Evaluation of the use of patient-focused simulation for student assessment in a surgery clerkship.

BACKGROUND: The purpose of this study was to evaluate the use of simulated patients in conjunction with anatomic and tissue task-training models to assess skills. METHODS: Faculty reviewed the objectives of the clerkship to identify skills to be acquired. Three cases were developed related to rectal examination, suturing, and inserting intravenous lines and nasogastric tubes. Student scores were based on their ability to gather data from simulated patients and perform procedures on simulation models. RESULTS: A total of 670 students were assessed between 2006 and 2009. Alpha reliability coefficients were .97 for Communication/Interpersonal Skills, .71 for Procedures, and .58 for Data Gathering. Students receiving low ratings from faculty in the clerkship had significantly (P < .001) lower simulation scores. There were significant (P < .001) relationships between scores and grades in other clerkships. CONCLUSIONS: The combination of simulated patients and simulation models yielded reliable scores for procedural and interpersonal skills, and evidence of validity related to clinical ratings.

The relationship between performance on a medical school's clinical skills assessment and USMLE Step 2 CS.

BACKGROUND: Little is known about the relationship between performance on clinical assessments during medical school and performance on similar licensing tests. METHOD: Correlation coefficients were computed and corrected for measurement error using data for 217 students who completed a school's clinical assessment and took the Step 2 Clinical Skills (CS) examination. RESULTS: Observed (and corrected) correlations between the two tests were 0.18 (0.32) for Data Gathering, 0.35 (0.75) for Documentation, and 0.32 (0.56) for Communication/Interpersonal Skills. The highest correlation within each test was between Documentation and Data Gathering. The lowest was between Documentation and Communication/ Interpersonal Skills. CONCLUSIONS: The pattern of correlations supports each test's construct validity. The low correlations suggest that the tests are not redundant, and do not support using the scores on the school's assessment to predict performance on Step 2 CS. Future studies of these relationships need to address the time between the two assessments and the effect of intervening remedial programs.