Clinical Evaluation of Diagnostic Tests and Biomarkers in CER

Comparative Effectiveness Research (CER) emphasizes the study of outcomes that are of primary relevance to patients. CER also aims to generate evidence that can inform both health policy and individual patient choices. The evaluation of diagnostic tests and biomarkers poses special challenges in CER and is not governed by as fully developed a paradigm as is the evaluation of therapy. In particular, tests generate information, which is subsequently incorporated into further diagnostic and therapeutic decision making. Although diagnostic testing may have a direct effect on patient outcome in some settings, most effects of diagnostic information are mediated by therapeutic decisions. Thus the assessment and comparison of the effectiveness of diagnostic tests need to account for the effects of subsequent interventions in order to identify the difference in outcome that can be reliably attributed to the test. In addition, diagnostic technology evolves very rapidly and creates a virtual moving target problem. In this presentation we will discuss the methodologic challenges in the design and analysis of clinical studies of diagnostic tests and biomarkers.