Do medical license and board examination questions acknowledge the existence of patient preferences? An assessment of underpinning assumptions

BackgroundShared decision making (SDM) is endorsed by several professional societies yet inclusion in graduate and post-graduate medical education is inadequate. Assessment drives the learning process, but there has not been,... [ view full abstract ]

Background

Shared decision making (SDM) is endorsed by several professional societies yet inclusion in graduate and post-graduate medical education is inadequate. Assessment drives the learning process, but there has not been, to our knowledge, an analysis of the extent to which standardized examination questions recognize the existence of “preference-sensitive” decisions. Do medical license and board examinations acknowledge the existence of patient preferences? Or do they assume the existence of a one-solution system, based only on professional interpretation of medical evidence? Our aim in this study was to assess whether examination questions were formulated without recognizing the role of informed patient preferences in achieving appropriate medical decisions.

Methods

Questions from US Medical Licensing Examination (USMLE) I, II, III and American Board of Internal Medicine (ABIM) were purchased from National Board of Medical Examiners and two popular online question banks (USMLEWorld, USMLE-Rx, and MKSAP), and reviewed by trained assessors. Questions were excluded that did not involve clinical management – those involving choice of therapy, actions or tests. Preference-sensitive questions were defined by whether the patient’s informed preferences should play a part in arriving at the correct answer. This did not include questions that asked how providers should inform patients why their recommendation was best.

Results

A total of 444 (47.4%) of 937 questions reviewed were management-related and thus eligible for assessment. The percentages from USMLEWorld, USMLE-Rx, MKSAP and NBME were 17%, 41%, 35%, and 6%, respectively. Of all management questions, 68 (16%) came from USMLE I, 123 (27%) from USMLE II, 23 (5%) from USMLE III, and 230 (52%) from ABIM. No questions in USMLE I, II or III, and only four (1.7%) in ABIM were preference-sensitive.  

Conclusion

This study shows that in this sample of questions, medical examinations do not acknowledge the role of informed patient preferences.