Effectiveness of Public and Co-operative Health Insurance Models in Nepal : A Concurrent Mixed Study

1 Department of Preventive Medicine, Wonju College of Medicine Yonsei University, Wonju, Gangwon 220-701, Korea 2 Institute for Poverty Alleviation and International Development, Yonsei University Wonju, Gangwon 220-710, Korea 3 Health Science Foundations and Study Center, Kathmandu, Nepal * Corresponding Author: Chun-Bae Kim, email: kimcb@yonsei.ac.kr and Chhabi Ranabhat, email: chhabir@gmail.com; Tel: +82337410347, Fax: +82337470409


Background
In voluntary health insurance (HI), community-based health insurance (CBHI) is attracting more and more attention in low-and middle-income countries as a means of improving health care utilization and to protect households against impoverishment caused by out-of-pocket medical expenditures.
The World Health Organization and the World Bank have continuously suggested reducing out-ofpocket payments and promoting universal health coverage.[1,2] Different health financing approaches have been developed to counter the detrimental effects of user fees introduced in the 1980s, but those efforts have not yet increased healthcare utilization, particularly among marginalised populations, and, moreover, sometimes lead to catastrophic health expenditures.[2][3][4] There are different models of health insurance; among them, CBHI is the most widely used among middleincome populations and in remote areas.CBHI has been implemented on a small scale in Nepal, but coverage remains relatively low.
As stated, there are different models of health insurance.Mandatory and single payer health insurance models are considered among the better approaches, but it is very difficult to collect premiums in low-and middle-income countries due to a lack of information technology.[5] Likewise, for people who are ultra-poor and live in rural areas, it is an additional challenge to enrol in health insurance.The only government approach is unable to provide universal healthcare because the primary responsibility for health falls upon each individual.The Obama administration purposed reducing healthcare costs and providing affordable health service using a cooperative concept in the US [6]; and, cooperative health insurance is replacing private health insurance in Canada.[7] Voluntary health insurances could be the step forward needed to lower risks associated with illness.
Two approaches to voluntary health insurance have been initiated both by the government and by local communities.Some studies have investigated the outcomes of CBHI in low-and middle-income countries, but their results have been inconsistent.Several previous reviews have evaluated the performance of CBHI in terms of enrolment, financial management, and sustainability.[8][9][10] One study of Laos indicated that government funded CBHI has low coverage, but the insured people typically have a significantly higher level of healthcare utilization, lower out-of-pocket payments (OPPs), lower incidence of catastrophic expenditures, and a lower propensity to employ coping mechanisms.[11] In a study in Ethiopia, out patients department (OPD) services increased, while inpatient department (IPD) services remained the same even after the implementation of CBHI.[12] Not only on low and middle countries, The Obama administration purposed reducing healthcare costs and providing affordable health service using a cooperative concept in the US [6]; and, cooperative health insurance is replacing private health insurance in Canada.[7] There are some successful reports of voluntary health insurance by community initiatives, such as in Vietnam [13] and Japan, [14] and government initiatives in developing countries.In China, a New Cooperative Medical Scheme (NCMS) was piloted, and it had mixed results.[15] Similarly, the state government of Karnataka in India also made health insurance available to its people and tried to reduce OPPs and catastrophic health expenditure (CHE).[16] Previous studies have largely focused on the single model of CBHI, but a comparison of outcomes on government and co-operative CBHI models are rare globally, and none have focused on Nepal.

