http://ijhpm.com Int J Health Policy Manag 2018, 7(5), 443–454 doi 10.15171/ijhpm.2017.117 Original Article Factors That Influence Enrolment and Retention in Ghana’ National Health Insurance Scheme Agnes Millicent Kotoh1*, Genevieve Cecilia Aryeetey1, Sjaak Van der Geest2 Abstract Background: The government of Ghana introduced the National Health Insurance Scheme (NHIS) in 2004 with the goal Article History: of achieving universal coverage within 5 years. Evidence, however, shows that expanding NHIS coverage and especially Received: 27 February 2017 retaining members have remained a challenge. A multilevel perspective was employed as a conceptual framework and Accepted: 23 September 2017 ePublished: 17 October 2017 methodological tool to examine why enrolment and retention in the NHIS remains low. Methods: A household survey was conducted after 20 months educational and promotional activities aimed at improving enrolment and retention rates in 15 communities in the Central and Eastern Regions (ERs) of Ghana. Observation, in- depth interviews and informal conversations were used to collect qualitative data. Forty key informants (community members, health providers and district health insurance schemes’ [DHISs] staff) purposely selected from two case- study communities in the Central Region (CR) were interviewed. Several community members, health providers and DHISs’ staff were also engaged in informal conversations in the other five communities in the region. Also, four staff of the Ministry of Health (MoH), Ghana Health Service (GHS) and National Health Insurance Authority (NHIA) were engaged in in-depth interviews. Descriptive statistics was used to analyse quantitative data. Qualitative data was analysed using thematic content analysis. Results: The results show that factors that influence enrolment and retention in the NHIS are multi-dimensional and cut across all stakeholders. People enrolled and renewed their membership because of NHIS’ benefits and health providers’ positive behaviour. Barriers to enrolment and retention included: poverty, traditional risk-sharing arrangements influence people to enrol or renew their membership only when they need healthcare, dissatisfaction about health providers’ behaviour and service delivery challenges. Conclusion: Given the multi-dimensional nature of barriers to enrolment and retention, we suggest that the NHIA should engage DHISs, health providers and other stakeholders to develop and implement intervention activities to eliminate corruption, shortage of drugs in health facilities and enforce the compulsory enrolment stated in the NHIS policy to move the scheme towards universal coverage. Keywords: National Health Insurance (NHI), Enrolment, Retention, Drugs, Ghana Copyright: © 2018 The Author(s); Published by Kerman University of Medical Sciences. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Citation: Kotoh AM, Aryeetey GC, Van der Geest S. Factors that influence enrolment and retention in Ghana’ National *Correspondence to:Agnes Millicent Kotoh Health Insurance Scheme. Int J Health Policy Manag. 2018;7(5):443–454. doi:10.15171/ijhpm.2017.117 Email: amkotoh@ug.edu.gh Key Messages Implications for policy makers • Despite the fact that the National Health Insurance Scheme (NHIS) has increased access to healthcare, it has not eliminated financial barrier to accessing services. • Corrupt practices especially extra payment at health facilities and payment for drugs persist. • Attaining universal coverage require appropriate interventions to address the issues of delay in claim payment and shortage of drugs. • The National Health Insurance Authority (NHIA) should effectively engage national and local stakeholders to create systems to improve service delivery and ensure prompt payment of claims to enable health facilities meet insured patients’ drug needs. • The government should resource district health insurance schemes (DHISs) adequately to enable them deliver NHIS cards promptly. Implications for the public The National Health Insurance Scheme (NHIS) is an effective means to increase healthcare access in low- and middle-income countries including Ghana but effective implementation to ensure the poor, the target of the scheme, benefit is needed. Exemption should be expanded to include poor households with large membership. Also, healthcare service delivery should be improved and the NHIS made attractive to both the rich and the poor. Intensive education is also needed to make people appreciate the operational principles of the NHIS and develop positive attitudes that will move the scheme towards universal coverage. Though social health insurance generates fierce political debates, our results show that Ghanaians are putting politics aside to enrol in the scheme. However, the politics that erupts after every election need to be stopped by appointing the chief executive as a public servant and not a political appointee who should leave office with a change in government. Full list of authors’ affiliations is available at the end of the article. Kotoh et al Background district health insurance schemes (DHISs) were established Over the last two decades many developing countries have in all districts in the country. The NHIS became operational been experimenting with social health insurance schemes in March 2004 to replace cash-and-carry and ensure equity (SHISs) to improve healthcare access. In Ghana, the National in healthcare access. The NHIS’ goal is: “Within the next five Health Insurance Scheme (NHIS) was introduced in 2004 years, every resident in Ghana shall belong to a health insurance in response to the criticism that user-fees,[1] particularly scheme that adequately covers him or her against out-of-pocket cash and carry,[2] improved the quality of care and drug payment ….”9,10 The NHI Act specifies the specific diseases and supply but denied many people access, especially vulnerable healthcare services covered and not covered.11,12 However, the groups.1-4 insured could only access healthcare in accredited facilities This criticism led many organisations including International within the district where they registered. Labour Organisation (ILO) and the World Health The NHIS premium is subsidised by 2.5% value added tax Organization (WHO) proposing the establishment of a (VAT). The SSNIT contributors pay 2.5% of their contributions national health insurance (NHI) to improve healthcare as premium (Table 1). Non-SSNIT contributors are expected access in Ghana. The Provisional National Defence Council to pay an income adjusted premium of between GH₵22 (about (PNDC) responded to this suggestion by contracting experts US$10) and GH₵48 (about US$22) per adult per annum; but to make recommendations for creating a NHI organisation. in practice everybody pays the minimum. Exemptions are They proposed the establishment of a centralised company provided for vulnerable: children below 18 years, the elderly to provide a compulsory social health insurance for all Social (70 years and above) and the core poor. Pregnant women and Security and National Insurance Trust (SSNIT)[3] contributors mentally challenged persons were added to the exemption and registered cocoa farmers. Rural-based community- category in 2008 and 2012 respectively.13,14 finance schemes were also proposed for non-formal sector The NHIS, the first nation-wide scheme in Africa initiated by workers. These proposals culminated in the establishment a government, covers both formal and informal sector workers of community-based health insurance schemes (CBHISs) in and operates under a nationalised system of service provision the early 1990s with support of the Ministry of Health (MoH) and financing with no co-payment. Healthcare is obtained and international donors. The National Democratic Congress from all public health, faith-based, quasi-government and (NDC) government also launched a pilot NHIS in four some private health facilities, pharmacy and chemist shops districts in Eastern Region (ER) in 1997. This scheme, though that have been accredited and operate under contract with the stalled, stimulated debate on the need to find an alternative National Health Insurance Authority (NHIA). sustainable healthcare financing system.5 The government, The NHIA established and manages the National Health inspired by the modest success of some CBHISs continued the Insurance Fund (NHIF). Premium payment account for 5%, initiative of establishing a viable NHIS.5-7 But this vision was SSNIT 17.5%, VAT 73% and interest on investment 5.2% of curtailed by a change in government in January 2001. NHIF’ inflows.15 The NHIS has become the dominant source The main opposition party, New Patriotic Party (NPP), having of income for public health facilities; accounting for 79.4% of made a commitment during the 2000 election campaign to internally generated funds (IGF) in 2010.16 and reduces out- implement the NHI, initiated the policy-making process. of-pocket health expenditures (OOPHEs).17 Consequently, the NHI Law (Act 650) was passed in 20038 and Ghana is a lower middle income country with estimated Table 1. Current Features and Operational Principles of the NHIS Features Description - DHISs are centrally administered by the NHIA but day-to-day administration is decentralised to the districts. Administration - NHIA functions as the insurer; provides NHIS cards and accreditation to service providers, negotiates benefit packages, cost of care, ensures quality service and pays service providers. - 2.5% VAT. - 2.5% SSNIT contribution. Funding - Money allocated to the NHIF by Parliament.