System-level factors associated with performance of prevention of mother-to-child transmission (PMTCT) services in Côte d’Ivoire

Abstract

Background Although the coverage of antiretrovirals (ARVs) for prevention of mother to child transmission (PMTCT) has increased globally after efforts, the uptake of PMTCT services remains low. We aim to identify system-level barriers to uptake HIV test and maternal prophylaxis during pregnancy in the existing national program in Côte d’Ivoire. Method Data of 46 sites during November 2011 to October 2015 were analyzed. Multiple Poisson regression models were used to 1) investigate systems-level factors associated with HIV testing and the delivery of maternal ARVs under Option B (Nov. 2012- Oct. 2015); 2) and highlight the differences in these associations between the Option A (Nov.2011-Oct.2012) and the Option B (Nov. 2012- Oct. 2015) periods. A significance level of 0.05 was used to determine factors associated with each outcome. Result In those 46 sites, 39 (85%) sites are public and 7 (15%) sites are privately-operated. During the Option B period, after adjusting for confounders, regional hospitals had higher rates of HIV testing compared to other health facilities (RR: 1.24; p=0.001) and lower rates of maternal ARV delivery (RR: 0.73; p=0.02). Conversely, privately-operated health facilities reported lower rates of HIV testing than public facilities (RR: 0.92; p<0.001), but higher rates of ARV delivery (RR: 1.23; p=0.008). Modifiable systems-level factors associated with rate of HIV testing included: the lack of on-site laboratory facilities (RR: 0.90; p=0.002); stock-outs of ARVs (RR: 0.88; p=0.046) and testing supplies (RR: 0.87; p=0.003). Stock-outs of general supplies were associated with low rates of delivery of maternal ARVs (RR: 0.79; p=0.046) as were stock-outs of ARVs, but with an inverse relation (RR: 1.41; p=0.005). While available charts, providing on-site CD4 tests, delay of ARV delivery were not significantly associated with either HIV test rate or delivery rate of appropriate maternal prophylaxis during the Option B period. Rural health facilities were associated with lower HIV testing rate (RR: 0.83; p=0.04), and big health facilities (>200 first antenatal care per month) were associated with higher delivery ratio of appropriate maternal prophylaxis (RR:1.39, p=0.01) as were available charts (RR:1.04, p=0.01) during the Option A period, but those associations were not significant during Option B period. Discussion HIV testing rates were higher in public health facilities and health facilities with a laboratory, while the privately-operated health facilities performed better in the delivery of appropriate maternal prophylaxis. Thus, when access to HIV tests is improving in low prevalence areas, especially in public facilities, specific attention should be paid to risks of loss to follow up from HIV testing to delivery of ARVs. Regional hospitals which served the largest population size had higher HIV testing rate but lower prophylaxis ratio. This implies bigger challenges to follow up pregnant women after HIV tests in facilities which serves a large population. Stock-outs of HIV test kits, general supply of ARVs limited the performance of HIV tests and maternal prophylaxis delivery. And supply of ARVs became more challenging when Option B guidelines were applied, which is likely to be an increasingly important issue as Option B+ guidelines are implemented. This study also highlights the need for further research regarding the association between system-level factors and adherence to ARVs, which might be one of the biggest challenges to improve PMTCT in Côte d’Ivoire under Option B+.