Statistical and methodological considerations for vision screening

The statistical measures and definitions currently in use for screening programs are discussed. Sensitivity, specificity, false negative, and false positive rates are defined and illustrated. The positive and negative predictive values and the over- and underrefferal rates are also discussed, and an example of these measures is provided. The relationship of rates to prevalence and expected screening results is shown. A method for calculating the standard error of sensitivity and specificity is also given.