A systematic review of entomological outcomes and sampling approaches used in the evaluation of cluster randomised controlled trials for malaria vector control products

Abstract Background Entomological outcomes are critical for understanding the biological mechanisms and operational performance of malaria vector control interventions. However, the quality and consistency of how these outcomes are measured and reported in cluster randomised controlled trials (cRCTs) is unclear. We conducted a systematic review of design features, sampling methods, and metrics used for measuring entomological outcomes in malaria vector control cRCTs. Our aims were to assess the consistency and quality of entomological study designs and examine how key design features influence the precision of reported entomological effect sizes. Methods Malaria cRCTs in the published literature were identified in four databases and two trial registries. We extracted information on entomological outcomes measured, entomological sampling strategies and trial design features. To evaluate overall design quality, we created an Entomological Study Design Risk (ESDR) metric that assessed trials across four domains related to power analysis and sampling design; with domain ratings combined to generate an overall ESDR score. We also used a linear mixed-effects model (LMM) to examine how the use of power analysis and the number of clusters and sampling points in entomological sampling influenced the precision of main entomological outcomes. Results Sixty-two cRCTs of malaria vector control interventions were identified between 1992 and 2021; with 74% being conducted in Africa. Thirteen different entomological outcomes and 12 sampling methods were used across trials: with considerable variation in the frequency of entomological sampling within trials. In total, 70% of the cRCTs were categorized as having high potential entomological design risk based on a combination of the lack of power analyses and inadequate or unclear randomisation of sampling points. The only significant predictor of the precision of entomological outcomes was the number of entomological clusters in which data were collected (greater precision as cluster number increased). Conclusions There is substantial variation in how entomological outcomes are measured and reported in cRCTs, and limited incorporation of key design features such as power analysis and random selection of sampling points for outcome measurement. Standardization where possible and clearer reporting guidelines for entomological components are therefore needed to improve the quality, comparability, and policy relevance for future trials.