Modeling Reactive Case Detection for Malaria Elimination: A Metapopulation Approach

Regional elimination of malaria is made difficult, in part, by high volumes of asymptomatic or otherwise non-treatment-seeking cases. Reactive case detection (RCD) - a geographically targeted disease control strategy - identifies non-treatment-seeking cases by screening individuals who live in close proximity to known cases. These known cases, or ‘index’ cases, are selected from a pool of individuals who seek treatment from a health facility. Research has demonstrated that RCD is most appropriate in low-transmission settings where cases are highly clustered. The existing body of literature leaves room to better characterize optimal RCD policy parameters under resource constraints. Particularly of need is analysis with explicit treatment of spatially heterogeneous transmission, which may be important in the presence of clustered transmission pockets. In this study, we introduce a spatially-explicit modeling framework, perform stability analysis, and present and a novel RCD algorithm. Through simulation, we study the effects of key parameters on RCD performance and provide insight to guide optimal policy making. Our results suggest that RCD is always more effective than random test-and-treat of equal screening intensity in low-to-moderate transmission environments. However, the resource-constrained trade-off between screening radius and number of index cases is relatively unimportant except in very low transmission environments. In such settings, it is optimal to follow as many index cases with as small a search radius as possible such that full screening capacity is utilized. While RCD may be a useful tool, malaria reduction and elimination is often best achieved by reducing transmission rates, improving access to health coverage, and strengthening health systems overall