Soil-transmitted helminth infection in South America: a systematic review and geostatistical meta-analysis

Background The four common soil-transmitted helminth species Ascaris lumbricoides, Trichuris trichiura, and the two hookworm species Ancylostoma duodenale and Necator americanus are endemic in South America, but their distribution, infection prevalence, and regional burden are poorly understood. We aimed to estimate the risk and number of people infected with A lumbricoides, T trichiura, and hookworm across South America. Methods We did a systematic review of reports on the prevalence of soil-transmitted helminth infection in South America published up to May 14,2012. We extracted and georeferenced relevant survey data and did a meta-analysis of the data to assess the geographical distribution of the infection risk with Bayesian geostatistical models. We used advanced Bayesian variable selection to identify environmental determinants that govern the distribution of soil-transmitted helminth infections. Findings We screened 4085 scientific papers and identified 174 articles containing relevant survey prevalence data. We georeferenced 6948 survey locations and entered the data into the open-access Global Neglected Tropical Diseases database. Survey data were sparse for the south of the continent and for the western coast, and we identified no relevant information for Uruguay and little data for smaller countries such as Suriname, Guyana, French Guiana, and Ecuador. Population-adjusted prevalence of infection with A lumbricoides was 15.6%, with T trichiura was 12.5%, and with hookworm was 11.9% from 2005 onwards. Risks of contracting soil-transmitted helminth infection have substantially reduced since 2005 (odds ratio 0.47 [95% Bayesian credible interval 0.46-0.47] for A lumbricoides, 0.54 [0.54-0.55] for T trichiura, and 0-58 [0-58-0-59] for hookworm infection). Interpretation Our findings offer important baseline support for spatial targeting of soil-transmitted helminthiasis control, and suggest that more information about the prevalence of soil-transmitted helminth infection is needed, especially in countries in which we estimate prevalence of infection to be high but for which current data are scarce.