Methodology for Comparative Effectiveness Research: Potential and Limitations

There is an increasing need for reliable information on the rela-tive value of medical interventions to assist decision making on both the individual and public health levels; addressing this need is the goal of comparative effectiveness research (CER). In reading the four articles1-4 describing CER methodology in this special series about CER, one may note that the methods described (and the arguments for and against certain ones) have been used for decades by statisti-cians, epidemiologists, and clinicians. Randomized Clinical Trials Allwould agree that, for a given question of interest in a specified clinical population, a randomized clinical trial (RCT) provides the most definitive evidence. Hahn and Schilsky1 describe some practice-changing RCTs involving chemotherapy, radiation therapy, diagnos-tic imaging, and cancer prevention; we would also mention RCTs of cancer screening.5,6 The results of these trials help individuals and