Nationally representative programmatic surveillance of mosquito and human behaviours that influence human exposure to malaria transmission and the impact of vector control across Tanzania.

backgroundIntegrated entomological and anthropological surveillance of interacting mosquito and human behaviours is critical for understanding malaria transmission risks and for tailoring intervention packages, but it is yet to be implemented at scale due to a lack of practical, affordable procedures for nationally representative monitoring. This study in Tanzania introduces and field-tests the first such scalable framework design and reports summaries of the earliest data generated, thus demonstrating its feasibility and utility for generating programmatically informative indicators of both mosquito and human behaviours.

methodsA single village was randomly selected from each of twenty-five ecologically and epidemiologically diverse sentinel districts distributed across Tanzania for simultaneous surveys of malaria vector and human behaviours, using a rolling cross-sectional design. Entomological and anthropological surveys were repeated approximately annually over three years, with each village surveyed on three separate occasions across wet and dry seasons to account for temporal variations between sites due to seasons. Mosquito electrocuting traps (MET) were employed for sampling mosquitoes both indoors and outdoors with human behaviour surveyed by standard questionnaire.

resultsEntomological and anthropological indicators were successfully and simultaneously measured in all 25 sentinel districts. Anopheles arabiensis was widespread across the country, whilst Anopheles gambiae s.s. and Anopheles funestus s.s. were more localized. Despite being collected in small numbers, blood-fed An. funestus s.s. and An. gambiae s.s. were predominantly human-fed, whereas An. arabiensis, An. quadriannulatus and An. leesoni fed predominantly on cattle. Higher proportions of host-seeking An. arabiensis were caught outdoors (64.9%) than An. gambiae s.s. (58.4%) or An. funestus s.s. (42.4%). Across all age groups, people were more likely to be outdoors during the early evening hours (18:00 to 21:00), with the highest proportions observed among adolescents (13-17 years) and adults (≥ 18 years) compared to under-fives (0-5 years) and children (6-12 years). Adolescent males were less likely to use bednet relative to other demographic groups. Higher temperatures were associated with reduced reported bednet use during the dry season but not during the wet season.

conclusionThis national surveillance framework proved highly effective for measuring relevant metrics of mosquito and human behaviours that can inform optimization of malaria control strategies.