Source:Choudhry NK, Fletcher RH, Soumerai SB. The relationship between clinical experience and quality of health care.
Ann Intern Med.
2005
;
142
:
260
–273.

One of the basic assumptions in medical education is that with clinical experience comes better clinical practice. The authors of this review from Harvard Medical School systematically searched the literature for reports of physician knowledge; adherence to standards of care for diagnosis, screening or prevention; adherence to standards of care for therapy; and patient health outcomes as they related to experience. Physician experience was variably defined in the studies as years in practice, time since graduation from residency or graduation from medical school, or age. For a study to be included, it was required to be an original report, measure 1 of the outcomes of interest, and include some variable of physician experience. The 59 articles they found reported on a total of 62 outcomes. However, the studies used multiple techniques for measurement, including some without a quantitative element, which precluded applying formal meta-analytic tools to combine the study results. Instead, the authors categorized the results for each of the 62 outcomes into 6 categories: consistently negative, partially negative, no effect, mixed effect, partially positive, and consistently positive. To be categorized as consistently negative, a study had to report all outcomes having a statistically significant decrease in performance with increasing years in practice or age. Overall, 52% of the studies reported negative associations between experience and performance for all outcomes. Twenty-one percent reported negative associations for some outcomes studied, 21% reported no association, 3% reported a concave association (steady increase for several years, then declining performance), 2% reported positive associations for some outcomes studied, and 2% reported positive associations for all outcomes studied. These negative associations held true for all types of outcomes analyzed (knowledge, adherence to standards of care, and health outcomes such as mortality), even when controlled for study quality. In particular, their conclusions did not seem to be influenced by whether a study used self-reported behavior (which presumably could result in an idealized reporting of practice, rather than true practice) or patient chart audit.

Dr. Robbins has disclosed no financial relationships relevant to this commentary

Studies such as this one have continued to demonstrate declining physician performance and knowledge with time. Clinicians’ knowledge and skills appear to have a half-life. The markers of knowledge, adherence, and outcomes used in the studies summarized by the authors of this review seem reasonable: eg, aspirin use and mortality rates after myocardial infarction, screening for cholesterol, and screening for renal disease in diabetes.

The most alarming finding in this study was the steady 0.5% increase in mortality of patients with myocardial infarction for each year since the treating physician graduated medical school. This was true whether the physician was an internist, family practitioner, or cardiologist. Although these data are derived primarily from outcomes of adult patients, the results are likely applicable to children and their practitioners as well. Now the search for why this decline...

You do not currently have access to this content.