Council on Medical Student Education in Pediatrics

COMSEP Logo

Search This Site

Journal Club

Williams, R G.; Klamen, D L.; Mayer, D; Valaski, M; Roberts, N K.
A Sampling Strategy for Promoting and Assessing Medical Student Retention or Physical Examination Skills. Academic Medicine. 82(10) Suppi:S22-S2S, October 2007.


Williams, R G.; Klamen, D L.; Mayer, D; Valaski, M; Roberts, N K.
A Sampling Strategy for Promoting and Assessing Medical Student Retention or Physical Examination Skills. Academic Medicine. 82(10) Suppi:S22-S2S, October 2007.

Reviewed by Michael Barone, MD

Acquisition of clinical skills requires explicit teaching and practice. However, regular and complete assessment of physical examination skills can be lime intensive for students and faculty. The authors sought to detennine the utility of a strategy for regu lar testing ofphysical examination skills on a sampling basis for purposes of promoting rehearsal and assessing retention of physical examination skill competency.

The investigators used total physical examination perfonnance data on 548 second year medical students (accrued from 2002-2004) at the University of Illinois at Chicago during the Essentials of Clinical Medicine course. Examination items were grouped into component "systems" (for example; Lungs, thorax, breast and abdomen) with the goal of creating "practical, logical, and psychometrically sound" samples that were good predictors of the gold standard (total exam performance). Whereas the total exam takes 120 minutes for a student to complete, the component exams ranged from 18-33 minutes in duration.

The average proportion of items correct on the total physical exam score was 88% with a reliability of 0.90 (very acceptable) each year. Sample item groups in the following categories were then examined: 1) Head, Neck, ENT 2) Lungs, Thorax, Breast, Abdomen, 3} Heart, Pulses, Vitals 4) Musculoskeletal 5) Neurologic. These component exams contained 21 -39 exam items compared with the 137-1 38 items of the full exam. The average percent correct for each of the component tests ranged from 82.67% to 90.67% with reliability estimates ranging from 0.69 to 0.91 (only the musculoskeletal and neurologic samples had reliabilities above 0.80, considered the necessary standard for a high stakes examination).

The correlation of total lest score with component test score ranged from 0.60-0.75. The authors also sought to detennine the sensitivity and specificity of accurate classification of students following the component exam using the total exam score as the gold standard. When this was done, sensitivity was defined as the percent ofdeficil!nt students classified correctly. Values ranged from 50-100% depending on the sample. Specificity, defined as the percent of students correctly classified as competent ranged from 58-8 t %. The authors conclude that these component examinations may push students to practice physical examination skills and could be incorporated imo OSCEs. Limitations of the study include its cross sectional design and limitation to second year students.

This paper represents a significant step in demonstrating that shorter. component examinations of clinical skills, especially musculoskeletal and neurologic exams, may correlate with students' "total" skill set. Using this infonnation, it seems worthwhile to administer smaller. component exams periodically. This can be done outside a course or perhaps as part of a clerkship OSeE. We must, however. make sure that these component exams have the necessary characteristics to be psychometrically sound (appropriate number of items, etc.). Students who fail these smaller exams may be candidates to repeat a comprehensive physical exam assessment. Students who pass these exams easily probably don' t warrant a comprehensive re-screening of their skills before graduation.

Ed note: As the authors note, the primary purpose fordeveloping and using these physical examination tests is primarily an instructional one. The next steps are I) to determine if students actually do the examination correctly while on their clinical rotations and 2) most importantly, can they interpret their findings. After seeing innumerable students carefully palpate the necks ofchildren and announce that the child has no lymph nodes, I am keenly interested in the application . of the skill (see also Margaret Golden's review in this edition of the Educator). WVR

Return to Journal Club