Confirmatory Testing Logic

Confirmatory testing logic applies Bayesian reasoning to sequential diagnostic testing. The post-test probability after a screening test becomes the pre-test probability for the confirmatory test. The first test is usually selected for sensitivity: it is designed to miss as few true cases as possible. That choice accepts some false-positive results. The second test is selected for specificity and, ideally, uses a different method or target analyte. It asks a new question rather than simply asking the first question again. Disease prevalence determines how much a positive result should change your thinking. Suppose a highly sensitive and specific test is used in 1,000 people when disease prevalence is 1%. Approximately 10 people have the disease; about 9.9 of them test positive. Of the 990 people without disease, approximately 9.9 may also test positive. A positive result in this setting is therefore not automatically persuasive. It may represent a false positive as often as a true positive. In a higher-risk population, the same test has a much higher positive predictive value. A second, informative test improves the situation because the group entering...