Anand Malpani, Martin Curry, DO, Thomas Tantillo,, Amod Jog, Ray Blanco, MD, Patrick Ha, MD, Joesph Califano, MD, Jeremy Richmon, MD, Rajesh Kumar, PhD. Johns Hopkins University, Greater Baltimore Medical Center
Introduction: Robotic surgery training studies in the literature often focus on massed training composed of a small set of console operated training tasks. To train our subjects in every aspects of robotic head and neck surgery, we report on a novel robotic surgery training regimen integrating objective skill assessment and consisting of four training modules of increasing complexity, including procedure specific training for transoral base of tongue surgery now being performed with the da Vinci robot.
Methods: Over approximately one year in 2010-2011, 8 otolaryngology residents from an academic hospital participated in four distinct phases of robotic surgery training 1) didactic module, 2) operational skills module, 3) patient side system setup module, and 4) ex-vivo surgical extirpation of a simulated “tumor” located in the base of a porcine tongue. Trainees performed four iterations of each module approximately at a week’s interval. In addition to trainees, baseline performance data was obtained for four experts with two executions of each training module. Endoscopic and operating room scene video and instrument motion was recorded for each module and analyzed using offline automated analysis. All recorded sessions were also assessed by multiple experts using structured assessment (OSATS Likert scale).
Results: Study results show experts and trainees are well separated at the beginning of each training module. Computed automated measures (for example, average task completion time 943sec, std.dev. 227sec for experts vesus 1464sec, std.dev. 484sec for trainees for module 1 at week 1, and expert margin measurement time of 19.5 sec compared to an average time of 62.6sec for trainees for module 2 at week 2) correlate with OSATS assessment for each module. Subjective assessment by experts, and measurement of margins for the removed tumor verified the clinical utility of the stage 3 surgical environment. A survey of trainees consistently rated it as very useful in progression to human operating room assistance.
Conclusions: Structured multi-module training may provide a more complete training regimen for robotic surgery residents. Anecdotally, trainees performing their initial human surgeries have reported favorably on utility of their training experience. Automated objective assessment also promise to reduce the overhead for mentors, and measurements show trainees improving towards the better expert scores over the course of the training in each aspect of robotic operation. In contrast to current cumulative statistics and timed training rotations, we aim to use our developed methods and metrics to create a proficiency based regimen, where each trainee is graduated to the next module upon performing within the standard deviation of the experts to enhance overall efficiency of the training regimen.
Session Number: Poster – Poster Presentations
Program Number: P587