The reliability of self-reported blood in urine and schistosomiasis as indicators of Schistosoma haematobium infection in school children: a study in Muheza District, Tanzania.

The use of self-reported blood in urine and schistosomiasis by school children was investigated as a tool to estimate the prevalence of infection with Schistosoma haematobium and to identify infected individuals. A general questionnaire about common health problems, including questions about blood in urine and schistosomiasis, was administered by teachers to 25443 children in 137 primary schools in Muheza District, Tanzania. The prevalence of reported schistosomiasis was calculated for each school and used to select 15 schools across a range in prevalence. All children in the 15 schools (n = 2370) were interviewed again by a nurse and gave a urine sample which was subjected to a quantitative microscopical examination for the eggs of S. haematobium by filtration. The prevalence of reported schistosomiasis by the interview in the 15 schools correlated strongly with the prevalence reported during the questionnaire survey. The prevalence of reported schistosomiasis in the interview was strongly correlated with the prevalence of infection determined by microscopy and consistently under-estimated the latter by around 20% across a range in prevalence from 22% to 93%. The sensitivity of diagnosis by an interview increased almost linearly with the prevalence of infection, so that when the prevalence was high, more infected children reported schistosomiasis. The percentage of children who were correct in their self-diagnosis was independent of the prevalence of infection and of the mean concentration of eggs in urine, and averaged 75%. These findings suggest that self-reported schistosomiasis is a useful method to estimate the prevalence of infection in schools and might be used to identify infected individuals.