To assess the cardiovascular physical examination skills of internal medicine housestaff.

DESIGN:

Cross-sectional assessment of housestaff performance on three valvular abnormality simulations conducted on the cardiology patient simulator, "Harvey." Evaluations were done at the beginning (session I) and end (session II) of the academic year.

SETTING:

Duke University Medical Center internal medicine training program.

SUBJECTS:

Sixty-three (59%) of 107 eligible internal medicine housestaff (postgraduate years 1 through 3) agreed to participate and completed session I; 60 (95%) completed session II.

MEASUREMENTS:

All volunteers were tested on three preprogrammed simulations (mitral regurgitation, mitral stenosis, and aortic regurgitation).

RESULTS:

The overall correct response rates for all housestaff were 52% for mitral regurgitation, 37% for mitral stenosis, and 54% for aortic regurgitation. No difference was noted in correct response rates between sessions I and II. For mitral regurgitation, correct assessment of the contour of the holosystolic murmur predicted a correct diagnosis (P = 0.002). For mitral stenosis, identification of an opening snap and proper characterization of the mitral area diastolic murmur predicted a correct diagnosis (P < 0.0001). No individual observations were noted for the aortic regurgitation simulation, whose identification by the housestaff was associated with a correct diagnosis.

CONCLUSIONS:

Medical educators have had a growing sense that proficiency in physical diagnostic skills is waning, but few data have examined the question critically.

OBJECTIVE, DESIGN, AND SETTING:

To compare the cardiac auscultatory proficiency of medical students and physicians in training. A multicenter cross-sectional assessment of students and house staff. A total of 8 internal medicine and 23 family practice programs of the mid-Atlantic area.

PARTICIPANTS:

A total of 453 physicians in training and 88 medical students.

INTERVENTIONS:

All participants listened to 12 cardiac events directly recorded from patients, which they identified by completing a multiple-choice questionnaire.

MAIN OUTCOME MEASURES:

scores were expressed as the percentage of participants, for year and type of training, who correctly identified each event. Cumulative scores were expressed as the total number of events correctly recognized. An adjusted score was calculated whenever participants selected not only the correct finding but also findings that are acoustically similar and yet absent.

RESULTS:

Trainees' cumulative scores ranged between 0 and 7 for both internal medicine and family practice residents (median, 2.5 and 2.0, respectively). Internal medicine residents had the highest cumulative adjusted scores for the 6 extra sounds and for all 12 cardiac events tested (P=.01 and .02, respectively). On average, internal medicine and family practice residents recognized 20% of all cardiac events; the number of correct identifications improved little with year of training and was not significantly higher than the number identified by medical students.

CONCLUSIONS: