Abstract
BACKGROUND: Reagent strip testing for microhaematuria has long been used for community diagnosis of Schistosoma haematobium. Sensitivities and specificities are reasonable, and hence, microhaematuria can serve as a proxy for S. haematobium infection. However, assessment of test performance in the context of the underlying S. haematobium prevalence is rare and test parameters other than sensitivity and specificity have been neglected. METHODS: Data about the association between microhaematuria and urine filtration results from three studies were compared and put into context with findings from a recent Cochrane review. Data were stratified by S. haematobium prevalence to identify prevalence-related differences in test performance. Kappa agreement and regression models were employed to compare data for different S. haematobium prevalence categories. RESULTS: We found a "background" prevalence of microhaematuria (13 %, on average) which does not seem to be associated with schistosomiasis in most settings, irrespective of the prevalence of S. haematobium. This background level of microhaematuria might be due to cases missed with urine filtration, or alternative causes apart from S. haematobium. Especially in very-low prevalence settings, positive results for microhaematuria likely give an inaccurate picture of the extent of S. haematobium, whereas negative results are a sound indicator for the absence of infection. CONCLUSIONS: Reagent strip testing for microhaematuria remains a good proxy for urogenital schistosomiasis, but implications of test results and scope of application differ depending on the setting in which reagent strips are employed. In very-low prevalence settings, microhaematuria is an unstable proxy for urogenital schistosomiasis and treatment decision should not be based on reagent strip test results alone. Our findings underscore the need for highly accurate diagnostic tools for settings targeted for elimination of urogenital schistosomiasis.