A multisite, multistakeholder validation of the accreditation council for graduate medical education competencies

PURPOSE: The Accreditation Council for Graduate Medical Education's (ACGME's) six-competency framework has not been validated across multiple stakeholders and sites. The objective of this study was to perform a multisite validation with five stakeholder groups. METHOD: This was a cross-sectional, observational study carried out from October to December, 2011, in the internal medicine residency continuity clinics of eight internal medicine residency programs in the Pacific Northwest, including a VA, two academic medical centers, a military medical center, and four private hospitals. The authors performed a cultural consensus analysis (CCA) and a convergent-discriminant analysis using previously developed statements based on internal medicine milestones related to the six competencies. Ten participants were included from each of five stakeholder groups: patients, nurses, residents, faculty members, and administrators from each training site (total: 400 participants). RESULTS: Moderate to high agreement and coherence for all groups were observed (CCA eigenvalue ratios ranging from 2.16 to 3.20); however, high differences in ranking order were seen between groups in four of the CCA statements, which may suggest between-group tension in these areas. Analyses revealed excellent construct validity (Zcontrast score of 5.323, P < .0001) for the six-competency framework. Average Spearman correlation between same-node statements was 0.012, and between different-node statements it was -0.096. CONCLUSIONS: The ACGME's six-competency framework has reasonable face and construct validity across multiple stakeholders and sites. Stakeholders appear to share a single mental model of competence in this learning environment. Data patterns suggest possible improvements to the competency-milestone framework.