Social Health Insurance for HIV Prevention and Treatment: Evidence from Kenya

Background: Social health insurance enhances and diversifies health financing in light of dwindling tax revenues, increasing healthcare costs, and burdensome out-of pocket payments especially for the poor. Established in 1966, Kenya’s National Hospital Insurance Fund (NHIF) is the oldest SHI in Africa and covers mainly inpatient services irrespective of pre-existing medical conditions. NHIF’s focus is important for persons living with HIV/AIDS (PLWHA) given that on average 51% of hospital beds in Kenya are occupied by PLWHA and HIV/AIDS is by far the leading cause of death. This is particularly true for HIV+ women, who continue to desire children, become pregnant, and give birth after knowing their HIV+ status; thus, requiring access to appropriate care during delivery so as to aid in prevention of mother to child transmission (PMTCT).

Aim: To evaluate the impact of Social Health Insurance (SHI) on HIV prevention and Treatment in Kenya

Objective: To investigate the extent to which affiliation to NHIF affects obstetric outcomes for post-partum HIV+ women and their newborn children, focusing specifically on the impact of NHIF on hospital/institutional delivery. The overall hypothesis is that NHIF improves access and utilization of institutional delivery services and improves obstetric and newborn outcomes.

Methods: We are using the Academic Model Providing Access to Healthcare’s (AMPATH’s) population to implement this study given its focus on provision of Antiretroviral Treatment (ART), training, and research.  AMPATH has a comprehensive electronic database, the AMPATH Medical Records System (AMRS), which has an extensive set of covariates for approximately 180,000 HIV+ persons. We will control for observable characteristics by choosing a homogenous "treatment"  group comprising AMPATH’s HIV+ pregnant women and their newborn children enrolled in NHIF, with the "control" group being a non-insured group consisting also of HIV+ pregnant women and their newborn children but not enrolled in NHIF. To further reduce selection on observables, we implement propensity score matching and use targeted maximum likelihood (TMLE) with propensity score methods for causal estimation.

Key Findings: AMPATH provides HIV testing to approximately 3,000 expectant mothers every month and annually treats about 6,600 HIV+ women with antiretroviral prophylaxis. AMPATH has approximately 14,927 HIV+ positive individuals insured by NHIF with 9,333 women (63%). We will finalize the propensity score matching analysis and expect to have the results in time for the conference.