Abstract Rationale Residents in respiratory medicine need to acquire skills for the adequate management of non-invasive ventilation during sleep and in patients presenting with acute respiratory failure. Assessment of these skills usually rely on written based evaluations that do not assess practical skills. Objective structured clinical examination (OSCE) may be an interesting tool as it captures both theoretical knowledge and practical decision-making skills. Here, we aimed to assess if a OSCE based assessment using a high-fidelity simulation station (SIM) would provide addition information on resident’s skills when compared to written evaluations. Methods As part of the routine assessment of pulmonology residents in Paris, France, a high-fidelity OSCE station using a lung simulator was added to the standard written evaluation. The written test comprised four short-answer questions and twelve flow-pressure curve analyses (32 points, scaled to 20). The OSCE scenario involved a patient usually treated with CPAP for severe OSA who required adjustment of ventilatory support due to acute respiratory failure in the context of a COPD exacerbation. Students had access to the mannequin, various masks, and ventilation devices, with real-time display of ventilatory curves on a large screen, allowing them to visualize the effects of their interventions live. Eight items were evaluated (20 points total).. Each resident completed the two modalities (SIM and written test). Scores obtained by the two different modalities and correlation between modalities were analyzed (respectively with a Wilcoxon matched value test and Spearman correlation). The study was approved by the French Society of Pulmonology Ethics Committee. Results Seven residents were assessed across the two modalities. Median scores (IQR) were 11.6 10-13.7 for the written test and 9.5 4-11.5, p = 0.047 for SIM (figure 1). Frequency distribution and coefficient of variation and were wider for SIM (43%) than for written test (23%). No significant difference was found in overall combined scores across students (22% of total variance, p = 0.33). Inter-modality correlations was non-significant: r = 0.54 (p = 0.22) Conclusion Median scores were lower for the simulation than for the written test, reflecting the greater difficulty of applying knowledge in a clinical context. The wider dispersion and lack of correlation between modalities suggest that they assess distinct yet complementary competences. This abstract is funded by: None
Bianquis et al. (Fri,) studied this question.