921Bull World Health Organ 2009;87:921–929 | doi:10.2471/BLT.08.058933 Use of Bayesian geostatistical prediction to estimate local variations in Schistosoma haematobium infection in western Africa

Objective To predict the subnational spatial variation in the number of people infected with Schistosoma haematobium in Burkina Faso, Mali and the Niger prior to national control programmes. Methods We used field survey data sets covering a contiguous area 2750 × 850 km and including 26 790 school-age children (5–14 years old) in 418 schools. The prevalence of high- and low-intensity infection and associated 95 % credible intervals (CrIs) were predicted using Bayesian geostatistical models. The number infected was determined from the predicted prevalence and the number of school-age children in each km. Findings The predicted number of school-age children with a low-intensity infection was 433 268 in Burkina Faso, 872 328 i