in the appropriate answer box.
1 = Strongly Disagree, 2 = Disagree, 3 = Agree, 4 = Strongly Agree
Financial sustainability 1 2 3 4 5
We have partnered with organizations that assist in collection of premiums
Premiums are not paid on time
The CBHIs is funded through a mix of contributions from county / national
government / donors and members contributions
Government and or donors‘ covers health cost for those who cannot afford
257
to pay premiums
Chronic conditions are covered by the CBHIs
The CBHI have put in place mechanisms to check whether the invoices sent
from the health facilities are correct
There are instances which health facilities tried to overstate the
reimbursement request amount
The CBHIs is part of a network of CBHIs
We have merged with other CBHIs
We are in partnership with NHIF
Besides treatment we finance community prevention, promotion and
rehabilitation activities
Any other, please indicate
Thank you for your responses
258
APPENDIX 5: SECONDARY DATA SHEET
Instructions
In parts I – IV of this questionnaire you have been provided with two types of questions:
1. Tick (√) the one which closely matches your opinion.
2. Tables which require that you indicate the values in each category for a specific year.
SECTION ONE: GENERAL QUESTIONS
1.1 How many households are targeted by the CBHIs? ………………………….
0-500 ( ) 501-1000 ( ) 1001-2000 ( )
1.2 How many households are covered by the CBHIs? ..................................
0-500 ( ) 501-1000 ( ) 1001-2000 ( )
1.3 Indicate the percentage of members that lives within the stated distance from the
participating health facilities.
Within 5km …………Between 5-10km ………………More than 10km ………………
1.4 Which methods of payment is used by the CBHIs?
i) Inpatient services
Capitation method ( ) Fee for service method ( ) Case based payment ( )
Mixed methods ( )
ii) Outpatient services
Capitation method ( ) Fee for service method ( ) Case based payment ( )
Mixed methods ( )
1.5 How many members of CBHI have drop out of NHIF in the last one year? ………
259
1.6 Indicate the insurance products offered by the CBHIs and premiums per product
CBHIs only products Composite product (NHIF and CBHIs)
Product Premium Product Premium
i CBHIs cover for small households
(public hospital services)
i Compounded CBHIs & NHIF cover
ii CBHIs cover for expanded households
(public hospital services)
ii Compounded CBHIs & NHIF expanded
cover
iii CBHIs general outpatient cover iii Cover with NHIF outside CBHIs
iv CBHIs general outpatient and inpatient
cover
iv Cover with NHIF outside CBHIs
(expanded benefits)
v CBHIs general outpatient and inpatient
cover with ambulance services
vi CBHIs general outpatient and inpatient
cover with ambulance services and
burial expenses
260
SECTION TWO: ENROLMENT AND CONTRIBUTIONS
Indicate the number of households enrolled and the contributions for the following years?
Item Population
2015 2014 2013 2012 2011 2010
i) Target Population
iv) Total members of the CBHIs
Members covered by both
CBHIs and NHIF
Members covered by CBHIs
only
SECTION THREE: MIX OF CONTRIBUTIONS
Indicate the amount of Contributions from each category
Category 2015 2014 2013 2012 2011 2010
Membership
contributions
Donor
NHIF
Government
261
SECTION FOUR: FINANCIAL STATUS OF CBHI SCHEME
What have been the income, reimbursements, costs and surplus/deficits for your CBHI for the following years?
2015 2014 2013 2012 2011 2010
Total income of the CBHI
scheme
Reimbursement of the cost of
OPD and inpatient services in
health facilities
Administrative cost
Total cost of the scheme
Surplus / Deficit
262
APPENDIX 6: List of CBHIs (registered and sampled) and their respective Networks
263
APPENDIX 7: Cross Loadings of Constructs
Cross loadings for Enrolment
Pattern Matrixa
Component
1 2 3 4 5
AFF1 0.915 -0.025 0 0.074 -0.198
AFF2 0.024 -0.035 0.056 0.013 -0.019
AFF3 0.031 -0.03 -0.086 0.07 0.013
AFF4 0.901 -0.027 0 0.071 -0.19
AFF5 -0.017 0.002 -0.102 0.095 0.02
MEM1 0.013 0.043 -0.003 0.057 -0.033
MEM2 -0.019 0.023 0.063 -0.053 -0.102
MEM3 -0.018 0.039 0.258 0.431 0.254
MEM4 0.014 0.017 -0.02 0.036 -0.071
TM1 0.873 0.018 -0.003 -0.038 0.207
TM2 -0.011 0.039 0.006 -0.039 0.031
TM3 0.884 -0.004 0.024 0.029 0.217
TM4 -0.011 0.039 0.006 -0.039 0.031
TRU1 0.87 0.012 0.002 -0.012 0.247
TRU2 0.905 -0.005 -0.001 0.045 -0.229
TRU3 0.849 0.003 0.029 -0.071 0.231
TRU4 0.905 0.013 -0.013 0.049 -0.222
TRU5 0.589 0.033 -0.045 -0.287 0.008
Extraction Method: Principal Component Analysis.
