Evaluating the utility, reliability, and validity of a resident performance evaluation instrument.

Abstract

The primary goal of this project was to create and evaluate the utility, validity, and reliability of an instrument assessing resident performance during child and adolescent psychiatry (CAP) training. In three stages, the instrument was developed and evaluated for utility in assessing skills critical to CAP. Items on the Global Assessment Evaluation (GAE) were derived from the six core competencies identified by the American College of Graduate Medical Education (ACGME). Secondary goals included producing an instrument that could be used by other CAP training programs and describing a process that could be followed by training programs to create their own assessment instruments. Faculty members developed a CAP resident performance evaluation instrument. In a three-stage process, faculty utilized the tool to evaluate residents during and after rotations. Statistical findings guided revisions to improve the utility and reliability of the instrument. For the final version of the GAE, intra-class correlation coefficients were calculated to assess inter-rater reliability, and Principal Components Analysis provided further insight into the dimensions of resident assessment. The final version of the GAE showed overall and construct validity by capturing significant differences among residents and matched faculty members' overall impressions of resident performance. Intra-class correlation coefficient values for the overall score (0.945) and individual scales showed good reliability. Resident performance was not correlated with rotation site or model of care. The GAE has proved a valuable instrument in tracking the progress, strengths, and weaknesses in resident performance over the course of training. Data from multiple evaluations over time provide useful information about resident performance in a way that one or few evaluations does not. This finding is consistent with the practice of semiannual reviews and the new Clinical Competency Committees, both of which are required by the ACGME.