From image to clinical judgment: a randomized near-peer teaching trial on systematic chest radiograph interpretation

Chest radiograph (CXR) interpretation remains a fundamental yet challenging skill for undergraduate medical students, requiring both perceptual and integrative reasoning. Near-peer instruction and structured checklists have been proposed to enhance learning, but evidence from randomized designs is limited. We conducted a randomized educational trial among final-year medical students (n = 50). All participants attended a two-hour lecture on CXR interpretation delivered by intensive care medicine residents. The intervention group received additional training using a systematic checklist for image analysis. Performance was assessed through a 16-item test (interpretation and diagnosis domains). Nonparametric tests and Spearman correlations were applied; internal consistency and item discrimination were analyzed. Compared with controls, the intervention group achieved higher median scores in interpretation (6 vs. 3), diagnosis (4 vs. 3), and total score (10 vs. 6) (all p < 0.05). Interpretation and diagnosis scores were not correlated (Spearman ρ = 0.12, p = 0.40). Cronbach’s alpha was 0.592, with 56% of items showing excellent discrimination. A brief, resident-led, checklist-guided session significantly improved CXR interpretation and overall performance. Structuring perceptual search through systematic frameworks can rapidly strengthen diagnostic foundations and represents a feasible, low-cost addition to medical curricula.