RESUMO
INTRODUCTION: The lack of standardized skill training reported by medical students in performing tube thoracostomies may be associated with higher complications. The ideal training model is yet to be determined. This study sought to evaluate three different models. METHODS: Between 2015 and 2017, 204 last-year medical students of Universidade de São Paulo with no prior training in tube thoracostomy were randomized into three groups: cadaver, pig, and synthetic models. All groups performed 1-d tube thoracostomy hands-on training and a 40-min theoretical class. The knowledge acquisition was measured by a comparison between a theoretical test before and 3 wk after the class, and the skills improvement was evaluated by a comparison between the skills test on the same day of the hands-on training and another after 24 wk (the retention skill test). A questionnaire was submitted to evaluate their satisfaction rate and self-reported confidence, as per a Likert scale. RESULTS: The theoretical post-test score was higher compared to the pretest score in all groups (P < 0.001). The retention skills test in the cadaver and synthetic groups decreased compared to the skills test (P = 0.01 and P = 0.007, respectively). There was no difference between the groups either in the theoretical test or in the skills test. Student satisfaction was higher in the cadaver and pig groups. The confidence perception increased in all groups after the training. CONCLUSIONS: The models used for tube thoracostomy training appear to have a similar impact on skills retention, knowledge acquisition, and confidence. Although the satisfaction rate is lower for the synthetic model, it has no biological risk or ethical issues and is more feasible.