Sensitivity and specificity in prevalence studies: The importance of considering uncertainty