- Income from investments. - Premium from non-SSNIT contributors, registration and administrative fees. - Donations from non-governmental organisation and individuals. Membership - Membership is compulsory for all residents in Ghana; unless private health insurance membership can be shown . Exemptions - SSNIT pensioners, non-SSNIT contributors above 70 years, children below 18 years, pregnant women and mentally challenged persons. Benefit package - Covers 95% of diseases reported in health facilities in Ghana. Payment to service providers - Pay service providers within four weeks of claim submission to DHISs. - The NHIA regulates premium and registration fees. Supervision - Health facilities submit quarterly reports to the NHIA. - DHISs submit annual reports to the NHIA who audits their accounts. Abbreviations: DHISs, district health insurance schemes; NHIA, National Health Insurance Authority; VAT, value added tax; SSNIT, Social Security and National Insurance Trust; NHIF, National Health Insurance Fund; NHIS, National Health Insurance Scheme; NHIF, National Health Insurance Fund. 444 International Journal of Health Policy and Management, 2018, 7(5), 443–454 Kotoh et al gross domestic product (GDP) of US$38 617 billion.18 Tax chosen to reflect the ecology (forest and coastal) and the main constitute 14.9% of GDP in 2011.19 The proportion of GDP economic activity (agriculture) of Southern Ghana. The CR spent on health in 2014 wa 3.6%.18 The per capita expenditure and ER had 70% and 72.7% economically active population on health was US$145.18 Per the 2010 Population and Housing aged 15 years and above.20 Unemployment rate was 8.1% and Census, 71.1% of the population aged 15 years and above were 8.4%21 and poverty incidence (PI) 18.8% and 21.2% in the CR economically active.20 The poverty incidence (PI) in 2012 was and ER respectively.21 The literacy rate in the CR and ER were 24.2% and poverty gap index 7.8%.21 81% and 78.2% respectively.20 Access to healthcare has increased through the establishment Each region has a regional hospital. All districts used in this of community-based health planning and services study have a hospital, health centres, clinics and CHPS zones. (CHPS)[4] zones nation-wide. However, human resource Almost all communities fall within the 5-km radius proximity challenges persist. With the ratio of one doctor to 10 402 to a health facility. and one nurse to 1599 persons in 2011,22 and other service Some key health indicactors in the two regions for 2011 were: delivery challenges, quality service is often comprised23 and infant mortality (per 1000 live births) was 61 and 50, maternal the country continuous to experience poor health indicators. mortality (deaths/100 000 live births) 520 and 538, and total Infant mortality rate was 41 deaths per 1000 live births, under fertility rate was 3.6% in the CR and ER respectively.21 five mortality 60 deaths per 1000 live births in 201424 and maternal deaths (380 per 100 000 life births) in 2013.22 Life Selection of Participants and Quantitative Data Collection expectancy at birth in 2014 was 60.3 years and 62.5 year for Since a health economist on the research project has discussed males and females respectively. The literacy rate was 74.1%.20 the sample size calculation in detail elsewhere,35 we limit Though studies have consistently shown that NHIS provides ourselves to giving the sample size and the selection process, quick access to healthcare,25,26 only 34% of the population and discuss the qualitative method in this paper. The sample was enrolled in 2011.27 In the light of the relatively low NHIS size of 3000 households for the whole project was based on premium and exemptions for vulnerable groups, factors that 80% power to detect a 5% difference in overall enrolment undermine enrolment and retention in the scheme need to be between intervention and control communities and 10% was explored. added to cater for non-response rate; making the total sample Previous studies attribute low enrolment in SHISs to poverty, 3300 households (110 for each community).35 inadequate information about health insurance and perceived The survey sample was selected as follows. First, one census poor quality of service.28-31 These researchers who often enumeration area (EA) referred to as community in this paper focused on one stakeholder, usually community members was randomly selected from each of the 30 districts with DHIS and used only quantitative methods, gain limited information offices using computer generated random numbers. Second, about the phenomenon of low enrolment in SHISs. This study all households were listed and 110 randomly selected. examined factors that influence enrolment and retention A structured questionnaire was administered to heads in the NHIS using a multi-level perspective (MLP) which of households. The survey data was used to categorise emphasises engagement of stakeholders at different levels in the communities into 15 interventions and 15 controls; studying healthcare32-34 to more convincing explanation to the considering their socio-economic status (SES), NHIS status phenomenon of low enrolment and high drop-out rate. and rural/urban location. The word ‘level,’ a metaphor, refers to the international, This paper is based on the follow-up survey carried out in national, regional and local tiers of social organisation.33 March 2011 after exposing the intervention communities to Engaging actors and focusing on events at these levels is 20 monthly educational and promotional activities to ensure the objective of the MLP. The authors emphasised that the people have adequate knowledge of NHIS’s operational actors’ varied interest and events interact to affect a social principles and benefit package. Qualitative data was collected phenomenon. We thus propose that in order to draw more between June 2009 and September 2011. The details of convincing conclusions about why enrolment is low and intervention activities are reported elsewhere.36 retaining members in the NHIS a challenge, it is necessary to focus on both local and national stakeholders of the scheme, Quantitative Data Analysis namely: community members, health providers, and staff of We used crosstabs to analyse enrolment status of individual DHISs, the NHIA, Ghana Health Service (GHS) and MoH household members, reasons for enrolling, not renewing and use ethnographic tools in the study. The findings of this membership and never enrolling. Chi-square test was used study will not only inform policy and intervention activities to determine the relationship between NHIS status, socio- to move NHIS towards universal coverage, but also help other economic and perceived health status. developing countries address problems confronting their The health economists estimated the PI using household SHISs. detailed monthly consumption expenditure on food and non- food items and dwelling characteristics (eg, water supply and Methods availability of electricity etc).37,38 These indicators conform Study Sites to the definitions of poverty in the Ghana Living Standard This ethnographic study was conducted in the Central Region Survey (GLSS) 6 data.21 The GLSS is an internationally (CR) and ER of Ghana. The two regions were strategically accepted method for estimating PI in developing countries. International Journal of Health Policy and Management, 2018, 7(5), 443–454 445 Kotoh et al The estimated monthly household expenditure represented CR. Informal conversation, unstructured spontaneous the total amount of money needed to meet food and non-food discussions with participants, provides the opportunity to ask consumption requirements of households.21 These estimates pertinent questions on different occasions. This eliminates were considered as proxies for household wealth. Principal the possibility