RESUMO
BACKGROUND: Although simulation-based bronchoscopy has been shown to be an effective training modality, formal assessment should still be performed as new technology emerges. We sought to validate a simulator in essential bronchoscopic tasks, and survey perceptions of bronchoscopists on simulation. METHODS: A cohort study at 2 medical centers used 3 groups to assess construct validity of the Simbionix Bronchoscopy Simulator: 7 first-year fellows with <10 bronchoscopies each (novice), 6 pulmonologists with 200 to 1000 bronchoscopies each (experienced), and 7 pulmonologists with >1000 bronchoscopies each (expert). Participants were tested in 4 tasks (1: scope manipulation, 2: guided anatomic navigation, 3: airway anatomy, and 4: lymph node anatomy). Participants were scored and surveyed on their impressions of simulation. The means and Kruskal-Wallis test among groups were compared by task item (P<0.05). RESULTS: There were statistically significant differences in mean ranks among groups for tasks 1 and 3. For task 1, final score, time, mid-lumen time, and wall hits were discriminative (P=0.006, 0.006, 0.012, and 0.014, respectively). For task 3, time, bronchial segments identified, bronchial segments incorrectly identified, and bronchial segments skipped were discriminative (P=0.04, 0.012, 0.013, and 0.013, respectively). There was no statistically significant difference for task 2 and task 4. All participants agreed that simulation training is helpful and should be incorporated into bronchoscopic training. CONCLUSIONS: The simulator demonstrated validity in differentiating skill in scope manipulation and airway anatomy, but did not discriminate skill levels in anatomic orientation or identification of lymph nodes. Bronchoscopy simulation was viewed as helpful by all levels and should be considered before performance on patients.