Evaluation of a push-pull system consisting of transfluthrin-treated eave ribbons and odour-baited traps for control of indoor- and outdoor-biting malaria vectors

Background: Push-pull strategies have been proposed as options to complement primary malaria prevention tools, indoor residual spraying (IRS) and long-lasting insecticide-treated nets (LLINs), by targeting particularly early-night biting and outdoor-biting mosquitoes. This study evaluated different configurations of a push-pull system consisting of spatial repellents [transfluthrin-treated eave ribbons (0.25 g/m 2 ai)] and odour-baited traps (CO 2 -baited BG-Malaria traps), against indoor-biting and outdoor-biting malaria vectors inside large semi-field systems. Methods: Two experimental huts were used to evaluate protective efficacy of the spatial repellents (push-only), traps (pull-only) or their combinations (push-pull), relative to controls. Adult volunteers sat outdoors (1830 h-2200 h) catching mosquitoes attempting to bite them (outdoor-biting risk), and then went indoors (2200 h-0630 h) to sleep under bed nets beside which CDC-light traps caught host-seeking mosquitoes (indoor-biting risk). Number of traps and their distance from huts were varied to optimize protection, and 500 laboratory-reared Anopheles arabiensis released nightly inside the semi-field chambers over 122 experimentation nights. Results: Push-pull offered higher protection than traps alone against indoor-biting (83.4% vs. 35.0%) and outdoor-biting (79% vs. 31%), but its advantage over repellents alone was non-existent against indoor-biting (83.4% vs. 81%) and modest for outdoor-biting (79% vs. 63%). Using two traps (1 per hut) offered higher protection than either one trap (0.5 per hut) or four traps (2 per hut). Compared to original distance (5 m from huts), efficacy of push-pull against indoor-biting peaked when traps were 15 m away, while efficacy against outdoor-biting peaked when traps were 30 m away. Conclusion: The best configuration of push-pull comprised transfluthrin-treated eave ribbons plus two traps, each at least 15 m from huts. Efficacy of push-pull was mainly due to the spatial repellent component. Adding odour-baited traps slightly improved personal protection indoors, but excessive trap densities increased exposure near users outdoors. Given the marginal efficacy gains over spatial repellents alone and complexity of push-pull, it may be prudent to promote just spatial repellents alongside existing interventions, e.g. LLINs or non-pyrethroid IRS. However, since both transfluthrin and traps also kill mosquitoes, and because transfluthrin can inhibit blood-feeding, field studies should be done to assess potential community-level benefits that push-pull or its components may offer to users and non-users.