Self-Learning Traditional vs Simulation-Based Education of Medical Residents for Mechanical Ventilation

Abstract

Background: Training of physicians on mechanical ventilation (MV) is highly variable among residency programs and little is known about best practices. In particular, few studies have compared online learning (eg, PowerPoint module) to computer-based simulation. The aim of this study was to evaluate the potential of a patient-ventilator simulator to facilitate MV education compared to traditional education with lecture slides, as a self-directed study of basic MV concepts. The hypothesis was the software-based patient-ventilator simulator results in more MV knowledge acquisition (ie, higher scores on a written test) compared to traditional education. Methods: The study was approved by the Cleveland Clinic (CCF) Institutional review board. Study participants were Internal Medicine Residents at CCF Main Campus and Fairview Hospital, whom we expect to have no substantive experience with MV and were enrolled during March and April 2019. Participants were randomly assigned to two learning tool groups; traditional lectures based on PowerPoint slides (TG) and patient-ventilator simulator programmed in Microsoft Excel (SG). The simulation-based group (SG) was also given an abbreviated set of slides that present the same content as the traditional group (TG) for definition of terms. In the SG, all graphic descriptions were presented with the simulator. Both groups were given the same pre-and post-test to evaluate knowledge acquisition. Outcome data were mean differences between the pre-test and post-test scores. Data were compared with t-tests and P < 0.05 indicated significance. Results: We enrolled 63 residents, of whom 33 were in the TG and 30 were in the SG. We excluded 5 residents due to not completing post-test questions. For the TG, the mean (SD) pre-test score (% correct) was 58.6 (11.8) and the post-test score was 78.5 (10.2). For the SG, the pre-test score was 58.8 (11.8) and the post- test score was 74.0 (9.9). The difference in pre-post test change (19.8 vs 15.2) was not significant (P = .17). The power of the study was only 0.145 due to the low sample size. Conclusions: This pilot study suggests that residents had relatively low knowledge of MV and gained substantially with either traditional or simulation methods. The lower mean scores of the simulation group may be explained by the unfamiliarity of the medium and longer time to complete. Further study is required to confirm these results and investigate underlying reasons.