Theoretical foundation
International Cooperative Alliance (ICA) defined that a co-operative is an autonomous association of persons united voluntarily to meet their common economic, social, and cultural needs and aspirations through a jointly owned and democratically-controlled enterprise [17].It further explored that cooperatives are based on the values of self-help, self-responsibility, democracy, equality, equity and solidarity.Similarly, cooperatives includes six basic principles; voluntary and open membership, democratic member control, member's economic participation, autonomy and independency, education training and information, cooperation among cooperatives and concern for the community.Previous experiences and after the assessment of Millinium Development Goal (MDG), there is a significant role of cooperative to reduce poverty and quality of health care.Cooperatives ensure healthy lives by creating the infrastructure for delivering healthcare services; financing healthcare and providing home-based healthcare services to people living with HIV/AIDS, among others [18].There are three types of health cooperatives; consumer, service purchasing and worker cooperative [19] and due to small scale of economy of the cooperative groups, service purchasing cooperatives are popular in low and middle income countries.On the other side, government health insurance is a program that should implement through government health care system through their personnel.Centre government provide hospital management committee determine some budget (subsidy) and premium and benefit package.Enrolment is volunteer and some discount is given to the minority and ultra poor [20].
There are some successful reports of voluntary health insurance by community initiatives, such as in Vietnam [13] and Japan, [14] and government initiatives in developing countries.In China, a New Cooperative Medical Scheme (NCMS) was piloted, and it had mixed results.[15] Similarly, the state government of Karnataka in India also made health insurance available to its people and tried to reduce OPPs and catastrophic health expenditure (CHE).[16] Previous studies have largely focused on the single model of CBHI, but a comparison of outcomes on government and co-operative CBHI models are rare globally, and none have focused on Nepal.In Nepal, there have been several experiments with CBHI reported since 2004 in both rural and urban settings.[21] Provider based health insurance was introduced in Nepal in 2003 through six pilot schemes offered by the government.At the same time, some community groups (coops) started CBHI by their own initiative supported by non-governmental organizations (NGOs).[22] CBHI schemes in Nepal complement a number of specialized programs for improving access to healthcare services.
The aim of the present study is to compare the outcomes of CBHI offered by government health facilities and those by community groups (co-ops) in terms of benefit packages, population coverage, inclusiveness, healthcare utilization, and promptness for treatment in Nepal.

Study design
The study is a descriptive cross-sectional of two CBHI models in 2014.

Study approaches
The researchers used a mix method approach both qualitative and quantitative data comparisons between two groups and synthesizing the results.[23] We used David Driscol et.al (2007) concurrent structure survey by applying open ended and closed questions with same respondents [24].This method integrates research questions, employs rigorous quantitative research assessing magnitude, and explores the meaning and understanding of concepts; it utilizes multiple processes, and combines these to draw on their strengths; and, frames the investigation within philosophical and theoretical positions.[25] Recently, Creswell [26] emphasized that such an approach is better for philosophical, theoretical, and methodological perspectives.Qualitative research seeks to understand how individuals explore and perceive their experiences.Quantitative research is more powerful for the generalization of the findings; whereas, a combined approach is more valid, reliable, and replicable.Thus, the mixed impact of both designs is more powerful than a single approach.The use of "mixing" procedure in health services research involves the integration, connection, and embedding of these 2 data components [27].

Study setting
Being a concurrent mix study, we obtained the data from institution and opinion from key informants.The institutions were selected purposively and total 12 institutions were taken in which 6 were government health centres and 6 were coops conducted CBHI.The qualitative information was taken from responsible (focal person) of each government health centres and member secretary

Sequential presentation of result
Concurrent mix study design: David 2007 from coops.Their responses were categorised and presented in narrative form after the data results and summary statements were presented from each group.

Study group and population
We selected two models of community-based health insurance in Nepal.

Co-operative prepayment health organization (Group B):
This group had six community groups governed by these communities in collaboration with private and government hospitals.This model can be defined as a zero-cost financing model that empowers these community groups.There was some support from Non Government Organizations (NGOs), but this was in-kind support only.
Funding came from their regular savings, subsidies, and donations from other organizations and some amount of benefits from group income generation activities.

Study area
The study area was selected from 12 pilot organizations that had managed government health facilities and co-operative organizations (Table 1).

