Session Title: Medical Education
Session Type: Abstract Submissions (ACR)
Background/Purpose: Published literature shows that internists have poor confidence in their ability to perform the musculoskeletal (MSK) exam. Confidence is an important self-assessment; lack of confidence drives individuals towards improvement efforts, but lack of confidence in physical exam skills can yield reliance on expensive imaging modalities and over-referral to specialists. Therefore, educators should develop effective interventions to teach the MSK exam and determine those efforts’ effects on practitioners’ confidence. There are no published validated surveys to assess confidence in performing a comprehensive MSK exam.
We designed a pilot MSK exam confidence survey based on a previously validated 15-item MSK student assessment for confidence in knee and shoulder exams. Our survey included 10 multi-item questions designed to measure confidence in several examination tasks for the knees, back/spine, shoulders, and hands/wrists. Confidence was measured on a 10-point Likert Scale (1=not at all confident to 10=very confident). The pilot survey was administered to internal medicine residents at Duke University before a didactic session on performing the MSK exam. We performed psychometric analyses of the instrument. Survey responses were analyzed in two dimensions: 1) the joint area in which confidence was assessed, and 2) specific task confidence (taking a pain history, assessing for swelling or range of motion, using information gained to make diagnoses or guide medical decision making, ability to perform provocative maneuvers).
19 surveys were collected (14 PGY1 residents, 4 PGY2, 1 PGY4). Only PGY1 responses were analyzed; 3 incomplete surveys were not analyzed. The discrimination for the entire test was 0.726 (joint domain R values 0.5 – 0.75), which is very good. The reliability was also high; Cronbach’s alpha for all scored items was 0.975 (joint domain values 0.906 – 0.925). For response analysis, confidence was highest for knee examination (mean 6.34) and lowest for hands (mean 5.92). In task domains, participants were most confident in assessing range of motion of joints (mean 7.71) and least confident in their ability to use the joint exam to guide medical decision making (5.05), identify and name deformities (5.182), use the joint exam to make diagnoses (5.27), and perform provocative diagnostic maneuvers (5.295). Extremes of confidence were rarely endorsed, giving an item frequency distribution that was highly skewed towards the center (overall item mean 6.016).
We established content and internal structure validity for a 10-question multi-item survey to assess residents’ confidence in their ability to perform a comprehensive musculoskeletal examination. Confidence was relatively similar among joint areas and varied more based on MSK exam task. Based on these results, we have modified the survey, contracting response options to 6 with a descriptive rather than numerical Likert scale to improve validity. Testing of this modified survey is ongoing. Future directions include validation by testing the relationship of measured confidence to actual performance of a comprehensive MSK exam.
L. G. Criscione-Schreiber,
K. S. O’Rourke,
« Back to 2013 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/development-of-a-tool-to-assess-internal-medicine-residents-confidence-in-their-musculoskeletal-examination-skills/