Trends in malaria indicators after scale-up of community-based malaria management in Afghanistan

Abstract

Background: The Community-Based Malaria Management (CBMM) strategy, introduced in 2013 and expanded to all health facilities and health posts in Afghanistan by 2016, aimed to deliver rapid diagnostic testing and more timely treatment to all communities nationwide. In this study, trends for several malaria outcome indicators were compared before and after the expansion of the CBMM strategy, using cross-sectional analysis of surveillance data. Methods: Generalized estimating equation (GEE) models with a Poisson distribution were used to assess trends of three key outcomes before (2012–2015) and after (2016–2019) CBMM expansion. These outcomes were annual malaria incidence rate (both all and confirmed malaria incidence), malaria death rate, and malaria test positivity rate. Additional variables assessed included annual blood examination rates (ABER) and malaria confirmation rate. Results: Average malaria incidence rates decreased from 13.1 before CBMM expansion to 10.0 per 1000 persons per year after CBMM expansion (P < 0.001). The time period after CBMM was expanded witnessed a 339% increase in confirmed malaria incidence as compared to the period before (IRR 3.39, 95% CI 2.18, 5.27; P < 0.001). In the period since the expansion of CBMM (2016–2019), overall malaria incidence rate declined by 19% each year (IRR 0.81, 95% CI 0.71,0.92; P = 0.001) and the malaria death rate declined by 85% each year (IRR 0.15, 95% CI 0.12, 0.20; P < 0.001). In comparing the before period to the after period, the ABER increased from 2.3 to 3.5 per 100 person/year, the malaria test positivity rate increased from 12.2 to 20.5%, and the confirmation rate increased from 21% before to 71% after CBMM. Conclusions: Afghanistan’s CBMM expansion to introduce rapid diagnostic tests and provide more timely treatment for malaria through all levels of care temporally correlates with significant improvement in multiple indicators of malaria control.