The educational effectiveness in flexible bronchoscopic intubation training using three-dimensional printed difficult airway models for anaesthesia residents: a single centre randomised controlled study.

BACKGROUND

Flexible bronchoscopic intubation (FBI) is a vital skill for anaesthesiologists managing difficult airways. Traditional training models often fail to replicate real difficult airway scenarios, limiting their educational value. This study aimed to assess the effectiveness of three-dimensional (3D) printed difficult airway models, developed from patient CT/MRI images, in improving FBI skills among anaesthesia residents compared to conventional manikin models.

METHODS

Twenty-eight anaesthesia residents were randomised equally into the 3D-printed model group and the AirSim manikin group. Each resident received a standardised lecture, followed by FBI training using their assigned model and then a crossover with the alternate model. Educational effectiveness was assessed using a global rating scale, performance on the ORSIM simulator scenarios, and a trainee-reported questionnaire measuring suitability, realism, and educational effectiveness.

RESULTS

The global rating scales of operative performance and FBI time improved significantly after training in both groups. The post-training performance on the ORSIM simulator did not differ significantly between both groups. Compared with the AirSim manikin model, the 3D-printed model had a higher trainee-reported educational effectiveness (8.6 ± 1.0 vs. 7.9 ± 1.0; p = 0.050). Exploratory post-hoc analyses suggested that prior FBI experience modified the training response.

CONCLUSIONS

Both 3D-printed difficult airway models and conventional manikins were associated with improvement in FBI performance, with no statistically significant between-group differences in objective skill-acquisition outcomes. Trainees perceived the 3D-printed model as more educationally effective.

TRIAL REGISTRATION: Not applicable. The study was launched based on simulation models and without enrollment of patients nor animals; hence the protocol registration number was not attributed.