Health insurance and health-seeking behavior: evidence from a randomized community-based insurance rollout in rural Burkina Faso

In 2004, a community-based health insurance (CBI) scheme was introduced in Nouna district, Burkina Faso, with the primary objective of improving access to facility-based health services. In order to overcome self-selection issues in the analysis of the behavioral effects of insurance, we combine four waves of the Nouna Health District Household Survey into a panel data set, and use the randomized timing of insurance rollout to estimate the causal effect of insurance coverage on health-seeking behavior. While we find a generally positive association between CBI affiliation and treatment seeking, we cannot reject the null that the introduction of health insurance does not have any effect on treatment seeking in general, and utilization of facility-based professional care, in particular. Low levels of health care provider satisfaction, poor perceived quality of care by enrollees, and ambiguity in the coverage level of the CBI benefit package appear to have contributed to these weak results. Our findings imply that the basic notion of insurance mechanically increasing facility-based professional care is not necessarily true empirically, and likely contingent on a large number of contextual factors affecting health-seeking behavior within households and communities.     

The feasibility of community-based health insurance in Burkina Faso

To ensure the acceptability of community-based insurance (CBI) by the community and its sustainability, a feasibility study of CBI was conducted in Burkina Faso, including preference for benefit package of CBI, costing of health services, costing of the benefit package and willingness-to-pay (WTP) for the package. Qualitative methods were used to collect information about preferences for the benefit package. Cost per unit health services, health demand obtained from household survey and physician-judged health needs were used to estimate the cost of the benefit package. The bidding game method was used to elicit household head's WTP for the package. We found that there were strong preferences for inclusion of high-cost healthservices such as operation, essential drugs and consultation fees in the benefit package. It is estimated that the cost of the package per capita was 1673 CFA (demand-based) and 9630 CFA (need-based), including 58% government subsidies (euro 1 = 655 CFA). The average household head with eight household members agreed to pay from 7500 (median) to 9769 CFA (mean) to join the CBI for his/her household. The WTP results were influenced by household characteristics, such as location, household size and age composition. Under certain assumptions (household as the enrolment unit, median household head's WTP as premium for the average household, 50% enrolment rate), it would be feasible to run CBI in Nouna, Burkina Faso if enrolees' health demand did not increase by more than 28% or if the underwriting of the initial losses was covered by extra funds.     

Inequality in willingness-to-pay for community-based health insurance

The purpose was to provide information for devising community-based health insurance (CBI) policies that reduce inequality in enrolment and further inequality in access to health services. A two-stage cluster sampling was used in the household survey. Inequalities in willingness-to-pay (WTP) for CBI are examined by expenditure quintile using data collected from a household survey. Interviews were conducted with 2414 individuals, 705 of whom were household heads. A bidding game method was used to elicit WTP. Individuals and households were assigned to 6-month expenditure quintiles. We found that mean and median individual WTP for CBI was significantly higher for higher spending quintiles, as was mean and median household WTP. The curves of cumulative percentage of individual and household WTP shifted rightwards for higher quintiles, implying that at any given premium the lower the quintile the lower the uptake of CBI. The Gini coefficient for individual WTP and household WTP was 0.15 and 0.08, respectively, and for individual 6-month expenditure and household 6-month expenditure is 0.68 and 0.63, respectively. The results imply that the premium needs to be adjusted for income; otherwise, a lower proportion of poor people will enrol in CBI and without exemptions or subsidies the poor will have less access to health services than the rich. Thus, exemptions and subsidies for the poor for enrolling in CBI are an important issue for decision-makers with an objective of improving equity of health and helping the poor to break out of the cycle of poverty. Since the distribution of WTP by household is less unequal than the distribution of WTP by individuals, the household might be a better unit of enrolment in terms of equity than the individual.     

