Encouraging maternal health service utilization: An evaluation of the Bangladesh voucher program

https://doi.org/10.1016/j.socscimed.2011.11.030Get rights and content

Abstract

With the ultimate goal of reducing maternal and neonatal mortality, many countries have recently adopted innovative financing mechanisms to encourage the use of professional maternal health services. The current study evaluates one such initiative – a pilot voucher program in Bangladesh. The program provides poor women with cash incentives and free access to antenatal, delivery, and postnatal care, as well as cash incentives for providers to offer these services.

We conducted a household survey of 2208 women who delivered in the 6 months before the survey (conducted in 2009) in 16 intervention and 16 matched comparison sub-districts. Probit and linear regressions are used to analyze the effects of residing in voucher sub-districts on the use of professional maternal health services and associated out-of-pocket expenditures. Using information on birth history, we conducted sensitivity analyses employing difference-in-differences methods, comparing women’s reported births before and after the program’s initiation in the intervention and comparison sub-districts.

We found that the program significantly increased the use of antenatal, delivery, and postnatal care with qualified providers. Compared to women in matched comparison sub-districts, women in intervention areas had a 46.4 percentage point higher probability of using a qualified provider and 13.6 percentage point higher probability of institutional delivery. They also paid approximately Taka 640 (US$ 9.43) less for maternal health services, equivalent to 64% of the sample’s average monthly household expenditure per capita. No significant effect of vouchers was found on the rate of Cesarean section.

Our findings therefore support voucher program expansion targeting the economically disadvantaged to improve the use of priority health services. The Bangladesh voucher program is a useful example for other developing countries interested in improving maternal health service utilization.

Highlights

► The Bangladesh pilot voucher program has improved the utilization of formal maternal health services and reduced the financial burden of care. ► Vouchers are effective in targeting the economically disadvantaged to improve the use of priority health services. ► Combining demand- and supply-side incentives are important in health intervention programs. ► Further evaluations are warranted to assess the cost-effectiveness of vouchers vs other types of interventions to improve maternal health.

Introduction

Despite significant improvements in the past two decades, most developing countries will not achieve the Millennium Development Goal 5 to reduce maternal mortality by 75% between 1990 and 2015 (Hogan et al., 2010, United Nations, 2009). A priority toward this end is to improve access to and use of quality health services and skilled assistance at birth (Donnay, 2000, Singh et al., 2009). However, only half of parturient women receive skilled assistance at delivery and many fewer receive postpartum care (Singh et al., 2009). Women in developing countries face multiple barriers to using formal health services, including limited physical and financial access, lack of voice and decision-making authority, and lack of education (Matsuoka et al., 2010, Priya, 2002, Sharma et al., 2005, Simkhada et al., 2006). The poor quality of available health services presents an additional deterrent (Singh et al., 2009). The complexity of these barriers suggests that any solution to increase maternal health service utilization must be comprehensive, taking into account both health system constraints on the supply side as well as financial, cultural, and knowledge barriers on the demand side.

In recent years, there has been a mounting interest in the use of vouchers and other innovative financing mechanisms to improve the effects of maternal and child health programs (Bellows et al., 2010, Bhatia and Gorter, 2007, Donaldson et al., 2008, Ir et al., 2010, Janisch et al., 2010). By providing a financial or in-kind reward conditional upon achievement of agreed performance goals, vouchers can be a promising holistic approach to encourage the use of cost-effective services by the poor and other disadvantaged populations (Gorter, 2003). Vouchers can work through supply or demand side or both. Supply-side incentives aims at improving quality and responsiveness of the service provision, while demand side incentives encourage utilization of services by not only lessening the financial burden, but also giving women the choice of providers and educating them of the benefits of using maternal health service. Vouchers are a particularly appealing approach for addressing barriers to maternal health services. The target population (pregnant women) is a vulnerable priority group that can be easily identified; a package of necessary maternal health services can be clearly defined; and the financial barriers to service use – particularly in the case of pregnancy complications – are substantial. To date, evidence suggests that many voucher programs have successfully encouraged deliveries with skilled birth attendants and institutional deliveries (Bellows et al., 2010, Ir et al., 2010). However, little evidence is based on rigorous evaluation studies, which makes it difficult to draw confident conclusions on the impact of voucher initiatives and make corresponding policy recommendations.

The current study presents an evaluation of a pilot voucher program in Bangladesh, a country where roughly 85% of all deliveries took place at home and only 18% of births were assisted by a qualified provider at the time of the program’s inception (NIPORT et al., 2009). Using current and past delivery information from women residing in voucher and non-voucher areas, we perform single and double differences estimates of the program’s effects on utilization of key maternal health services and associated out-of-pocket (OOP) expenditure.

In the following, we provide a brief description of the Bangladesh voucher program. We next present our data and methods, to be followed by the estimation results. We conclude with a discussion of the study’s strengths and weaknesses, as well as the implications of its findings.

Section snippets

Bangladesh voucher program for maternal health

Known in the country as “Demand-Side Financing program”, the Bangladesh pilot voucher program aims to encourage the use of maternal health services, in particular qualified birth attendance, and mitigate the financial costs of delivery. It started out in 2004 in 2 sub-districts, expanded significantly to 21 in 2007, and is now functioning in 46 out of 489 sub-districts in the country. Covering a population of roughly 10 million people, the program is financed with national and pooled donor

Data

Data for this evaluation come from a household survey conducted in 2009 in 32 sub-districts, of which 16 have been implementing the pilot since mid-2007 and 16 have not had the program. The 16 intervention sub-districts (8 universal and 8 means-testing) were selected to the survey to represent all divisions of the country. Control sub-districts were matched with intervention sub-districts on geographical proximity (i.e. from the same district), number of beds in the sub-district hospital, and

Results

Table 3 provides a descriptive statistics of the study sample. Women in the intervention group were younger (23.7 vs. 24.6 years old, p < 0.001) and their babies’ birth order was lower (2.0 vs. 2.2, p = 0.002) than those in the comparison group. More intervention group women were in the two poorest quintiles than in the comparison group (42.8% vs. 37.3%, p = 0.053 for the overall test of quintile variable). More intervention than comparison group sub-district health complexes provided EOC (81.4% vs.

Discussion

We find that Bangladesh’s pilot voucher program has had large and positive effects on the utilization of maternal health services in the short time since its initiation. As a result of the program, pregnant women were significantly more likely to seek care from qualified providers for ANC, delivery, and PNC. The estimated increase in the probability of delivering with a qualified provider ranges from 35 to 46 percentage points. In relative terms, this represents a more than one hundred percent

Acknowledgment

The evaluation study was funded by the Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH and implemented by Abt Associates Inc. (USA) and RTM International (Bangladesh). The manuscript preparation was supported by Abt Associates’ Daniel McGillis Development and Dissemination Grant Program. The authors wish to thank Abt Associates’ members of the Journal Authors Support Group, in particular Jacob Klerman and Slavea Chankova, for their comments on the earlier draft of the paper. We

References (25)

  • D. Filmer et al.

    Estimating wealth effect without expenditure data – or tears: an application to educational enrollments in states of India

    Demography

    (2001)
  • A.C. Gorter

    A background paper on competitive voucher schemes for health, Instituto CentroAmerica de la Salud

    (August 2003)
  • Cited by (75)

    View all citing articles on Scopus
    View full text