Current Gaps in Survey Design and Analysis for Molecular Xenomonitoring of Vector-Borne Neglected Tropical Diseases: A Systematic Review.

Abstract

Objectives: Molecular xenomonitoring is a surveillance method for vector-borne diseases where vectors are tested for molecular pathogen markers. Testing is typically on pools (groups) of vectors. Molecular xenomonitoring is a sensitive and efficient complement to human-based surveillance. However, existing statistical guidance for the appropriate design and analysis of molecular xenomonitoring surveys has key gaps. We reviewed the literature to understand the common objectives, survey designs, and analysis methods for molecular xenomonitoring surveys for two vector-borne neglected tropical diseases: lymphatic filariasis and onchocerciasis.

Methods: We searched peer-reviewed literature for studies published between 1999 and 2022 that presented the results of surveys that collected vectors in field surveys and used a molecular test for the presence of the causative pathogens for lymphatic filariasis and onchocerciasis.

Results: Out of 1225 works identified in the database search, a total of 76 studies (lymphatic filariasis: 45; onchocerciasis: 31) across 30 countries were included in the review. The five most common objectives were determination of elimination status after mass drug administration, comparison of vector and human infection indicators, evaluation of an intervention, comparison of vector collection methods and comparison of laboratory techniques. Nearly all studies used a cluster or hierarchical sampling framework to collect vectors (72/76), but very few studies accounted for this in their designs (2/76) or analysis (1/76). While few studies justified the number of vectors included in each pool (5/76), nearly all studies accounted for pooled testing when calculating pathogen prevalence from results (69/76). Few studies justified the number or selection of collection sites or total sample size (16/76).

Conclusions: Published molecular xenomonitoring surveys for lymphatic filariasis and onchocerciasis had varied objectives, study designs and analysis methods, but proper consideration of survey design was frequently missing from the analysis. There is a need for statistical tools and guidance to enable appropriate design and analysis of molecular xenomonitoring surveys while accounting for disease, objective and context-specific considerations.