Gender's effect on willingness-to-pay for community-based insurance in Burkina Faso

The purpose was to study gender's effect on willingness-to-pay (WTP) for community-based insurance (CBI) in order to provide information for deciding enrolment unit and setting premium in Burkina Faso. A two-stage cluster sampling was used in the household survey, with each household having the same probability of being selected. One thousand one hundred and seventy-eight men and 1236 women in the 800 households were interviewed. The bidding game approach was used to elicit WTP. We found that compared to male, female had less education, lower income and expenditure, less episodes of diseases and lower ratio of becoming household head, but higher marriage rate. These characteristics influenced the WTP difference between men and women. Men were willing to pay 3666 CFA ($4.89) to join CBI, 928 CFA higher than women were. Education and economic status positively influenced WTP, implying higher years of schooling and economic status and higher WTP. Age and distance to health facility negatively influenced WTP, thus higher age and longer distance and less WTP. Based on the results from this study, we suggest that CBI should be enrolled on the basis of households or villages in order to protect vulnerable persons, such as the aged, women and the poor. In setting premium a policy-maker needs to take into account costs of the CBI benefits package, possible subsidies from government and other agencies and WTP information. WTP should never be taken as a premium because it only provides some information for the respondents' financial acceptability for a certain benefits package.     

Applying the net-benefit framework for assessing cost-effectiveness of interventions towards universal health coverage

ABSTRACT: In assessing the cost-effectiveness of an intervention, the interpretation and handling of uncertainties of the traditional summary measure, the Incremental Cost Effectiveness Ratio (ICER), can be problematic. This is particularly the case with strategies towards universal health coverage in which the decision makers are typically concerned with coverage and equity issues. We explored the feasibility and relative advantages of the net-benefit framework (NBF) (compared to the more traditional Incremental Cost-Effectiveness Ratio, ICER) in presenting results of cost-effectiveness analysis of a community based health insurance (CBHI) scheme in Nouna, a rural district of Burkina Faso. Data were collected from April to December 2007 from Nouna's longitudinal Demographic Surveillance System on utilization of health services, membership of the CBHI, covariates, and CBHI costs. The incremental cost of a 1 increase in utilization of health services by household members of the CBHI was 433,000 XOF ($1000 approximately). The incremental cost varies significantly by covariates. The probability of the CBHI achieving a 1% increase in utilization of health services, when the ceiling ratio is $1,000, is barely 30% for households in Nouna villages compared to 90% for households in Nouna town. Compared to the ICER, the NBF provides more useful information for policy making.     

"To enrol or not to enrol?": A qualitative investigation of demand for health insurance in rural West Africa

In spite of the fact that feeble levels of participation have long been identified as a major constraint to the successful long-term implementation of community-based health insurance (CBI) in low-income countries, evidence on determinants of enrolment in CBI is still lacking. The application of econometric modelling has provided a partial answer to the question, but on its own it has proved to be insufficient to guide policy making. This paper aims to fill this gap in knowledge using qualitative research methods. In-depth interviews with 32 household heads were conducted in the Nouna Health District, Burkina Faso, West Africa to assess determinants of enrolment in a newly established CBI scheme. The findings highlight that factors previously neglected in the literature, such as institutional rigidities and socio-cultural practices, play an important role in shaping the decision to enrol. The discussion of the findings focuses on the policy implications, offering concrete recommendations to maximise enrolment, within and beyond Burkina Faso.     

Health insurance enrolment in the Upper West Region of Ghana: Does food security matter?

Toward achieving universal health coverage, Ghana's national health insurance has been acclaimed as a pro‐poor scheme, yet been criticized for leaving the poor behind. Arising from this is how poverty has been operationalized and how poor people are targeted for enrolment into the scheme. We examine the role of food insecurity (not currently considered) as a multidimensional vulnerability concept on enrolment into Ghana's health insurance using binary logistics regression on cross‐sectional survey of household heads (n = 1438) in the Upper West Region of Ghana. Our analyses show that heads of severely food‐insecure households were significantly less likely to enroll in national health insurance scheme (NHIS) relative to households who reported being food‐secure (OR = 0.36, P < .05). We also found education, occupation, and religion as significant predictors of health insurance enrolment. Based on our findings, it is crucial to incorporate food security status in the identification of vulnerable people for free enrolment in Ghana's health insurance.     

