Reducing Mosquito-Borne Disease Transmission To Humans: A Systematic Review Of Cluster Randomised Controlled Studies That Assess Interventions Other Than Non-Targeted Insecticide

Background Mosquito control interventions are widely used to reduce mosquito-borne diseases. It is unclear what combination of interventions are most effective in reducing human disease. A novel intervention study for Buruli ulcer targeting mosquito vectors was proposed for a Buruli ulcer-endemic area of Victoria, Australia. The local community expressed a preference for avoiding widespread residual spraying of pyrethroids. To inform the design of a future cluster randomised control study (cRCT) for Buruli ulcer prevention in Victoria, we conducted a systematic literature review. Aims The aim was to describe cRCT designs which investigated interventions other than non-targeted insecticide for reducing mosquito-borne disease transmission, and comment on the strengths and weaknesses of these study designs. Methods Five medical research databases were searched for eligible literature from the earliest available sources up to 5 July 2019 (Medline, Embase, Web of Science, EBM Reviews, CAB Direct). Reference lists of identified studies were hand searched. Eligible studies were cRCTs using targeted chemical or biological mosquito control interventions, or mosquito breeding source reduction, with the occurrence of mosquito-borne disease as an outcome. Results Eight eligible cRCTs, conducted between 1994-2013 were identified in a variety of settings in the Americas and Asia. Interventions to reduce dengue transmission were mass adult trapping and source reduction. Interventions to reduce malaria transmission were largescale larvicide administration and (topical and spatial) repellent use. Three studies showed the intervention was associated with statistically significant reductions in the disease of interest and entomological indicators. High community engagement with the intervention were common to all three. In two studies, large buffer zones reduced contamination between study arms. Heterogeneity was reduced through increasing study cluster numbers, cluster matching and randomisation. Conclusion High community engagement is vital for a cRCT reducing mosquito-borne disease with a mosquito control intervention. These findings support a mosquito breeding source reduction intervention for Aedes control in a future study of Buruli ulcer prevention if local communities are supportive and very engaged. Regular administration of larvicide to sites unsuited to source reduction may supplement the intervention.Author summary Mosquito control interventions are widely used to reduce mosquito-borne diseases, but it is unclear what combination of interventions are most effective in reducing human disease. Given the wide range of mosquito species and the diseases they transmit, different interventions strategies have been implemented across many regions globally, with varying degrees of success. This literature review identified three intervention studies which did not include non-targeted use of insecticide and were associated with statistically significant reductions in the disease of interest and in entomological indicators following the intervention. High community engagement is vital for the success of a cluster randomised control study aiming to reduce mosquito-borne disease with a mosquito control intervention, such as breeding source reduction for Aedes control. In two studies, large buffer zones reduced contamination between study intervention and control arms. Differences between the study arms were reduced through increasing study cluster numbers, cluster matching and randomisation. Regular administration of larvicide to potential breeding sites that are unsuitable for source reduction may supplement this intervention strategy.