Study tools and technique
Three sets of study tools were prepared for information collection.
Review of the records: Data were obtained from the logbooks, ledgers, enrolment registers, and meeting minutes of the health facilities and co-operative groups.From these, the characteristics and coverage levels of the enrolees were determined.The characteristics gathered included the enrolee age, gender, religion, and ethnicity.The following information was collected from each group's records:

Data management
The data were divided into two parts.The numeric data were exported into Excel and analysed using the Statistical Package for the Social Sciences (SPSS) version 20.Likewise, the qualitative information was categorized into different groups.Common information was presented in a narrative form.

Data analysis
In the first phase, the descriptive findings were plotted based on different categories related to enrolment, population coverage, and population composition.In the second phase, numeric data were analysed and tested statistically by applying the Mann-Whitney U test as a two-group comparison.The qualitative information was analysed by inductive method and after analysed the results were integrated [28].

Research outcome analysis
This study was group-based, and the mean comparison of variables was included as like below research framework.

Validity and reliability
The results were synthesized carefully according to the data synthesis model of both qualitative and quantitative information.Numeric data were repeatedly checked by the appointed researcher and were cross-verified by other researchers.The tools were pre-tested in a similar community using video recordings.Necessary changes were made to the tools after the pilot study.

Ethical consideration
Study approval was granted by the Department of Health Service (DoHS) of Teku Kathmandu, Nepal and a formal letter was sent to the co-operatives and government health facilities.The research started after approval.

Results
Our study compared the outcomes of the government and the co-operative health insurance plans in Nepal.In the government CBHI, 1,422 households with 22,691 individuals were enrolled and 2,152 households with 10,106 individuals were enrolled in the co-ops.The composition of male and female was almost equal, domination of adult group, Hindu religion and mixed ethnicity were mostly involved in both CBHI models and enrolment was not satisfactory (low), and population coverage was minimal.The benefit package was slightly higher for the co-op group, while population coverage was better for the government group.Inclusiveness, overall management, and proxy activities were better in the co-op group (Table 2).

Group B
Community finance CBHI six groups

Outcome
• Benefit package amount  Table 3 shows a comparison of the mean of each group variable examined in numeric data.The proportion of overall inclusiveness for the government group was significantly higher (p<0.05)than that of the co-op group.However, the amount of negotiation and average response for treatment after illness/injury (initial and refer) to reach hospital was significantly better for the co-op group than the government health facilities.

Results from the key informant interviews
A similar result was found during the key informant interviews.Both representatives expressed that there was less interest among people in new enrolment.During the observations and conversations with key informants, the following results were found.
Although many people receive services from the government health facilities, the services offered were not promoted to insured patients to promote health insurance.Health workers and other personnel lacked experience in health insurance management and realized that they could not provide priority to insured patients or encourage new enrolment because the staff already had high There are no extra human resources for health insurance, and we have to perform duties related to public health and curative services at this health facility.We have no time to go out in the community to convince people to obtain health insurance or other special benefit for insured people.The enrolment of individuals in the higher social classes is rare, but the levels of participation from other groups are acceptable.We are taking extra responsibility work regarding health insurance but receive no additional incentives.There is a very limited budget from the government; and, they do not release it time.We are also not trained in health insurance and have no previous experience with it, so we have difficulty recording and reporting as required.However, if the government mandates health insurance for all health facilities, the program will become consistent, and trained human resources will be appointed.
Community groups had different experiences.Some NGOs supported the building of capabilities, such as training, workshops, and materials (whereas, the government did not provide such support to its group).These co-op groups maintained records properly in comparison with the government health facilities, but the recording systems were not consistent.At the time of this study, they were just starting to use computer-based records for each patient, expenditures, and other important decisions.They had collected more funds from their members and minimized administrative costs.
Due to their strong negotiating skills, they had saved some money during service contracts with hospitals and invested this into other income generation activities.There was quite a different conversation with the representative member from the co-ops: The government does not provide us with any subsidy.The NGOs provide some occasional support, and they rely on us to train our staff and provide some materials for the office.We have managed all activities related to health insurance by ourselves.We have started some income generation activities in our group such as vegetable farming, poultry farming, and live stock management, and we have our own ambulance to carry people who need transport to the hospital.We manage all efforts for our group members; volunteers do almost all the work and therefore there is minimum cost in administration.
Up to now, nobody in our group has been deprived of treatment during an illness or faced catastrophic health expenditures.If the government provided some subsidy, we could also enrol ultra-poor groups.
With our limited resources, we are doing our best.
The study revealed that new enrolment was not increased, health care utilization rate was increased and the benefit package was almost same in both groups.The overall inclusiveness was higher for the government HI, but enrolment from the religious minority, proportion of negotiated amount during treatment were significantly higher (p<0.05).During illness, the response time to reach hospital was significantly faster in cooperative health insurance than government health insurance.
Qualitative findings showed that level of participation, accountability, transparency and recording system was better in cooperative health insurance than public.

