The study built on similar previous studies focusing on health care financing conducted by the Population Council in collaboration with Government of Kenya and other partners. These include evaluation of the impact of reproductive health vouchers programs in five countries (Bangladesh, Cambodia, Kenya, Tanzania and Uganda); evaluation of performance-based finance (PBF) pilot program in Samburu County in Kenya; the Health Insurance Subsidy Programme (HISP) in Kenya and Health Access for Kenyans through Innovation (HAKI). The household survey in this study used the same tools and was conducted in the same counties where the evaluation of the reproductive health vouchers program was undertaken.
The current study used a mixed methods approach. The qualitative component involved targeted interviews to identify the process and complexities of implementing the user fee removal policy and to better understand the factors that influence women to seek facility-based deliveries with and without user fees.
The quantitative component involved a large household survey. 2,156 women aged 18-49 years from the reproductive health vouchers evaluation sites were included in the study. The sample size calculation is based on the 2008 national proportion of facility-based births among all births, which was the most recent data available at the time the study was conceptualized. An assumption was made that the national figure is representative of the proportion of facility-based births in the voucher project sites. To detect a 14% increase proportion of facility-based births, an estimated 1,078 women voucher sites and 1,078 from comparable non-voucher sites would be sufficient to reject the null hypothesis that the proportion of facility-based births in voucher and comparison sites are equal with probability (power) of 0.8 and Type I error probability of 0.05.
A total 25 policymakers, stakeholders and facility in-charges (managers) aged 18 years and above were invited for interview. The sample size for in-depth interviews was derived based on purposive sampling procedures, with respondents being selected according to the criteria described elsewhere in the study protocol. Similarly, 25 focus groups were held with voucher and non-voucher clients.