Multi-test screening and the chances of being normal

Screening programs involve responsibility for appropriate action on abnormal test results. When multiple-test screening batteries are used, a simple probability formula is commonly used to predict the proportion of healthy individuals who will have one or more abnormal test results occur by chance alone. This formula is valid only when the assumptions upon which it rests are met in the population being tested. In many situations the assumptions are not met and the formula overestimates the occurrence of abnormal results in healthy populations. Data for three screening programs involving blood chemistry test batteries on 769 patients document this overestimate and its magnitude. Clinical judgment, not misapplied probability theory, should guide the physician's strategy in evaluating abnormal results of screening tests.