Making student mistakes didactically useful in bedside teaching: relevance of learning content, teacher and patient presence

Abstract Background Student mistakes are common in bedside teaching (BST) and can serve as important learning opportunities. However, little is known about how frequently mistakes occur in authentic clinical settings, what types of learning content they involve, and how teacher and patient presence shape teachers’ handling of these mistakes. This study investigated (RQ1) the relevance of learning content during student mistakes, (RQ2) associations between mistake types and teachers’ reactions, and (RQ3) whether patient presence influences clinical teachers’ responses to mistakes. Methods We conducted an observational study based on video recordings of 36 BST sessions across internal medicine, neurology, and orthopaedics (78.4 h). Using a hierarchical coding scheme, we identified mistakes, classified their learning content and mistake types, and coded teacher reactions as well as teacher and patient presence. Interrater reliability ranged from substantial to excellent. Descriptive statistics were used to determine the frequency of mistakes per hour and their distribution across the learning content. Associations between mistake types and teacher reactions were analysed via Spearman’s rank correlations (RQ2). Mann–Whitney U tests (RQ3) were used when examining whether teachers’ mistake-related behaviours differed depending on the presence or absence of the patient. Results Across all BST sessions, the students made an average of 4.9 mistakes per hour. Mistakes occurred most often during theoretical knowledge and clinical reasoning, predominantly as reproduction mistakes, whereas an incorrect application of skills was most frequent during clinical examination and case presentations. Teacher presence was positively associated with the number of observed mistakes, particularly reproduction mistakes. During the physical examination, the students were supervised for only 72% of the active practice time. Teachers’ reactions - including elaboration, directness of feedback, and time provided for correction - did not differ significantly depending on patient presence. Conclusions Student mistakes are common in BST and hold considerable didactic value; however, this value is not fully realised when supervision is inconsistent or when feedback remains limited. Enhancing clinical teachers’ capacity to detect errors, providing structured and supportive feedback, and adapting their approach to the presence of patients may strengthen their ability to learn from mistakes. Interventions that cultivate psychological safety and promote effective error-management practices could further improve the pedagogical quality of BST.