Discussion
This study clearly shows the characteristics of the two models of CBHI in Nepal.The population coverage was significantly higher in the government conducted CBHI, but inclusiveness and institutional capacity were stronger in the co-ops.Health insurance has been in operation in Nepal for a long time on a small scale, but the Ministry of Health and Population (MoHP) has been unable to establish milestone target for universal health care (UHC).This is because achieving UHC is still challenging.The whole labour market has not been defined and is not linked through a banking system.As a result, it is difficult to collect health insurance premiums and about one quarter of the population is still using traditional and complementary medicine in Nepal.[29] Likewise, there is a lack of a legal framework as well as operating guidelines for health insurance in Nepal.Thus, the CBHI approach could be an effective approach to achieve curative health service.The current enrolment status indicates that the government was unable to convince people to share the costs during illness.Existing government CBHI programs are not attractive to people and the co-op CBHI had poor coverage as well but with a positive direction.User fees, community drug programs, and free health service policies in past have created confusion among individuals looking to enrol in CBHI.[30] There are some pros and cons concerning the community health-financing scheme.We found comparatively large coverage and relatively flexible premiums (in terms of payment schedules) and subsidies for the ultra-poor in the government run CBHI.However, in this model, the local communities were unable to take ownership and there was very low utilization of the resources.In contrast, in the co-ops pre-payment CBHI, this approach engendered trust and a feeling of ownership.
A similar conclusion was drawn by Mebratie [31] in a systematic review published in 2013.
In looking at our results and comparing these to other studies, we found similar trends.Enrolment is the first step in any CBHI model.Based on the large number of enrolees, the average enrolment was significantly higher in the government CBHI model, but the number of disadvantaged minorities was significantly higher in the co-op group.This finding is similar to results in a study in India that women in self-help groups found more inclusiveness among minority populations.[32] In both groups, health care utilization increased significantly (up to two times), and this condition was also observed in the Sky community group in Cambodia [33], the Grameen Bank group in Bangladesh [34], the government Amhims group in Ghana [35], Jaminan Kesehatan Aceh (JKA) scheme in Indonesia [36] the Mutelleus Government Center in Rwanda [37], Kerala India [38] and Vietnam.[39] At the same time, coverage of the population in the catchment area was low in both groups.The same trend was observed in the People's Democratic Republic of Laos.[11] New enrolment and retention of current enrolees was low in both models due to uncertain financial viability, quality of care, long waiting time for treatment, and proper management of health care providers, similarly to conditions in Ghana.[40] Yeshavani is co-operative CBHI provider in India, [41] and Urban Resident Basic Medical Insurance (URBMI) has the same function in China [42] as like government CBHI model in Nepal.In both cases, people from remote areas and those in a higher education class were not interested in enrolment, similarly to a trend in Mali [43] and findings are similar trend with our results.The number of new enrolees has been decreasing in the Hanang district of Tanzania, [44] as in the government CBHI in our study.Enrolment of members of the Terai disadvantaged group was significantly higher in the co-op insured group; this finding is similar to the results from the SEWA group in India, [45] but the Nouna community health organization in Burkina Faso [46] and the Les mutealle de santé in Senegal [47] have not been able to cover disadvantaged groups.
The discounts provided during service contracting with hospitals were significantly larger (p<0.05) for the co-op group (by effective negotiation skills) versus the government public health centres.This is not only beneficial for the sustainability of the health insurance industry but also empowers individual and institutional capabilities.
In India, it was concluded that people's negotiating power reduced the costs and improved the quality of service.[48] In Canada, per patient cost was 17% lower than the average price; hospitalization rates were up to 30% lower, and 21% less money was spent on prescription drugs [49] in co-operative health insurance versus private health insurance.In China, the operating capacity of a CBHI is significantly higher in a new co-operative medical scheme (NCMS) than it is in a government health insurance scheme.[50] In addition, health service quality and co-operative health insurance in Canada are equally concerned with profit generating activities.[51] The results from Canada and China were similar to our findings.
The health insurance industry in Nepal has been in existence for a long time, but coverage is still low and there are only a few successful CBHI models.The health insurance plans offered by co-operative groups are a newly growing strategic movement in health service in the country.With proper subsidies from the government and long-term contracts with advanced hospitals, health insurance offered by community groups could be more effective as an almost zero cost financing model and health equity and quality of service in Nepal could be more accessible to people.Nevertheless, there are some limitations in our study.There was not sufficient variables to compare co-operative and government health insurance.However, as a whole, CBHI has achieved limited success in terms of community participation, self-health care, and social unity.Thus, significant support is still necessary for both CBHI models as shown in the case of East Africa.[52] There are different minority groups in Nepal and inconsistence region by region.So, overall inclusiveness, religious minority and disadvantage Terai group are not equal proportion in both groups.The results of this study are from the supply side (government health facilities and community groups) and conclusions may differ when the demand side (consumers) is examined.Additionally, some information from the key people interviewed may be more subjective and thereby there is a risk in generalization.