of participants adjusting their response on component analysis, a statistical procedure used to determine purpose or holding back information on sensitive issues that weights for a linear index of a set of variables was employed are critical to the study. to estimate households’ SES scores.39,40 The households were ranked into five wealth quintiles based on their SES scores Qualitative Data Analysis (core poor, poor, average, rich and very rich).37,38 All the Observations and conversations were jotted down and analysis was done in SPSS version 16. elaborated into field note books at the end of each day in line with standard ethnographic studies.41 Second, digitally Key Informants and Qualitative Data Collection recorded interviews were transcribed verbatim and subjected Qualitative data was collected in seven communities in the to content analysis to elicit the common themes emanating CR by the first author using observation, in-depth interviews from the data. Third, typical comments by participants were and informal conversations. She paid special attention to condensed into meaningful summary statements and placed two strategically selected communities: an urban fishing under the appropriate theme. Fourth, all authors reviewed the community and a rural farming community in the forest zone analysis to ensure that the themes and summary statements and their respective health facilities and DHIS offices as case reflect the participants’ views. studies (Table 2). The data was collected at multiple levels using both Forty local key informants were interviewed (20 from each case quantitative and qualitative methods in order to capture what study site: 11 community members, 7 health providers and 2 happens at various levels and uncover aspects of issues that DHIS’ staff). In addition, four national level key informants are not immediately obvious.42 In this study, triangulation (one each from the MoH and GHS headquarters and two was used to verify responses by asking different categories from the NHIA) were interviewed. The community members of participants the same questions using interviews and were cautiously selected to ensure they fairly represent the informal conversation to capture all dimensions of factors target population: currently insured, previously insured and that influence health insurance decisions. never insured. Educational level, SES and health status were Four main steps were taken to ensure validity and authenticity also considered. Key informants who were health providers, of the data collected. (1) The questionnaire and interview staff of DHISs, GHS, NHIA and MoH were purposely selected guide were translated using the back-translation method and based on their work schedule. pre-tested. (2) The first author conducted all interviews. (3) Several community members, health providers and DHISs’ Triangulation was used to verify responses. (4) Self-reflection, staff who were not regarded as key informants were engaged etic[5] and emic[6] perspectives were used to ensure objectivity in informal conversation in the seven communities in the when writing the reports. Table 2. Profile of Key Informants in Each Case Study Community and Results and Discussion National Level Background Characteristics of Households and Individuals Covered in the Survey Position/Employment Number Local Stakeholders (n = 20) Of the 6790 individuals covered by the survey, 46.3% were Fishermen or farmers 3 under 18 years and 4% were 70 years and above. Females Traders/farmers 2 formed 52.5% (Table 3). These statistics roughly agree with Community members Unemployed 1 the 2010 Population and Housing Census which reports that Traditional and Opinion Leaders 3 38.3% and 4.7% of the population were under 15 years and 65 Religious Leaders 2 years and above respectively.20 Staff Nurse 1 Community Health Nurse 1 Enrolment in the National Health Insurance Scheme and Health facility staff Midwife 1 Membership Renewal Medical Assistant 1 Community members typically described the NHIS as: Physician 1 ‘District Health Disease Control Officer 1 “A good arrangement that ensures everyone has access Directorate’ staff DDHSs 1 to healthcare.” However, this did not translate into high Public Relations Officer 1 enrolment and regular renewal of membership. Of the 6790 DHIS staff Manager 1 individuals covered by the survey, 40.3% were currently National Level Stakeholders (n = 4) insured and 22.4% previously enrolled. MoH 1 Significant differences were observed between enrolment of GHS 1 poor and rich respondents as well as the sick and healthy. NHIA 2 Lower enrolment was reported among the poor categories. Of Abbreviations: DHIS, District Health Insurance Scheme; MoH, Ministry the 1392 poorest individuals covered, 17.6% were currently of Health; GHS, Ghana Health Service; NHIA, National Health Insurance Authority; DDHSs, District Director of Health Services. insured compared to 44.4% out of the 1299 richest (P = .000). 446 International Journal of Health Policy and Management, 2018, 7(5), 443–454 Kotoh et al Table 3. Background Characteristics of Households and Individuals Benefits of the National Health Insurance Scheme Covered in the Survey Community members mentioned that the NHIS gives access Overall (All Individuals), N = 6790 Percent to healthcare and financial relief from catastrophic payments. Age, n = 6790 An insured diabetic patient described NHIS’ benefits as: “The 0-17 46.3 premium compared to the cost of healthcare is reasonable. I 18-69 49.7 don’t have to spend all my money paying hospital bills. Insurance 70+ 4.0 helps me get my drugs. I don’t have crises anymore.” Gender, n = 6790 Health providers confirmed these assertions and added that Male 47.5 they encouraged people to enrol because the NHIS reduces Female 52.5 complications among insured patients. A physician described Highest level of education of heads of households, n =1562 the NHIS’ benefits as: “One good thing about NHIS is that more None 26.3 people report to the hospital early with fewer complications and Primary 20.6 come for review regularly.” Secondary 44.4 Our survey results support these accounts and show that more Tertiary 8.7 than two-thirds of respondents across the five SES enrolled Marital status of heads of households, n = 1562 because the NHIS provides financial protection against ill Single 9.0 health (Table 5). Our results corroborate the findings that Married 62.0 the NHIS improved access to formal care, 25,26 significantly Divorce 6.8 reduced out-of-pocket payment (OOPP). 43 Durairaj et al Separated 5.5 observed a decline in hospital deaths among insured patients Widowed 12.5 owing to early treatment. 26 These evidence underscore the Cohabiting 4.2 importance of the NHIS as a safety net for residents in Ghana. Positive Health Provider-Patent Interaction Membership non-renewal rate was lower among the poor: Insured patients mentioned that they enrolled and remained poorest (15.4%) and richest (23.8%) (P = .000). in the NHIS because of some health providers’ positive Respondents with perceived poor health enrol and renew their behaviour towards them. One of them shared his experience membership compared to healthy respondents (P = .000). Out as follows: “I registered because the physician assistant (PA) of the 145 respondents perceived as having poor health, 73% persuaded me to enrol. The first thing they ask you is whether were currently insured and 10.8% previously insured while you have insurance card.” The PA confirmed this assertion 39.2% of the 6206 respondents with perceived good health and said: “I show patients my insurance card and educate them were currently insured and 22.9% previously insured (Table about the benefits to encourage them to enrol and renew their 4). membership regularly to enable them seek care early and avoid These statistics and qualitative data revealed factors that complications.” influence enrolment and retention in the NHIS are multi- These comments indicate that when health providers behave dimensional. The factors are discussed under two themes: favourably towards insured patients, people are encouraged enablers and barriers. to enrol. Thus, perceptions about quality of service ultimately influence people to enrol in SHISs.5 Enablers of Enrolment and Retention in the National Health Insurance Scheme Barriers to Enrolment and Retention in the NHIS at the Enablers are factors that inspired people to enrol and remain Community Level in the NHIS. These are benefits derived from the NHIS and Barriers are factors that discourage people from enrolling positive health provider-patient interaction. in the NHIS and renewing their membership regularly. The Table 4. NHIS Status by Socio-Economic and Perceived