RESUMO
Objective: To explore the impact of mosquito collection methods, sampling intensity and target genus on molecular xenomonitoring detection of parasites causing lymphatic filariasis. Methods: We systematically searched five databases for studies that used two or more collection strategies for sampling wild mosquitoes, and employed molecular methods to assess the molecular xenomonitoring prevalence of parasites responsible for lymphatic filariasis. We performed generic inverse variance meta-analyses and explored sources of heterogeneity using subgroup analyses. We assessed methodological quality and certainty of evidence. Findings: We identified 25 eligible studies, with 172 083 mosquitoes analysed. We observed significantly higher molecular xenomonitoring prevalence with collection methods that target bloodfed mosquitoes compared to methods that target unfed mosquitoes (prevalence ratio: 3.53; 95% confidence interval, CI: 1.52-8.24), but no significant difference compared with gravid collection methods (prevalence ratio: 1.54; 95% CI: 0.46-5.16). Regarding genus, we observed significantly higher molecular xenomonitoring prevalence for anopheline mosquitoes compared to culicine mosquitoes in areas where Anopheles species are the primary vector (prevalence ratio: 6.91; 95% CI: 1.73-27.52). One study provided evidence that reducing the number of sampling sites did not significantly affect molecular xenomonitoring prevalence. Evidence of differences in molecular xenomonitoring prevalence between sampling strategies was considered to be of low certainty, due partly to inherent limitations of observational studies that were not explicitly designed for these comparisons. Conclusion: The choice of sampling strategy can significantly affect molecular xenomonitoring results. Further research is needed to inform the optimum strategy in light of logistical constraints and epidemiological contexts.