The differences in characteristics between health-care users and non-users: implication for introducing community-based health insurance in Burkina Faso

The purposes of this study are to describe the characteristics of different health-care users, to explain such characteristics using a health demand model and to estimate the price-related probability change for different types of health care in order to provide policy guidance for the introduction of community-based health insurance (CBI) in Burkina Faso. Data were collected from a household survey using a two stage cluster sampling approach. Household interviews were carried out during April and May 2003. In the interviewed 7,939 individuals in 988 households, there were 558 people reported one or more illness episodes; two-thirds of these people did not seek professional care. Health care non-users display lower household income and expenditure, older age and lower perceived severity of disease. The main reason for choosing no-care and self-care was ‘not enough money’. Multinomial logistic regression confirms these observations. Higher household cash-income, higher perceived severity of disease and acute disease significantly increased the probability of using western care. Older age and higher price-cash income ratio significantly increased the probability of no-care or self-care. If CBI were introduced the probability of using western care would increase by 4.33% and the probability of using self-care would reduce by 3.98%. The price-related probability change of using western care for lower income people is higher than for higher income although the quantity changed is relatively small. In conclusion, the introduction of CBI might increase the use of medical services, especially for the poor. Co-payment for the rich might be necessary. Premium adjusted for income or subsidies for the poor can be considered in order to absorb a greater number of poor households into CBI and further improve equity in terms of enrolment. However, the role of CBI in Burkina Faso is rather limited: it might only increase utilisation of western health care by a probability of 4%.     

The impact of reducing financial barriers on utilisation of a primary health care facility in Rwanda

This study investigates the impact of subsidising community-based health insurance (mutuelle) enrolment, removing point-of-service co-payments, and improving service delivery on health facility utilisation rates in Mayange, a sector of rural Rwanda of approximately 25,000 people divided among five 'imidugudu' or small villages. While comprehensive service upgrades were introduced in the Mayange Health Centre between April 2006 and February 2007, utilisation rates remained similar to comparison sites. Between February 2007 and April 2007, subsidies for mutuelle enrolment established virtually 100% coverage. Immediately after co-payments were eliminated in February 2007, patient visits levelled at a rate triple the previous value. Regression analyses using data from Mayange and two comparison sites indicate that removing financial barriers resulted in about 0.6 additional annual visits for curative care per capita. Although based on a single local pilot, these findings suggest that in order to achieve improved health outcomes, key short-term objectives include improved service delivery and reduced financial barriers. Based on this pilot, higher utilisation rates may be affected if broader swaths of the population are enrolled in mutuelle and co-payments are eliminated. Health leaders in Rwanda should consider further studies to determine if the impact of eliminating co-payments and increasing subsidies for mutuelle enrolment as seen in Mayange holds at greater levels of scale. Broader studies to better elucidate the impact of enrolment subsidies and co-payment subsidies on utilisation, health outcomes, and costs would also provide policy insights.     

Catastrophic household expenditure for health care in a low-income society: a study from Nouna District, Burkina Faso

OBJECTIVE: To quantify the extent of catastrophic household health care expenditure and determine the factors responsible for it in Nouna District, Burkina Faso. METHODS: We used the Nouna Health District Household Survey to collect data on 800 households during 2000-01 for our analysis. The determinants of household catastrophic expenditure were identified by multivariate logistic regression method. FINDINGS: Even at very low levels of health care utilization and modest amount of health expenditure, 6-15% of total households in Nouna District incurred catastrophic health expenditure. The key determinants of catastrophic health expenditure were economic status, household health care utilization especially for modern medical care, illness episodes in an adult household member and presence of a member with chronic illness. CONCLUSION: We conclude that the poorest members of the community incurred catastrophic health expenses. Setting only one threshold/cut-off value to determine catastrophic health expenses may result in inaccurate estimation leading to misinterpretation of important factors. Our findings have important policy implications and can be used to ensure better access to health services and a higher degree of financial protection for low-income groups against the economic impact of illness.     