Health Status Currently Insured Previously Insured Never Insured P Overall N = 6790 40.3 22.4 37.3 Socio-economic categories N = 6790 Poorest 1392 (20.5) 17.6 15.4 67.0 .000 Poor 1362 (20.1) 31.3 18.4 50.3 Average 1336 (19.7) 35.0 22.1 42.9 Rich 1401 (20.6) 46.4 23.7 29.9 Richest 1299 (19.1) 44.4 23.8 30.9 Perceived health status N = 6788 Good health 6206 (92.9) 39.2 22.9 37.9 .000 Average health 339 (5.0) 56.3 14.2 29.5 Poor health 145 (2.1) 73.0 10.8 16.2 Abbreviation: NHIS, National Health Insurance Scheme. International Journal of Health Policy and Management, 2018, 7(5), 443–454 447 Kotoh et al Table 5. Reasons for Enrolling, not Renewing Membership and Never Enrolling in the NHIS by SES N = 1562 Poorest Poor Average Rich Richest Reasons for enrolling in the NHIS, n = 619 Financial protection against illness 71.0 79.1 77.5 74.1 72.4 It a better than cash and carry 29.0 20.9 21.1 25.4 26.5 The school insured my child 0 0 1.4 0 0.6 Community opinion leaders asked me to join 0 0 0 0 0.6 Employer paid 0 0 0 0.5 0 Reasons for not renewing membership in the NHIS, n = 319 Could not afford renewal payment 75.0 63.2 65.5 56.6 65.2 Not satisfied with service 25.0 5.3 6.9 10.5 13.0 Difficulty in accessing services 0 5.3 3.4 5.3 4.3 No transport money 0 0 0 0 1.3 Inappropriate timing of premium payment 0 5.3 3.4 1.3 0 Had to buy drugs outside facility 0 0 10.5 2.6 6.9 Did not use service last year 0 21.0 10.3 18.4 6.6 Others 0 0 0 5.3 2.7 Reasons for never enrolling in the NHIS, n = 608 Cannot afford premium 100 96.8 61.3 70.5 66.7 Covered elsewhere 0 0 0 2.3 3.7 Mostly healthy do not need to be ensured 0 0 16.1 12.4 13.1 No scheme in the area 0 0 0 0 1.7 No close facility in the area 0 0 0 1.1 0 No confidence in the scheme 0 3.2 12.9 10.3 3.7 Registration point too far 0 0 3.2 0 7.4 Others 0 0 6.5 3.4 3.7 Abbreviations: NHIS, National Health Insurance Scheme; SES, socio-economic status. appreciation of the NHIS’ benefits was not marked by high dependants, low enrolment and retention was an attitudinal enrolment and retention rates. Aside poverty, which is often issue and not a matter of poverty. An uninsured cocoa farmer reported in health economics literature as the main cause of said: “It is sad that most of us can pay the premium but we don’t. low enrolment,28-31 this study reveals more hidden factors. I registered my children and wife because they need healthcare.” These are influence of traditional risk-sharing arrangements, I also observed that heads of households knew that OOPP is corruption among health providers, service delivery more expensive but waited and rushed to enrol or renew their challenges and politics. membership for only members who needed healthcare. Also, the higher non-renewal rates among the average and rich Poverty/“No Money to Pay Premium” categories in this study contrasts health economics literature Though poverty contributes to low enrolment, it is only an in sub-Sahara Africa, which gives excessive weight to poverty important factor among the poorest and some poor households as the main factor for low coverage of health insurance with many members. When community key informants were schemes.44,45 questioned about why enrolment and retention in the NHIS Two explanations are here in place. First, ‘poverty’ is not only is low, ‘No money to pay premium’ was normally their first an issue of lack of money but also lack of control over one’s response. Our survey results confirmed the ‘No money to own life uncertainties. As a result, people wait until they need pay’ reason and show that 63.2% poor and 65.2% richest healthcare before enrolling. For example, the majority (65.2%) respondents did not renew their membership in the NHIS of the richest respondents said they could not renew their and 70.5% rich and 66.7% richest never enrolled because of membership because of poverty. Secondly, the ‘no money to ‘poverty’ (Table 5) but when I engaged key informants and pay’ statement had little to do with lack of money. It was a others who by local standards were not very poor in informal socially acceptable response. conversation, they gave additional reasons: ‘I’m not often sick’ That said, a critical analysis of the situation of some poor and “I’m waiting for a while.” Critical analysis of their living household heads who can pay the heavily subsidised premium conditions revealed that some poor household heads could but did not, revealed that the issue was beyond the premium. pay the heavily subsidised premium for all their members The poor do not have substantial income but due to the social but did not. I also observed that in many of the households, responsibility of caring for close relatives in Ghana, the poor only those who needed healthcare were registered and (usually the men) also have many dependents (their wives, renewed their membership regularly. So, I engaged them in children and other dependents) to enrol. Given their low discussions to explore their motives. Many participants stated income, enrolling all these dependents is unaffordable to that apart from the core poor and poor households with many them. A cocoa farmer and a father of six children (one above 448 International Journal of Health Policy and Management, 2018, 7(5), 443–454 Kotoh et al 18 years) and two dependents above 18 years explained that SHISs.48 But just like previous exemption policies which were he did not have money to pay for everybody so he enrolled his not successful,49-51 the purpose of the NHIS as a safety net has wife and four children who needed healthcare. failed to reach the poor and ensure that they have access to In addition, all the poor who were engaged in informal healthcare when sick. conversation complained that the extra payments for healthcare services and drugs discouraged them from Negative Influence of Traditional Risk-Sharing Arrangements renewing their membership. A fisherman said: ‘We struggle The fact that people quickly enrolled or renewed their to enrol but when we go to the hospital, we pay for drugs or membership when sick, indicates that they accept the NHIS as are given prescription to look for the drugs in accredited better than OOPP, but devised strategies to derive maximum pharmacy shops which we sometimes pay for. This discouraged benefit with minimal contribution. Almost all non-SSNIT many people from enrolling and renewing their membership.’ members including the richest pay the minimum premium. This supports the observation that compulsory or voluntary An uninsured cocoa farmer provided insight into the actual informal payment is a barrier to healthcare access for poor situation in the following comment: “Most of us can enrol and families; about 25% of healthcare users in Ghana pay illegal renew our membership during harvest time but we do not. We fees to public health providers.46 wait until we are seriously sick and rush to pay the minimum Our results show that poverty as a barrier to enrolment and premium.” I also met people who enrolled because they retention was most important only for the core poor. Some needed healthcare. During one of my routine observational of them did menial jobs and did not have regular income. visits to a DHIS office, I met a man who looked desperate and Others were totally unemployed and occasionally supported asked him what was wrong. He responded angrily: “Madam I by family members and neighbours. The majority (75%) of thought I didn’t need health insurance till I fell sick. Now I have previously enrolled and all never enrolled core poor said they registered and need the card for surgery but its delaying.” could not afford the cost of premium (Table 6). Appraisal Our survey results support these comments and show of the qualitative data revealed that some of them could not that those who perceive themselves as healthy enrol less even enrol their sick household members or take them to the and have higher drop-out rate: 39.2% with perceived good hospital. A health volunteer describes the situation of the core health were currently insured and 22.9% previously insured poor as: compared to 73% with poor health status being currently “Many of the core poor do not have a stable source of income. insured and 10.8% previously enrolled (P = .000) (Table 4). They are occasionally supported by family members and This corroborates Kusi and colleagues’ finding in their 2011 friends. Even to get one meal a day is a problem. They cannot study in three districts across Ghana that 73.9% of household afford the cost of premium. Meanwhile they are not given members whose perceived health status was poor were likely exemption.” to be insured compared to 49.2% with excellent health.31 A core poor woman explained why she could not enrol her