Purpose To assess correlation between (1) standard evaluations of internal medicine trainees and (2) results from an objective performance audit of preventive services.
Methods We used standardized patient chart abstraction (1,102 charts) to assess the percent of residents' patients receiving an appropriate, indicated preventive health service during their continuity clinic visits. Resident-level performance audit data were linked to their individual evaluation measures including selection interview scores, subjective ratings of performance by clinic faculty, USMLE Step-1 and in-training examination scores.
Results Of the 130 residents, 70% were male, 42% were interns, and 34% intended to practice primary care. Mean (standard deviation) clinic faculty rating was 4.25 (SD 0.55), range 3 to 5 on a five-point scale (1 = worst, 5 = best), and selection scores varied from 2.7 to 6.5 (mean = 4.2 SD = 0.56). Knowledge testing mean scores were USMLE Step-1 = 220 (SD 17) and in-training examination = 57 (8.2). The Table summarizes the mean percent (SD) of residents' patients who received the measure and correlation of individual resident performance measures with the other standard evaluation measures (*p < .05).
Conclusions Standard ratings and tests of knowledge were not strongly correlated with objective assessment of actual performance of these six preventive services. Thus, the performance audit seemed be a unique measure. Future research should assess correlation of standard measures with audit of additional performance measures.