Conclusion
Community health insurance programs have multiple and long-term impacts because they can reduce the financial pressure on the government and be responsible for maintaining health and empowering people.More importantly, it is applicable to reduce the out of pocket expenditure and reduce catastrophic health expenditures [53] for newly designated community groups in developing countries, especially those in remote areas and for people in urban areas who cannot afford private health care.Ultimately, successful CBHI models could be milestone to achieve Universal Health Care (UHC).
A): This group had six pilot areas, governed by the Ministry of Health and Population (MoHP) of Nepal, and based on a population surveyed more than five years prior.Primary health centres and district-level hospitals offered benefit packages of health insurance using their own management.They operated on government financing.

Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 6 August 2016 doi:10.20944/preprints201608.0068.v1 1
. Did you have any targets for new enrolment in the last year?1.1.If yes, did you have a special scheme for new enrolment?2.What was the amount of the benefit packages you provided to the enrolee?This checklist was prepared to identify the physical infrastructure, office setting, recording and reporting status, membership cards, registration, patient records, bank ledgers, and observation of income generation activities.Informal question-answer sessions, individual relationships, and breakfast and break time were utilized as techniques.
6.1.If yes, what kinds of agreements do you have (annually, bi-annually, etc.)? 7. Do you negotiate with health service providers within their standard price of treatment?7.1.If yes, how much (Rupee or percentage)?8. Do you have your own ambulance service to use during emergencies?Key informant interview guidelines: The key informant interview guidelines were used to assess management's experience with CBHI.The informants were focal person of government health institution and member secretary of coops.In particular, the problems and challenges were obtained from the key informant interview process 1.When did you initiate the CBHI? 2. What are the problems with your CBHI? 3. Who supports your group, and what kind of support do they provide?4. What is your expectation from the government?5. What suggestions do you have for the government and any supporting organizations? 6.How do you sustain the program in terms of governance and financial support?Observation checklist:

Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 6 August 2016 doi:10.20944/preprints201608.0068.v1 workloads
. The allocated budget was not recorded properly and used for administrative purposes (travelling and daily allowance).The experience from the representative focal person at the government health facilities was captured as follows: