Robotic endovascular peripheral arterial interventions: a proposal of a new learning model.

Abstract

Objective: This study tests a suitable model for training robot-assisted peripheral vascular interventions and examines the learning curves of endovascular surgeons with different levels of previous experience and main focus of work, analyzing procedure time, fluoroscopy time, use of contrast, and radiation emission.

Methods: Sixteen endovascular surgeons with different previous experience and training performed nine manual and 18 robotic angioplasties using the CorPath GRX platform on a 3D-printed life-size immersed infragenicular arterial phantom.

Results: All participants considered the model reliable. When analyzing manual angioplasty outcomes, the juniors took significantly longer to perform angioplasties than the seniors (p=0.044). Among the seniors, interventionists were faster only on the first angioplasty (p=0.046). Analysis of the robotic angioplasty results showed that only one junior failed to cannulate one of the target arteries once. The total duration, fluoroscopy time, and radiation emission did not differ between juniors and seniors (p=0.095, p=0.60, and p=0.64, respectively). In addition, the learning curves for the maximum benefit required two attempts for procedure duration, one for fluoroscopy time, and three for radiation emission. There were no significant differences between senior vascular surgeons and interventionists. Among juniors, residents had a significantly lower procedure duration (p=0.042) and radiation emission (p=0.046) only for the first angioplasty.

Conclusion: The learning curves for robotic peripheral arterial interventions were short, with a plateau for the procedure and fluoroscopy times and radiation emission after the third attempt. We observed no differences in the learning curves in relation to previous experience or training.