Diagnostic Trials

The term diagnostic trial is generally used in two different ways. A diagnostic trial type I describes studies that evaluate accuracy of diagnostic tests in detecting disease or its severity. Primary endpoints for these studies are generally test accuracy outcomes measured in terms of sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operating characteristics curves. Although establishing an accurate diagnosis or excluding disease is a critical first step to manage a health problem, medical decision-makers generally rely on a larger evidence base of empirical data that includes how tests impact patient health outcomes, such as morbidity, mortality, functional status, and quality of life. Therefore, the diagnostic trial type II evaluates the value of test results to guide or determine treatment decisions within a broader management strategy. Typically, differences in diagnostic accuracy result in differences in delivery of treatment, and ultimately affect disease prognosis and patient outcomes. As such, in the diagnostic trial type II, the downstream consequences of tests followed by treatment decisions are evaluated together in a joint construct. These diagnostic randomized clinical trials or test-treatment trials are considered the gold standard of proof for the clinical effectiveness or clinical utility of diagnostic tests. In this chapter, we define the variety of accuracy measures used for assessing diagnostic tests, summarize guidance on sample size calculation, and bring attention to the importance of more accurate reporting of study results.