Can non-routine data collection help decision-making for gambiense sleeping sickness? Using active adaptive management to assess the potential of vector control

Gambiense human African trypanosomiasis (gHAT) is a vector-borne disease, spread by tsetse, found in West and Central Africa. It has been targeted for elimination of transmission (EoT) by the World Health Organization: the reduction to zero of the incidence of infection, with minimal risk of reintroduction. Many high-risk regions have implemented vector control (VC), in addition to medical interventions, to reduce the spread of gHAT and reach EoT. There exists uncertainty around the effectiveness of VC interventions, and so the impact of VC can be measured through deploying traps, which estimate changes in the tsetse population. In this study, we use mathematical transmission modelling to assess the added monetary value of entomological monitoring data in strategy selection for two health zones of the Democratic Republic of Congo, illustratively chosen because of their different levels of uncertainty on VC effectiveness. We used active adaptive management (AAM) - the process of updating our strategy based on collected data (i.e. choosing to continue or stop VC based on perceived effectiveness), where posterior distributions on VC effectiveness were calculated with an adaptive MCMC method. Next, we used an established gHAT transmission model along with the policy objective of maximising the intervention's net monetary benefit to calculate the value gained by various monitoring strategies. We found that the strategy that included VC and twice-yearly entomological monitoring was cost-effective at a willingness to pay threshold of $153 for the health zone of Mulumba, where we were able to determine the optimal monitoring strategy, but it was not cost-effective in the health zone of Vanga. This study is the first application of AAM to monitoring data for a vector-bone infection, presenting a novel framework for quantitative assessment of how VC monitoring could be used to improve the cost-effectiveness of interventions in other regions.