Spatiotemporal distribution and population at risk of soil-transmitted helminth infections following an eight-year school-based deworming programme in Burundi, 2007-2014

Background: Investigating the effect of successive annual deworming rounds on the spatiotemporal distribution of infection prevalence and numbers at risk for soil-transmitted helminths (STHs) can help identify communities nearing elimination and those needing further interventions. In this study, we aim to quantify the impact of an 8-year mass drug administration (MDA) programme (from 2007 to 2014) on the spatiotemporal distribution of prevalence of STH infections and to estimate the number of school-aged children infected with STHs in Burundi. Methods: During annual longitudinal school-based surveys in Burundi between 2007 and 2011, STH infection and anthropometric data for a total of 40,656 children were collected; these data were supplemented with data from a national survey conducted in 2014. Bayesian model based geostatistics (MBG) were used to generate predictive prevalence maps for each STH species and year. The numbers of children at-risk of infection per district between 2008 and 2014 were estimated as the product of the predictive prevalence maps and population density maps. Results: Overall, the degree of spatial clustering of STH infections decreased between 2008 and 2011; in 2014 the geographical clusters of all STH infections reappeared. The reduction in prevalence was small for Ascaris lumbricoides and Trichuris trichiura in the centre and central north of the country. Our predictive prevalence maps for hookworm indicate a reduction in prevalence along the periphery of the country. The predicted number of children infected with any STH species decreased substantially between 2007 and 2011, but in 2014 there was an increase in the predicted number of children infected with A. lumbricoides and T. trichiura. In 2014, the districts with the highest predicted number of children infected with A. lumbricoides, T. trichiura and hookworms were Kibuye district (n = 128,903), Mabayi district (n = 35,302) and Kiremba (n = 87,511), respectively. Conclusions: While the MDA programme in Burundi resulted in a reduction in STH prevalence, this reduction was spatiotemporally heterogeneous, with some pockets of high prevalence remaining, suggesting that treatment coverage and complementary interventions should be evaluated to improve impact.