Evaluating Medical Training Programs by the Quality of Care Delivered by Their Alumni

MEDICAL EDUCATION SERVES MANY GOALS, AND all of them are difficult to measure. For that reason, assertions that one medical school or residency program is better than another may have many different meanings. Nevertheless, most stakeholders, including prospective trainees, health systems, and patients, could be justified in expecting that graduates of good training programs generally take care of patients well, and that graduates of better training programs generally take care of patients better. The evaluation of medical education programs ought to keep this goal in sight. Because patient outcomes are already used to judge the performance of physicians and hospitals, a logical extension might be to evaluate training programs in part by the downstream outcomes of patients treated by their graduates. A key issue is whether patients have better outcomes if they receive care from physicians trained at top institutions. Although the institutional effect of a training site might not have as much predictive value as other physician characteristics, training pedigree might distinguish physicians, and it is easily observable. This idea has intuitive appeal, but also presents considerable challenges to its development and possible use. Several conceptual issues related to clinical quality-based evaluation of medical training must be considered, leading to a proposed research agenda.