Purpose: How to give feedback is widely taught and assessed during Faculty Development programs. As part of such programs, clinical teachers can attend objective structured teaching sessions (OSTEs), during which they are asked to give feedback to simulated residents on different tasks. Study aimed at: -analysing the feedback content provided during these OSTEs; -evaluating the impact of the training phase, medical discipline, or observed task; -assessing the alignment between feedback content addressed by clinical teachers and content identified as essential by experts.

Methods: We conducted a multimethod study. Clinical teachers (N=89) from five departments were trained to give feedback to residents in a six-month training program. Before and after training, they completed three OSTE stations which focused on tasks involving communication, interprofessional, physical exam or procedural skills. We analysed feedback content descriptively. ANOVA test was applied to evaluate feedback contents' influencing factors (ie participants' training phase, medical discipline, type of task addressed). For each OSTE, we analysed the percentage of items identified as essential by 3 experts that were addressed by clinical teachers during the feedback.

Results: We analysed 317 feedback sessions and coded 5388 occurrences. Feedback content distribution was: targeted content (73%), other clinical content (20%), learning strategies (4%), and self-management/other (3%). Feedback was often negative (73%). The training phase did not influence the content addressed while the topic of the observed task and clinical teachers' specialization slightly did. Alignment between content identified by experts and addressed by clinical teachers during OSTEs was low (3-38%).

Conclusion: Clinical teachers give mostly negative and targeted feedback according to the task. The poor alignment in selecting key content to be addressed is striking and should be further explored since clinical teachers may address elements of competence more according to their personal preferences than to residents' needs and context priorities.