A cross-sectional study of malaria endemicity and health system readiness to deliver services in Kenya, Namibia and Senegal
Despite good progress towards elimination, malaria continues to contribute substantially to the sub-Saharan African disease burden. Sustaining previous gains requires continued readiness to deliver malaria services in response to actual disease burden, which in turn contributes to health systems strengthening. This study investigates a health system innovation. We examined whether malaria prevalence, or endemicity, is a driver of health facility readiness to deliver malaria services. To estimate this association, we geo-linked cross-sectional facility survey data to endemicity data for Kenya, Namibia and Senegal. We tested the validity and reliability of the primary study outcome, the malaria service readiness index and mapped service readiness components in a geographic information system. We conducted a weighted multivariable linear regression analysis of the relationship between endemicity and malaria service readiness, stratified for urban or rural facility location. As endemicity increased in rural areas, there was a concurrent, modest increase in service readiness at the facility level [β: 0.028; (95% CI 0.008, 0.047)], whereas no relationship existed in urban settings. Private-for-profit facilities were generally less prepared than public [β: −0.102; (95% CI − 0.154, −0.050)]. Most facilities had the necessary supplies to diagnose malaria, yet availability of malaria guidelines and adequately trained staff as well as medicines and commodities varied. Findings require cautious interpretation outside the study sample, which was a more limited subset of the original surveys’ sampling schemes. Our approach and findings may be used by national malaria programs to identify low performing facilities in malarious areas for targeted service delivery interventions. This study demonstrates use of existing data sources to evaluate health system performance and to identify within- and cross-country variations for targeted interventions.