Background: Flexible bronchoscopy is an essential procedural skill to master for the management of respiratory illness. Bronchoscopy simulators hasten the attainment of competencies. Whether augmented reality (AR) offers advantages over virtual simulation (VR) during training is unknown. We tested this premise in novice bronchoscopists by assessing end-of-training performance after mastery learning on either simulator. Methods: A single-centre, single-blinded, randomised control trial. Thirty bronchoscopists were randomised to a VR (n=15) or AR (n=15) simulator training group. Each had up to 2 hours to practice full (diagnostic completion /18 segments, DC), orderly (structured progress, SP) and timely (procedure time, PT) segmental navigation using pre-defined mastery learning protocols. Experts assessed recorded complete test bronchoscopies on 2 different endobronchial airway training models (BronchoBoy®, Koken). Cognitive Load during training was compared. Results: The VR group outperformed the AR group in the first test (Bronchoboy), marginally for diagnostic completeness (DC) (median difference, p value) (1, p=0.0022) but not structured progress (SP) (1, 0.6887), procedure time (PT) (46 s, 0.8119) or mean intersegmental time (MIT) (2.15 s, 0.9760). In the second test (Koken), the AR group (trained on the Koken), was faster PT (-48 s, 0.0377) but no different from VR in DC, SP or MIT. The VR group trained for more time than AR (90+/-30min v 80+/-30 min, p=0.0468). The VR group had a significantly lower mental demand in training than AR (median difference, p value) (15 NASA-TLX points, P=0.023) but higher physical demand score (35 NASA-TLX points, P=0.004) Conclusion: Training by novices on VR or AR bronchoscopy simulators led to thorough, timely and moderately ordered test segmental navigation. VR training portended high and slightly more thorough navigation, with lower mental but higher physical demand and longer time spent training. Further comparative studies of VR and AR bronchoscopic skills training should seek preferential and predictive factors for optimal clinical proficiency.
Naik et al. (Fri,) studied this question.