Using inpatient administrative data for ABIM research: Methodological and ethical concerns.

Abstract

The American Board of Internal Medicine (ABIM) has recently published studies in high-impact medical journals demonstrating an association between initial and continuing certification exam performance and outcomes of hospitalized patients. We share methodological and ethical concerns regarding the conclusions of this research. Methodological concerns include the conceptual approach of attributing outcomes of hospitalized persons to individual physicians' test performance and not considering the potential impacts of confounding and bias. Ethical concerns relate to waiving of informed consent from diplomats, the potential harm with linking data from exams to claims, and the professional harms from the stratification of exam scores. Because of the importance of certification, we suggest that the ABIM, hospitalist diplomates, and professional societies collaboratively take action to set standards for research regarding the impact of certification on outcomes to demonstrate the value of certification in routine clinical practice.