Authors
Enoch Yu, Zier Zhou, Ingrid de Vries, Christina Rogoza, Lindsay Crawford, Andrew D Chung, Benjamin Y M Kwan
Published in
Current problems in diagnostic radiology. Nov 13, 2026. Epub Nov 13, 2026.
Abstract
High-quality narrative feedback is essential for Entrustable Professional Activity (EPA) assessments in competency-based medical education (CBME), providing residents with specific guidance on performance, identifying areas for improvement, and informing competency committees about progression. Despite widespread EPA implementation, feedback quality in radiology has not been studied, and understanding underlying factors influencing quality, especially radiology-specific contexts or components, may identify targets to improve consistency and effectiveness of assessment to enable feedback.
This retrospective study analyzed EPA assessments for radiology residents (2019-2024). Feedback quality was scored using the Quality of Assessment for Learning (QuAL; 0-5) tool by two independent reviewers. We assessed associations of QuAL with EPA characteristics like assessor timelag or case complexity using correlation, trend tests, and ordinal, logistic, or LASSO-regularized multivariate regression.
A total of 3,638 assessments from 27 residents and 332 assessors were analyzed. Overall narrative feedback quality was moderate (mean QuAL 2.00±1.67). 75.6% contained comments, while 24.4% were blank, 42.4% had suggestions, and 31.5% had comments connecting suggestions to evidence. Timely completion was associated with higher feedback quality across all QuAL domains (detail: ρ=-0.23; suggestions: OR=0.99/day; connections: OR=0.99/day; blank comments: OR=1.008/day; all p<0.001). Lower entrustment and higher case complexity were associated with having suggestions and connections (all p<0.001). Higher QuAL scores were seen for fluoroscopy and lumbar punctures and lower for simulation or spine/DVT settings. Cross-validated multivariable models showed moderate discrimination for detail, suggestions, connections, and blank comments (AUC 0.57-0.68), improving substantially with inclusion of assessor identity (0.72-0.85).
Timeliness and assessor identity were the strongest factors for narrative feedback quality, whereas trainee and contextual factors had smaller effects. Interventions targeting timely completion and assessor-level or setting-level variability may improve consistency and quality of narrative feedback in radiology EPA assessments.
PMID:
43052168
Bibliographic data and abstract were imported from PubMed on 19 Nov 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 1
- Comments 0