|Past two years||Past Year||Past 30 Days|
|Full Text Views||14||10||0|
An investigation of a population-based stratified random sample of 11,970 inhabitants of Minya Governorate in Egypt included examining urine specimens for Schistosoma haematobium ova using the filtration technique and for hematuria and proteinuria using reagent strips. Age- and gender-specific sensitivity, specificity, predictive values, and likelihood ratios of reagent strip-detected hematuria and proteinuria as indicators of S. haematobium infection were assessed. Results showed that in this population with a 9.0% infection rate, sensitivity (from 45.9 to 70.9) and specificity (from 65.9 to 86.3) were much lower than in previous reports. The most clinically valuable parameters, positive predictive value (PPV), negative predictive value (NPV) and likelihood ratio (LR), were highest with the presence of hematuria ≥ +++ and proteinuria ≥ ++, especially in males less than 20 years of age. Using the LR allowed effective altering of the pretest probability of infection among age and gender subgroups. The study concluded that reagent strip-detected hematuria and proteinuria, although valuable, are less reliable predictors of S. haematobium within communities than previously reported and their usefulness is improved with stratification by age and gender. In addition, predictive values and likelihood ratios provide practical information for predicting the presence or absence of infection within population subgroups.