Do USMLE scores reflect an individual's clinical skills?
Standardized patients are frequently used to assess a medical student's clinical skills, including history taking, performing the physical examination, and interpersonal skills. There has been concern that medical school faculty may not be well suited to document these skills during routine evaluations because they infrequently observe students actually performing a history and physical secondary to time restraints and because the criteria by which students are judged may vary between different physicians. In part, this was the stimulus for the creation of the Clinical Skills portion of the USMLE Step 2. This raises the question of whether the USMLE adequately assesses one's clinical skills.
A 2005 study examined the relationship of performance on the Step 1 and Step 2 CK tests with performance in standardized patient situations in PGY1 year internal medicine residents. The results of both USMLE tests were found to correlate with each other (ie, students who performed well on Step 1 tended to perform well on Step 2). However, both tests had very low correlation with performance on a standardized patient evaluation. Despite being the more clinically oriented exam, Step 2 was no better correlated with standardized patient performance than Step 1.
However, the findings of a 2003 study comparing program director ratings of residents to their performance in medical school yielded somewhat contradictory results. This study found a correlation between class rank and Step 1 and Step 2 scores with program director rankings of skills during residency. This discrepancy is potentially attributable to the fact that what standardized patient exams assess may differ from what residency program directors value in a resident. Taken together, these studies imply that students with higher GPAs or test scores likely have the intelligence and motivation to acquire more of the clinical skills needed to become excellent physicians, although higher test scores do not by themselves mean that a student will have better clinical skills.
Paolo A, Bonamino
G. Measuring Outcomes of Undergraduate Medical Education: Residency Directors' Ratings
of First-year Residents. Academic Medicine Issue: 78(1), January 2003, p 90–95
Rifkin WD, Rifkin
A. Correlation between Housestaff Performance on the United States Medical Licensing
Examination and Standardized Patient Encounters. The Mount Sinai Journal of Medicine
72(5); January 2005, pp. 47-49.