Inequality decomposition and geographic targeting with applications to China and Vietnam

How far are income‐related inequalities in the health sector due to gaps between poor and less poor areas, rather than due to differences between poor and less poor people within areas? This note sets out a method for answering this question, and illustrates it with two empirical examples. The disproportionate accrual of health subsidies to Vietnam's better‐off is found to be largely due to the fact that richer provinces have larger per capita subsidies, while pro‐rich inequalities in health insurance coverage in rural China are found to be largely due to the fact that better‐off villages have been more successful at preventing the collapse of their insurance schemes. Copyright © 2005 John Wiley & Sons, Ltd.     

Sanitation,financial incentives and health spillovers: A cluster randomised trial

Poor sanitation and its consequent negative health outcomes continue to plague the developing world. Drawing on the finding that financial subsidies have changed behaviour in other health contexts, we conducted a clustered randomised trial in 160 villages in Lao PDR to evaluate the effectiveness of combining financial incentives with Community-Led Total Sanitation (CLTS), a widely-conducted behaviour change program. Villages were randomly allocated to four groups, all of which received CLTS but differed in the type of subsidy offered (none, household, village or both). Using data from a random sample of households with young children and village administrative data, we show that household incentives increased sanitation take-up among the poor, whereas a village incentive increased take-up primarily among the non-poor. Improved sanitation produced positive health spillovers - a 10 percentage point increase in village sanitation coverage decreased the probability of childhood stunting by 3 percentage points.     

Community-based health insurance in poor rural China: the distribution of net benefits

The collapse of China's Cooperative Medical System (CMS) in 1978 resulted in the lack of an organized financing scheme for health care, adversely affecting rural farmers' access to health care, especially among the poor. The Chinese government recently announced a policy to re-establish some forms of community-based insurance (CBI). Many existing schemes involve low premiums but high co-payments. We hypothesized that such benefit design leads to unequal distribution of the "net benefits" (NB)--benefits net of payment--because even though low premiums are more affordable to poor farmers, high co-payments may have a significant deterrent effect on the poor in the use of services in CBI. To test this hypothesis empirically, we estimated the probability of farmers joining a re-established CBI using logistic regression, and the utilization of health care services for those who joined the scheme using the two-part model. Based on the estimations, we predicted the distribution of NB among those who joined the CBI and for the entire population in the community. Our data came from a household survey of 4160 members of 1173 households conducted in six villages in Fengshan Township, Guizhou Province, China. Three principal findings emerged from this study. First, income is an important factor influencing farmers' decision to join a CBI despite the premium representing a very small fraction of household income. Secondly, both income and health status influence enrollees' utilization of health services: richer/sicker participants obtain greater NB from the CBI than poorer/healthier members, meaning that the poorer/healthier participants subsidize the rich/sick. Thirdly, wealthy farmers benefit the most from the CBI with low premium and high co-payment features at every level of health status. In conclusion, policy recommendations related to the improvement of the benefit distribution of CBI schemes are made based on the results from this study.     

Does health insurance improve health?: Evidence from a randomized community-based insurance rollout in rural Burkina Faso

From 2004 to 2006, a community-based health insurance (CBI) scheme was rolled out in Nouna District, Burkina Faso, with the objective of improving access to health services and population health. We explore the random timing of the insurance rollout generated by the stepped wedge cluster-randomized design to evaluate the welfare and health impact of the insurance program. Our results suggest that the insurance had limited effects on average out-of-pocket expenditures in the target areas, but substantially reduced the likelihood of catastrophic health expenditure. The introduction of the insurance scheme did not have any effect on health outcomes for children and young adults, but appears to have increased mortality among individuals aged 65 and older. The negative health effects of the program appear to be primarily driven by the adverse provider incentives generated by the scheme and the resulting decline in the quality of care received by patients.     

The impact of reducing financial barriers on utilisation of a primary health care facility in Rwanda

Abstract

This study investigates the impact of subsidising community-based health insurance (mutuelle) enrolment, removing point-of-service co-payments, and improving service delivery on health facility utilisation rates in Mayange, a sector of rural Rwanda of approximately 25,000 people divided among five ‘imidugudu’ or small villages. While comprehensive service upgrades were introduced in the Mayange Health Centre between April 2006 and February 2007, utilisation rates remained similar to comparison sites. Between February 2007 and April 2007, subsidies for mutuelle enrolment established virtually 100% coverage. Immediately after co-payments were eliminated in February 2007, patient visits levelled at a rate triple the previous value. Regression analyses using data from Mayange and two comparison sites indicate that removing financial barriers resulted in about 0.6 additional annual visits for curative care per capita.

Although based on a single local pilot, these findings suggest that in order to achieve improved health outcomes, key short-term objectives include improved service delivery and reduced financial barriers. Based on this pilot, higher utilisation rates may be affected if broader swaths of the population are enrolled in mutuelle and co-payments are eliminated. Health leaders in Rwanda should consider further studies to determine if the impact of eliminating co-payments and increasing subsidies for mutuelle enrolment as seen in Mayange holds at greater levels of scale. Broader studies to better elucidate the impact of enrolment subsidies and co-payment subsidies on utilisation, health outcomes, and costs would also provide policy insights.     

Shame or subsidy revisited: social mobilization for sanitation in Orissa, India

Objective

To determine the effectiveness of a sanitation campaign that combines “shaming” (i.e. emotional motivators) with subsidies for poor households in rural Orissa, an Indian state with a disproportionately high share of India’s child mortality.

Methods

Using a cluster-randomized design, we selected 20 treatment and 20 control villages in the coastal district of Bhadrak, rural Orissa, for a total sample of 1050 households. We collected sanitation and health data before and after a community-led sanitation project, and we used a difference-in-difference estimator to determine the extent to which the campaign influenced the number of households building and using a latrine.

Findings

Latrine ownership did not increase in control villages, but in treatment villages it rose from 6% to 32% in the overall sample, from 5% to 36% in households below the poverty line (eligible for a government subsidy) and from 7% to 26% in households above the poverty line (not eligible for a government subsidy).

Conclusion

Subsidies can overcome serious budget constraints but are not necessary to spur action, for shaming can be very effective by harnessing the power of social pressure and peer monitoring. Through a combination of shaming and subsidies, social marketing can improve sanitation worldwide.     

The Affordable Care Act's insurance market regulations' effect on coverage

Much of the debate surrounding reform of the Patient Protection and Affordable Care Act (ACA) revolves around its insurance market regulation. This paper studies the impact on health insurance coverage of those provisions. Using data from the American Community Survey, years 2008–2015, I focus on individuals, ages 26 to 64, who are ineligible for the subsidies or Medicaid expansions included in the ACA to isolate the effect of its market regulation. To account for time trends, I utilize a differences‐in‐differences approach with a control group of residents of Massachusetts who were already subject to a similarly regulated health insurance market. I find that the ACA's regulations caused an increase of 0.95 percentage points in health insurance coverage for my sample in 2014. This increase was concentrated among younger individuals, suggesting that the law's regulations ameliorated adverse selection in the individual health insurance market.     