Also, the 2014 nation-wide Demographic and Health Survey household members as follows: “I do menial jobs and have no (DHS) reported that 48% and 62% of men and women money to enrol my five children. One of them died because I respectively were currently enrolled. It must however be had no money to take her to the hospital.” Our results thus noted that the DHS covers only women of reproductive provide credible evidence that the core poor, who need health age (15-49 years) who enjoy free enrolment for pregnant insurance most, could genuinely not afford the premium and women. Therefore, the 62% enrolment for women is not so need exemption. However, the exemption is not reaching representative of all women. The survey also reported that 1% them. None of the core poor I engaged in conversation in the of the respondents were covered by other types of insurance.24 seven communities visited benefitted from exemption. When Blanchet et al also found that less than 25% of women under I questioned DHIS staff why they do not grant exemption to 30 years and about 45% over the age of 60 were enrolled in the core poor, one of them said: “We need money so if we go the NHIS.52 These results indicate that low enrolment is a about saying we want people to exempt they won’t pay and how country-wide phenomenon. The question then is: Is it really do we get money?” Witter and Garshong also reported that the case that 65.2% of richest previously enrolled and 66.7% only one per cent of the population were granted exemption richest never enrolled respondents genuinely could not pay in 2008.47 Our results corroborate WHO’s finding that the heavily subsidised premium? exemption is crucial to ensure that the poor are enrolled in All key informants attributed low enrolment and high drop- Table 6. Opinion Related to Quality of Service Currently Insured (n = 619) Previously Insured (n = 319) Never Insured (n = 608) N = 1562 Agree Neutral Disagree Agree Neutral Disagree Agree Neutral Disagree The insured still have to buy drugs 64.0 11.6 24.4 58.2 10.8 31.0 50.0 19.8 30.2 Attitude of health staff should be improved 76.6 14.7 8.7 87.0 9.6 3.4 74.7 18.3 7.0 Availability of drugs should be improved 83.7 11.2 5.1 90.1 7.1 2.8 78.6 17.7 3.7 Expect prompt treatment at the facility 75.8 11.4 12.8 68.8 18.5 12.7 64.7 21.6 13.8 International Journal of Health Policy and Management, 2018, 7(5), 443–454 449 Kotoh et al out rate to the negative influence of traditional risk-sharing drugs from the chemist shop. I can go and show you the arrangements used to manage livelihood activities and life nurse.” events. An analysis of the operations of two of such groups: I could not follow it up, but I asked a nurse from the facility ‘Pataase’[7] in fishing communities and ‘Nnoboa’[8] in farming to react to the complaint. She replied: “I won’t deny it. Some communities shows that though they have health insurance of us ask insured patients to pay cash because we don’t want elements, their risk-sharing principles are not the same. People to fill the complex insurance form.” In another incident, a join ‘Nnoboa’ when they need support on their farms and pull health volunteer called me to intervene and collect money out from the group until another farming season. Benefits are an insured patient was forced to pay. I followed-up to have commensurate to one’s contribution and not need. ‘Pataase’ concrete evidence to support the earlier reports I had received. focus on life events, mainly death. People join and remain in The provider gave ‘a face-saving’ explanation and quickly ‘Pataase’ because they are sure of benefitting. The benefits refunded the money. cover the funeral cost of members, their spouse, children and parents. A member argued: “Death is a certain event but for Payment for Drugs Inside and Outside Health Facilities ill health, you may or may not fall sick.” A community leader Paying for drugs was a common complaint among insured explained low enrolment in the NHIS and retention as follows: patients. A community member told me: “I had to pay for “Health insurance is not part of our culture. People join malaria drugs even though I know it is wrong.” Though some ‘Nnoboa’ if they need help in their farms. Benefits are health providers denied these allegations, others confirmed according to one’s contribution. You remain a member only the practice and lamented on how these undermined the when there is work to be done.” NHIS’ credibility. One of them said: “Some of us sell drugs These arguments show that though solidarity and reciprocity that are covered by the NHIS to insured patients, ‘pocket the are predominant features of both traditional risk-sharing money’ and charge the DHISs.” A medical officer confirmed arrangements and health insurance, the former do not the allegations and said: “One of my patients reported to the help convince people to join and remain in the NHIS when nurse that she was not given all her drugs but given prescription healthy. Platteau’s review of concepts underlying traditional to look for it outside. We followed-up and found that it was true, risk-sharing reveal that traditional mutual support schemes yet the DHIS was billed.” are based on balanced reciprocity (people receive as much To explore these allegations further, I asked DHIS officials for benefit as they contribute),53 while insurance is based on their reaction. One of them lamented: “Collecting illegal fees conditional reciprocity (members receive a return only if they and payment for drugs undermine our efforts. Some of the people fall sick). Our study shows that people’s reaction to health we struggle to enrol do not renew their membership because of insurance is influenced by the principles of traditional risk- the extra payments.” These assertions support the evidence sharing arrangements. The logic of not enrolling was that that payment of unauthorised fees has been a problem in the people perceive NHIS’ benefits as limited to the individual health sector in Ghana.23,55,56 Our study revealed that some so their investment might not benefit them. As contended, health providers only pursue their parochial interests and not a well-established cultural perspective limits the possibilities the achievement of NHIS’ goals. Our findings thus illuminate for thinking and acting in new situations.54 Our study the observation that corruption undermines achievement of also shows that existing knowledge and practices largely public policy goals.57 determine one’s reaction to new policies and not simply its benefits. The incentive for enrolling in the NHIS is largely Service Delivery Challenges at the District Health Insurance informed by the motive of benefitting and not of sharing Schemes cost so people enrol when sick and opt out when well. This Inadequate office accommodation, equipment and materials undermine the fundamental principle of health insurance; undermined the efficient functioning of DHISs. They were regular contribution into a common fund based on income, unable to deliver NHIS cards to their clients promptly. It was whether one benefits or not. common to see many clients waiting for hours at the DHIS offices to get their NHIS cards. A DHIS staff explained: Accusation of Corruption Among Health Providers “Inadequate equipment and registration materials make it Community members mentioned that health workers exploit difficult for us to deliver the cards to clients promptly.” insured patients. They cited illegal payments for drugs and Community members on their part expressed worry about other services. the delay in getting their cards. I engaged some people I met in DHISs’ offices looking frustrated in a conversation. One of Payment of Illegal Fees them told me: “This is the second time I’m coming here without Throughout the fieldwork, the issue of extra payments getting my card. I continue to pay at the hospital because of the by insured patients was prevalent. During an informal delay.” This study thus shows that service delivery challenges conversation, an insured woman narrated her experience at a do not only frustrate DHIS staff but also discourage people health facility as follows: from enrolling and remaining in the NHIS. “I went to the hospital in the evening because of a sudden The ongoing biometric registration to replace the old system stomach pain. The nurse refused to accept my insurance card is expected to introduce efficiency into the process but it because I was late and demanded cash. I left and bought seems not to solve the problem of delay in getting NHIS cards. 