Rotation Method: Promax with Kaiser Normalization.
a. Rotation converged in 5 iterations.
264
APPENDIX 7: Cross Loadings of Constructs (Continued)
Extraction Method: Principal Component Analysis
Rotation Method: Promax with Kaiser Normalization
a. Rotation converged in 3 Iterations
Cross loadings for Strategic Purchasing
Pattern Matrixa
Component
1
SP1 .783
SP2 .876
SP3 .861
SP4 .849
SP5 .476
Cross loadings for Mix of Contribution
Pattern Matrixa
Component
1 2
MC1 .327 -.209
MC2 .984 .084
MC3 .256 .068
MC4 .964 .084
Cross loadings for Risk pooling Pattern Matrix
a
Component
1 2 3 4
ERP1 .801 .103 .003 -.114
ERP2 .015 .052 -.786 -.221
ERP3 .971 -.039 -.016 -.019
ERP4 -.044 .995 -.027 .010
ERP5 .015 .055 .024 -.249
ERP6 .745 .055 .027 .041
ERP7 .018 .045 -.001 .141
ERP8 .823 -.050 .014 .145
265
APPENDIX 7: Cross Loadings of Constructs (Continued)
Extraction Method: Principal Component Analysis
Rotation Method: Promax with Kaiser Normalization
a. Rotation converged in 3 Iterations
Cross loadings for Government Stewardship
Pattern Matrixa
Component
1 2
COF1 .884 -.015
COF2 .904 -.018
COF3 -.002 .099
TR1 .942 -.001
TR2 .964 .009
TR3 .978 .031
TR4 .948 .008
TR5 .045 -.019
TR6 .070 .000
TR7 .012 .024
MO1 .933 -.046
MO2 .936 -.037
AD1 .916 .034
AD2 .945 .027
AD3 .952 -.006
AD4 .924 -.021
AD5 .969 .009
AD6 .973 .018
AD7 .941 .029
AD8 .931 -.008
AD9 .937 -.023
AD10 .057 .002
266
APPENDIX 7: Cross Loadings of Constructs (Continued)
Extraction Method: Principal Component Analysis. Rotation Method: Promax with Kaiser Normalization. a. Rotation converged in 6 iterations.
Cross loadings for Equity in HealthCare
Pattern Matrixa
Component
1 2 3 4
ACC1 .027 -.008 .003 .085
ACC2 .063 .038 .001 .234
ACC3 .081 .008 .010 .055
ACC4 .071 -.049 -.014 .714
AMC1 .099 -.008 .992 -.075
AMC2 .096 .002 .016 .020
AMC3 .004 -.008 .009 -.140
AMC4 .006 .020 .097 -.014
AMC5 -.011 .007 .097 .016
AMC6 .040 .002 .978 .038
AMC7 .037 .817 -.073 .006
AMC8 -.024 -.021 -.007 .091
FS1 .004 .915 .009 -.140
FS2 .398 .002 -.001 .038
FS3 -.001 .099 .011 -.010
FS4 -.001 .099 .011 -.010
FS5 .089 .750 -.068 -.048
FS6 .015 .014 .016 -.082
FS7 .020 .075 -.066 .008
FS8 .321 .015 -.001 .640
FS9 -.044 .091 -.084 -.061
FS10 .988 .988 -.006 -.205
FS11 -.039 .910 .035 .071
QOC1 -.001 -.003 .011 .040
QOC2 -.001 -.001 .011 -.010
QOC3 -.001 -.084 .011 -.010
QOC4 -.001 -.006 .011 -.010
QOC5 -.001 .035 .011 -.010
QOC6 -.001 .011 .011 -.010
QOC7 -.001 .035 .011 -.010
QOC8 .914 -.007 .038 .028
QOC9 .886 .033 -.060 .077
QOC10 .095 -.003 .011 .040