Implementation of the redesigned Community Health Fund in the Dodoma region of Tanzania: A qualitative study of views from rural communities

The need to understand how an intervention is received by the beneficiary community is well recognised and particularly neglected in the micro–health insurance (MHI) domain. This study explored the views and reactions of the beneficiary community of the redesigned Community Health Fund (CHF) implemented in the Dodoma region of Tanzania. We collected data from focus group discussions with 24 groups of villagers (CHF members and nonmembers) and in‐depth interviews with 12 key informants (enrolment officers and health care workers). The transcribed material was analysed thematically. We found that participants highly appreciate the scheme, but to be resolved are the challenges posed by the implementation strategies adopted. The responses of the community were nested within a complex pathway relating to their interaction with the implementation strategies and their ongoing reflections regarding the benefits of the scheme. Community reactions ranged from accepting to rejecting the scheme, demanding the right to receive benefit packages once enrolled, and dropping out of the scheme when it failed to meet their expectations. Reported drivers of the responses included intensity of CHF communication activities, management of enrolment procedures, delivery of benefit packages, critical features of the scheme, and contextual factors (health system and socio‐political context). This study highlights that scheme design and implementation strategies that address people's needs, voices, and values can improve uptake of MHI interventions. The study adds to the knowledge base on implementing MHI initiatives and could promote interests in assessing the response to interventions within the MHI domain and beyond.     

Health insurance reform and retirement: Evidence from the Affordable Care Act

The Affordable Care Act (ACA) has provided millions of Americans with medical insurance but may have led to an increase in retirement among older individuals who are utilizing the newly available coverage options as a substitute for employer‐provided insurance. Using data from the American Community Survey from 2009–2016, this hypothesis is tested by estimating the effect of the premium subsidies and Medicaid expansions of the ACA on retirement transitions for the non‐Medicare eligible cohort of older Americans aged 55–64. Research results indicate a 2% and 8% decrease in labor force participation resulting from the premium subsidies and Medicaid expansions, respectively. Slightly larger estimates are found among a subgroup of adult couples. The study also finds suggestive evidence of crowd‐out of employer‐sponsored insurance by subsidized marketplace plans but finds no such effects from the Medicaid expansions.     

Assessment of factors associated with complete immunization coverage in children aged 12-23 months: a cross-sectional study in Nouna district,Burkina Faso

Background

The Expanded Program on Immunization (EPI) is still in need of improvement. In Burkina Faso in 2003, for example, the Nouna health district had an immunization coverage rate of 31.5%, compared to the national rate of 52%. This study identifies specific factors associated with immunization status in Nouna health district in order to advance improved intervention strategies in this district and in those with similar environmental and social contexts.

Methods

A cross-sectional study was undertaken in 41 rural communities and one semi-urban area (urban in the text). Data on 476 children aged 12 to 23 months were analyzed from a representative sample of 489, drawn from the Nouna Health Research Centre's Demographic Surveillance System (DSS) database. The vaccination history of these children was examined. The relationships between their immunization status and social, economic and various contextual variables associated with their parents and households were assessed using Chi square test, Pearson correlation and logistic regression.

Results

The total immunization coverage was 50.2% (CI, 45.71; 54.69). Parental knowledge of the preventive value of immunization was positively related to complete immunization status (p = 0.03) in rural areas. Children of parents who reported a perception of communication problems surrounding immunization had a lower immunization coverage rate (p < 0.001). No distance related difference exists in terms of complete immunization coverage within villages and between villages outside the site of the health centres. Children of non-educated fathers in rural areas have higher rates of complete immunization coverage than those in the urban area (p = 0.028). Good communication about immunization and the importance of availability of immunization booklets, as well as economic and religious factors appear to positively affect children's immunization status.

Conclusion

Vaccination sites in remote areas are intended to provide a greater opportunity for children to access vaccination services. These efforts, however, are often hampered by the poor economic conditions of households and insufficient communication and knowledge regarding immunization issues. While comprehensive communication may improve understanding about immunization, it is necessary that local interventions also take into account religious specificities and critical economic periods. Particular approaches that take into consideration these distinctions need to be applied in both rural and urban settings.

Abstract in French

See the full article online for a translation of this abstract in French.