Search

Objective: To explore designs for evaluating the prognostic and predictive value of medical tests and their effect on patient outcome., Study Design: Theoretical analysis with examples from the medical literature., Results: For evaluating the prognostic value of a test, one can include the test at baseline in prognostic studies. To evaluate the value of test in predicting treatment outcome, the test results can be used as baseline information in randomized controlled trials of treatment. To compare the prognostic or predictive value of two or more tests, the test result combinations can be used as baseline information. To evaluate the effect on patient outcome, randomized controlled trials of test strategies are an option. Randomization can apply to all tested or be restricted to specific subgroups, such as those with discordant test results, to increase the efficiency of trials., Conclusion: The prognostic and predictive value of medical tests can and should be evaluated, to demonstrate the test's ability to guide clinical decision making and to improve patient outcome. Various randomized designs can be used to evaluate the effects on testing on patient outcome.

Whether or not patients are better off from undergoing a diagnostic test will depend on how test information is used to guide subsequent decisions on starting, stopping, or modifying treatment. Consequently, the practical value of a diagnostic test can only be assessed by taking into account subsequent health outcomes. In the appraisal of diagnostic test studies, it is essential to discriminate between studies that report on the accuracy of a diagnostic test and studies that report on health outcomes of strategies that incorporate diagnostic tests. In a study that reports on diagnostic accuracy, a cohort of patients is subjected to at least two diagnostic tests: the index test and the reference test, the latter usually being the best method available to detect the target condition. The accuracy of the index test can be expressed in terms of sensitivity, specificity, or likelihood ratios. Studies that compare two or more strategies that incorporate diagnostic tests as well as therapeutic interventions should be approached differently. Such studies do not require expression of test accuracy in terms of sensitivity and specificity. The merit of diagnostic tests evaluated in such studies can be expressed by comparing relevant outcomes of both strategies. The effectiveness of such strategies can be compared similarly as the effectiveness of treatment. However, due to the fact that the effect of a diagnostic test on health outcome is not as direct as the effect of treatment on health outcome, the design of outcome studies reporting on diagnostic tests requires special attention. It is important to establish a clear link between the result of the test under study and subsequent therapeutic management. Furthermore, trial efficiency can be improved by moving the point of randomization from the decision point, whether or not to test, to the point where a decision has to be made regarding what to do with the positive test results.

In this era of evidence-based medicine, diagnostic tests cannot escape close scrutiny of their effectiveness. Sensitivity and specificity have up till now played a central role in the evaluation of diagnostic tests. These terms are not without their shortcomings when it comes to the characterisation of a test's true worth for patients. Randomised clinical trials are increasingly used for evaluation of medical tests and outlining of strategy. The indirect relationship between test results and health outcome creates additional challenges for designers of such trials.