450 International Journal of Health Policy and Management, 2018, 7(5), 443–454 Kotoh et al Anecdotal reports and my observation reveal that people still of roaming looking for accredited pharmacy shops to obtain wait in long queues to register and do not sometimes get the prescribed drugs covered by the NHIS which they sometimes cards immediately as expected due to inadequate equipment pay for as follows: “We go around looking for prescribed drugs and shortage of materials. which we often pay for. Since we enrolled to avoid paying money when sick, these payments discourage us from renewing our Healthcare Service Delivery Challenges membership.” Our survey results corroborate these complaints. The assumption that the NHIS will improve quality of service About 83.7% of insured respondents and 90.3% of previously was not evident. Health providers’ heavy workload and insured agreed to the statement that availability of drugs at shortage of drugs on the NHI Drug List at health facilities health facilities should be improved (Table 6). undermined the quality of service. These results illustrate a significant misunderstanding regarding what insured patients, health providers and policy Heavy Workload and Long Waiting Time makers thought would stimulate enrolment and retain Many people, who had no access to formal care or cut their members. Health facilities anticipated prompt payment of treatment short because of user-fees,1-4 have access now. claims to enable them meet insured patients’ drug needs. Though an improvement in utilisation is desirable, it has Insured patients expect to receive all prescribed drugs at increased health providers’ workload. We found that the NHIS health facilities while policy-makers thought accredited shops was implemented within an overburdened health system could augment drug shortages at health facilities. without adequate resources to handle the growing patient Provision of drugs has been established to be critical in the numbers. Patients waiting for hours at health facilities was a appreciation of service delivery. Van der Geest et al write: common sight. Insured patients argued that they anticipate “Medical practitioners see pharmaceuticals as indispensable in prompt treatment, but they were rather made to wait longer their encounter with sick people... patients and their relatives than the uninsured. Their common complaint at busy health expect medicines to solve their problems.”59 This study also facilities was: “Health providers make us wait while they shows that drugs are a critical component of quality service attend to those who pay cash.” Our survey results confirmed and its shortage in health facilities increased dissatisfaction these assertions. The majority (75.8%) currently insured and among insured patients. This contradicts policy makers’ (68.8%) previously insured respondents said they expect assumption that the NHIS will improve quality of healthcare prompt treatment at health facilities. These results clearly and indicate that SHISs do not automatically lead to quality show that health facilities are arenas of social relations that service. This study thus supports the finding that health affect not only clients’ well-being58 but also health insurance insurance has weak or no effect on quality of service60 and decision making. that patients’ satisfaction about the quality of care determines Health providers were divided on the genuineness of insured the degree of participation in health insurance.61 patients’ complaints. A District Director of Health Services’ (DDHSs) argued: “Many more people come to the hospital and Politics and Enrolment in the National Health Insurance we spend time filling forms for insured patients; prolonging the Scheme time spent treating them. They don’t realise this and complain Social health insurance generates fierce political debates at about delays.” There were others who, though acknowledged international, national and local levels. In Ghana, the NHIS’s the increased workload, admitted insured patients’ concerns. political stake was very high during policy making and the One of them said: “These complaints are genuine. Some of us initial stage of implementation.62 Though, the political furor see insured patients as giving us extra work and give preference disappeared from public discourse after its introduction, it to uninsured patients.” still persists in subtle ways among national level stakeholders. Generally, clinicians are used to hurriedly writing a few words Participants who were NPP (the party in government) and so the additional task of filling the NHIS form, psychologically NDC (the main opposition party) sympathisers mentioned drew negative reaction towards insured patients. Some were that their decision to enrol or not to enrol was influenced hostile to insured patients; others collected unofficial fees or by the politics that surrounded NHIS’ introduction but this demanded cash payments to avoid filling the forms. changed later. An NDC supporter said: “I didn’t register when insurance was introduced because of Shortage of Drugs on National Health Insurance Drugs List politics. The NPP said all kinds of things about NDC who Shortage of drugs on the NHI Drug List undermine health first brought the idea. Though NDC could not implement providers’ desire to provide quality service to insured patients. health insurance, they should credit them for introducing the Health providers attributed the shortage to delays in claim idea.” payment which was expected to be within four weeks after One of the NPP sympathisers also explained why people claim submission to DHISs but this does not happen. A PA were not enrolling in the NHIS as follows: “Don’t mind NDC described the delay in payment as follows: “Only 60% of April supporters they thought that the NHIS will die when a new bill was paid in August and the remaining 40% this month government comes to power. Now they are registering because it [September]. I don’t know when May bill will be paid. All these benefits them.” Nobody mentioned that he did not have a valid make it impossible to meet patients’ drug needs.” NHIS card because of politics. Insured patients on their part, expressed their dissatisfaction However, national level key informants (staff of the MoH International Journal of Health Policy and Management, 2018, 7(5), 443–454 451 Kotoh et al and GHS) were concerned that politics is undermining the access expanded. Abiiro and Mcintyre observe: effective collaboration needed to develop efficient systems “Though the OTPP potentially can lead to increases in to improve quality of service, stimulate enrolment and retain NHIS coverage, especially within the informal sector... people in the NHIS. A GHS staff reacted to community sustainability will largely depend on how it is designed.... members’ and health providers’ complaints as follows: “These The government and the policy drivers need to... examine complaints are true. It is because we [MoH, GHS and the NHIA] its feasibility and long-term sustainability within the current are not meeting to develop systems to improve service delivery.” Ghanaian economic context.”66 The MoH staff I spoke with confirmed health providers’ The considerable controversy generated about OTPP’s complaints and argued that: implication for NHIS’ long term sustainability gradually “The biggest challenge we face is that the NHIA doesn’t share disappeared from public discourse possibly due to lack information. Even now that the politics that surrounded of a policy document in the public domain to justify its NHIS’ introduction is over, they still do not effectively engage contribution to achieving NHIS’ goals. us [MoH and GHS] to build systems to improve service Limitation of the Study delivery. They take decsions and inform us later.” The study covered two regions in Southern Ghana. Since the The NHIA staff also responded to these complaints as follows: socio-economic and cultural differences between the South “The NHIS is pro-poor. The premiums is low enough for and the North might affect health insurance decision making all Ghanaians to enrol. Exemptions are also provided. But, differently, our conclusions should be interpreted with health providers’ corrupt practices result in delays in claims caution. reimbursement and shortage of drugs. Enrolment could be improved if facilities help in the efficient management of their Conclusion stock levels and service providers improve their attitudes The study revealed that the NHIS’ implementation arena is towards insured patients.” littered with multi-dimensional factors located at multiple Responding to the question why the NHIA does not share levels. People enrolled because of NHIS’ benefits and positive information and meet regularly with the MoH and DHS health provider-patient interaction. Apart from the core poor officials, Another staff replied: “They read politics into and poor households with many dependants, poverty was whatever we do. We inform them about what we do. They not a reason for not enrolling in the NHIS and renewing delay payment of claims is to ensure that claims are thoroughly membership, instead, the negative influence of traditional checked.” risk-sharing arrangements, corruption, shortage of drugs, and A critical analysis of these comments and previous studies politics are the serious challenges that need to be addressed. reveal that the NHIS is highly centralised; involvement We thus suggest the following interventions to improve of stakeholders and technocrats in the implementation is enrolment and retention rates in the NHIS. minimal.7,60 and the checks and balances needed to address The evidence that insurance reduces medical complications, gaps was often missing. As Brinkerhoff observes after but inundated with corruption, requires the establishment analysing the US health system, effective engagement of of a national health system free of cash transaction and all stakeholders ensures accountability and helps reveal gaps that residents in Ghana compelled to belong to a health insurance require intervention to improve service delivery63 However, in scheme. The compulsory enrolment would be acceptable the case of Ghana, effective collaboration among stakeholders if the NHIS resonates with quality care and makes it easier to ensure independent decision making and development of and faster for the insured to access healthcare and not efficient systems is often sacrificed for political actors control the reverse. This requires that the NHIA engage local and of the NHIS. national stakeholders to create systems that improve service Also, the politics of the NHIS came up four years into NHIS’ delivery, prompt payment of claims to enable health facilities implementation when the opposition NDC also made two meet insured patients’ drug requirements and stop corrupt pledges in its manifesto for the next general elections in 2008 practices. Also, the government should resource DHISs as follows: adequately to enable them deliver NHIS cards promptly to “Our universal health insurance scheme will guarantee make NHIS attractive to both the rich and the poor. Above all, access to free healthcare in all public health institutions. the politics that erupts after every election should be stopped It will not be district-specific and will allow for one-time by appointing NHIA’s chief executive as a public servant and premium payment (OTPP).”64,65 not a political appointee who leaves office when there is a Upon resumming office in January 2009, the government was change in government. These measures could move NHIS able to nationalise the NHIS card and the insured now access towards the universal coverage that was announced in 2004. healthcare in every accredited facility irrespective of where Finally, we propose more qualitative study to explore further they registered. The OTPP which means that individuals will the effect of traditional mutual support arrangements and pay premium once in their life time implies the elimination shortage of drugs on enrolment and retention in health of annual premium, was fiercely debated and its implications insurance. for NHIS’ long term financial sustainability questioned. The government on his part argued that the high cost of premium Ethical issues The research protocol was approved by the Ghana Health Service Ethic collection from the informal sector can be surmounted and Committee (ID No. GHS-ERC: 12-1-09). Informed consent was obtained from 452 International Journal of Health Policy and Management, 2018, 7(5), 443–454 Kotoh et al all participants. Anonymity and confidentiality were guaranteed. 8. Government of Ghana. The National Health Insurance Act: Act 650. Accra: Ghana Government; 2003. Competing interests 9. Ministry of Health. Policy framework for the establishment of Authors declare that they have no competing interests. health insurance in Ghana. Accra: Ministry of Health; 2002. Authors’ contributions 10. Ministry of Health. National Health Insurance Policy Framework AMK, GCA, and SVG were involved in the conceptualisation and study design. for Ghana. Revised edition. Accra: Ministry of Health; 2004. The quantitative data collection and analysis was done by AMK and GCA. AMK 11. National Health Insurance Scheme website. http://www.nhis.gov. was solely responsible for the qualitative data collection and its analysis and gh/benefits.aspx. Accessed September 3, 2016. wrote the first draft paper. The other authors provided critical comments during 12. Dzakpasu S, Soremekun S, Manu A, et al. Impact of free the analysis and the interpretation of the data and revising the earlier draft. All delivery care on health facility delivery and insurance coverage authors read and approved the final manuscript. in Ghana’s Brong Ahafo Region. PLoS One. 2012;7(11):e49430. doi:10.1371/journal.pone.0049430 Authors’ affiliations 13. Government of Ghana. National Health Insurance Act: Act 852. 1School of Public Health, University of Ghana, Legon, Ghana. 2Department Accra: Government of Ghana; 2012. of Sociology and Anthropology, University of Amsterdam, Amsterdam, The 14. National Health Insurance Regulations. http://www.ilo.org/gimi/ Netherlands. RessShowRessource.do?ressourceId=11967. Accessed June Endnotes 2, 2008. Published 2004. [1] User-fees refer to OOPPs for some healthcare services at the point of 15. National Health Insurance Authority. NHIA Annual Report 2009. utilisation. Accra: National Health Authority; 2010. [2] Cash and carry led to OOPP for full cost of drugs in public health facilities. 16. World Health Organisation. The World Health Report 2010: It was a WHO and UNICEF initiative adopted by African Health Ministers in Health Systems Financing: The path to universal coverage. Bamako, Mali, in 1987. The policy was expected to improve drug supplies in Geneva: World Health Organization; 2010. public health facilities. 17. World Health Organization Global Health Expenditure database [3] Social Security and National Insurance (SSNIT) is a government pension website. http://apps.who.int/nha/database. Accessed May 20, scheme in Ghana that most formal sector workers and their employers 2017. contribute to. [4] 18. Life expectancy at birth, total (years). The World Bank CHPS zone is a national programme of community-based care provided by resident nurses who are referred to as community health officers. CHPS, website. http://data.worldbank.org/indicator/SP.DYN.LE00. introduced in 1999, reduces geographical barriers to access to healthcare and IN?locations=GH. Accessed May 31, 2017. provides basic level preventive and curative services for minor ailments at the 19. Tax revenue (% of GDP). The World Bank website. community and household levels. h t tps : / /da ta .wor ldbank.org / ind ica tor /GC.TAX.TOTL. [5] Etic account is a description of a phenomenon in terms of its meaning to the GD.ZS?locations=GH&view=chart. Accessed September 6, observing outsider. 2 0 1 7 . [6] Emic perspectives means describing behaviours and understandings in terms 20. Ghana Statistical Service. 2010 Population and Housing of meaningful experiences to the actor. Census: National Analytical Report. Accra: Ghana Statistical [7] Pataase is an association of mostly fishermen who come together to support Service; 2013. each other in times of economic stress. They provide financial assistance to 21. Ghana Statistical Service. Trends and patterns of poverty in members. Their main focus is funeral costs of members or their close relatives. [8] Nnoboa (support weeding) is a mutual help group. The name refers to local Ghana 2012-2013: Report of Ghana Living Standards Survey. support networks formed by farmers to support each other on their farms. In Accra: Ghana Statistical Service; 2007. these groups, trust and commitment to achieve individual and group objectives 22. World Health Organisation Country Office For Ghana Annual is the key principle. Report 2014. http://www.afro.who.int/sites/default/files/2017-05/ ghana-annual-report-2014-final.pdf. Assessed October 4, 2017. References 23. Agyepong IA, Nagai RA. “We charge them; otherwise we 1. Nyonator F, Kutzin J. Health for some? The effects of user fees in cannot run the hospital” front line workers, clients and health the Volta Region of Ghana. Health Policy Plan. 1999;14(4):329- financing policy implementation gaps in Ghana. Health Policy. 341. 2011;99(3):226-233. doi:10.1016/j.healthpol.2010.09.018 2. Asenso-Okyere WK, Anum A, Osei-Akoto I, Adukonu A. Cost 24. Ghana Statistical Service, Ghana Health Service, ICF recovery in Ghana: are there any changes in health care seeking International. Ghana Demographic Health Survey 2014. behaviour? Health Policy Plan. 1998;13(2):181-188. Rockville, Maryland, US: GSS, GHS and ICF International; 2015. 3. Waddington CJ, Enyimayew KA. A price to pay: The impact of 25. Mensah J, Oppong JR, Schmidt CM. Ghana’s National Health user charges in the Ashanti-Akim district, Ghana. Int J Health Insurance Scheme in the context of the health MDGs: an Plann Manage. 1989;4(1):17-47. doi:10.1002/hpm.4740040104 empirical evaluation using propensity score matching. Health 4. Waddington CJ, Enyimayew KA. A price to pay, part 2: the impact Econ. 2010;19 Suppl:95-106. doi:10.1002/hec.1633 of user charges in the Volta region of Ghana. Int J Health Plann 26. Durairaj V, D’ameida S, Kirigia J. Ghana’s approach to social Manage. 1990;5(4):287-312. doi:10.1002/hpm.4740050405 health protection. World Health Report Background, Paper No. 5. Arhinful DK. The Solidarity of Self-interest: Social and Cultural 2. Geneva: WHO; 2010. Feasibility of Rural Health Insurance in Ghana. Leiden: African 27. National Health Insurance Authority Annual Report 2011, 2012. Studies Centre; 2003. Status of implementation of the NHIS. http://www.nhis.gov.gh/ 6. Arhin-Tenkorang D. Health insurance for the informal sector in files/annualreport2011.pdf. Assessed September 2015. Africa: Design features, risk protection and resource mobilisation. 28. Asante F, Aikins M. Does the NHIS cover the poor? Accra: Geneva: Commission on Macroeconomics and Health, World DANIDA Health Sector Support Office; 2008. Health Organization; 2001. 29. Sarpong N, Loag W, Fobil J, et al. National health insurance 7. Agyepong IA, Adjei S. Public social policy development and coverage and socio-economic status in a rural district of Ghana. implementation: a case study of the Ghana National Health Trop Med Int Health. 2010;15(2):191-197. doi:10.1111/j.1365- Insurance scheme. Health Policy Plan. 2008;23(2):150-160. 3156.2009.02439.x doi:10.1093/heapol/czn002 30. Nguyen H, Knowles J. Demand for voluntary health insurance in International Journal of Health Policy and Management, 2018, 7(5), 443–454 453 Kotoh et al developing countries: the case of Vietnam’s school-age children perspectives. Geneva: WHO; 2003. and adolescent student health insurance program. Soc Sci Med. 49. Badasu DM. Implementation of Ghana’s health user fee policy 2010;71(12):2074-2082. doi:10.1016/j.socscimed.2010.09.033 and the exemption of the poor: problems and prospects. African 31. Kusi A, Enemark U, Hansen KS, Asante FA. Refusal to enrol in Population Studies. 2004;19:285-302. Ghana’s National Health Insurance Scheme: is affordability the 50. Aikins M, Arhinful D. Review of exemption policy: A report on problem? Int J Equity Health. 2015;14:2. doi:10.1186/s12939- the annual health sector review 2005. Ghana: Ministry of Health; 014-0130-2 2006. 32. Baer HA, Singer M, Johnsen JH. Toward a critical medical 51. Kanchebe Derbile E, van der Geest S. Repackaging exemptions anthropology. Soc Sci Med. 1986;23(2):95-98. under National Health Insurance in Ghana: how can access 33. Van der Geest S, Speckmann JD, Streefland PH. Primary health to care for the poor be improved? Health Policy Plan. care in a multi-level perspective: towards a research agenda. 2013;28(6):586-595. doi:10.1093/heapol/czs098 Soc Sci Med. 1990;30(9):1025-1034. 52. Blanchet NJ, Fink G, Osei-Akoto I. The effect of Ghana’s National 34. Press I. Levels of explanation and cautions for a critical clinical Health Insurance Scheme on health care utilisation. Ghana Med anthropology. Soc Sci Med. 1990;30(9):1001-1009. J. 2012;46(2):76-84. 35. Jehu-Appiah C. Reaching the poor in Ghana’s National Health 53. Platteau JP. Mutual insurance as an elusive concept in Insurance Scheme: Equity aspects and strategies to improve traditional rural communities. J Dev Stud. 1997;33(6):764-796. enrolment [PhD Thesis]. Nijmegen: Radboud University; 2012. doi:10.1080/00220389708422495 36. Kotoh AM. Improving health insurance coverage in Ghana: a 54. van der Geest S. Is paying for health care culturally acceptable case study. Leiden: African Studies Centre; 2013. in Sub-Sahara Africa? Money and tradition. Soc Sci Med. 37. Jehu-Appiah C, Aryeetey G, Spaan E, Agyepong I, Baltussen R. 1992;34(6):667-673. Efficeincy, equity and feasibility of strategies to identify the poor: 55. Agyepong IA. Reforming health service delivery at district level An application to premium exemptions under National Health in Ghana: the perspective of a district medical officer. Health Insurance Scheme in Ghana. Health Policy Plan. 2010;95:166- Policy Plan. 1999;14(1):59-69. 173. 56. Aikins M, Okang G. Utilisation and cost of healthcare under 38. Aryeetey GC, Jehu-Appiah C, Spaan E, D’Exelle B, Agyepong district health insurance: A case study of Brong Ahafo and I, Baltussen R. Identification of poor households for premium Eastern Regions. Accra: JSA Consultants; 2006. exemptions in Ghana’s National Health Insurance Scheme: 57. Aryee JRA. Public sector management in Africa. Economic empirical analysis of three strategies. Trop Med Int Health. research working paper series. No. 82. Tunisia: African 2010;15(12):1544-1552. doi:10.1111/j.1365-3156.2010.02663.x Development Bank; 2005. 39. McKenzie DJ. Measuring inequality with asset indicators. J Popul 58. van der Geest S, Finkler K. Hospital ethnography: introduction. Econ. 2005;18(2):229-260. doi:10.1007/s00148-005-0224-7 Soc Sci Med. 2004;59(10):1995-2001. doi:10.1016/j. 40. Vyas S, Kumaranayake L. Constructing socio-economic status socscimed.2004.03.004 indices: how to use principal components analysis. Health Policy 59. van der Geest S, Whyte SR, Hardon A. The anthropology of Plan. 2006;21(6):459-468. doi:10.1093/heapol/czl029 pharmaceuticals: A biographical approach. Annu Rev Anthropol. 41. Emerson RM, Frits RI, Shaw LL. Writing ethnography field notes. 1996;25:153-178. doi:10.1146/annurev.anthro.25.1.153 Chicago: University of Chicago Press; 2005. 60. Ekman B. Community-based health insurance in low-income 42. Graneheim UH, Lundman B. Qualitative content analysis in countries: a systematic review of the evidence. Health Policy nursing research: concepts, procedures and measures to Plan. 2004;19(5):249-270. achieve trustworthiness. Nurse Educ Today. 2004;24(2):105- 61. D’Ambruoso L, Achadi A, Adisasmita A, Izati Y, Makowiecka K, 112. doi:10.1016/j.nedt.2003.10.001 Hussein J. Assessing quality of care provided by Indonesian 43. Sulzbach S, Garshong B, Owusu-Banahene G. Effect of National village midwives with a confidential enquiry. Midwifery. Health Insurance Act in Ghana: Baseline report. Bethesda: ABT 2009;25(5):528-539. doi:10.1016/j.midw.2007.08.008 associates Inc; 2005. 62. Rajkotia Y. The political development of the Ghanaian National 44. Jütting JP. Do Community-based Health Insurance Schemes Health Insurance System: Lessons in health governance. Improve Poor People’s Access to Health Care? Evidence From Bethesda, MD: Abt Associates Inc; 2007. Rural Senegal. World Dev. 2004;32(2):273-288. doi:10.1016/j. 63. Brinkerhoff DW. Accountability and health systems: toward worlddev.2003.10.001 conceptual clarity and policy relevance. Health Policy Plan. 45. Basaza R, Criel B, Van der Stuyft P. Community health insurance 2004;19(6):371-379. doi:10.1093/heapol/czh052 in Uganda: why does enrolment remain low? A view from 64. National Democratic Congress. NDC Manifesto. A better Ghana: beneath. Health Policy. 2008;87(2):172-184. doi:10.1016/j. Investing in people, jobs and the economy. 2nd ed. Accra: healthpol.2007.12.008 National Democratic Congress; 2008. 46. Save the Children U.K. Helping children survive. Supporting poor 65. National Health Insurance Authority. Delivery on the NHIS families to overcome barriers to maternal, new born and child promise of one-time premium payment, financial sustainability health services. 2008. http://www.savethechildren.org.uk/sites/ of the NHIS based on recent financial assessment of the NHIS. default/files/docs/Helping_Children_Survive_1.pdf. Accessed A presentation to stakeholders Elmina. Ghana: National Health March 2, 2012. Published 2008. Insurance Authority; 2010. 47. Witter S, Garshong B. Something old or something new? 66. Abiiro GA, McIntyre D. Universal financial protection through Social health insurance in Ghana. BMC Int Health Hum Rights. National Health Insurance: a stakeholder analysis of the 2009;9:20. doi:10.1186/1472-698x-9-20 proposed one-time premium payment policy in Ghana. Health 48. World Health Organisation. Community-based health insurance Policy Plan. 2013;28(3):263-278. doi:10.1093/heapol/czs059 schemes in developing countries: Facts, problems and 454 International Journal of Health Policy and Management, 2018, 7(5), 443–454