MP35-10 IMPACT OF A STANDARDIZED TRAINING TUTORIAL ON AUTOMATED PERFORMANCE METRICS AND COGNITIVE WORKLOAD DURING ROBOTIC VESICOURETHRAL ANASTOMOSIS

Abstract

INTRODUCTION AND OBJECTIVES: We previously created a robotic vesicourethral anastomosis (VUA) tutorial with standardized hand gestures for each stitch based on the most efficient automated performance metrics (APMs) and least tissue trauma. In this study, we validate the training effect of this tutorial on robotic surgical novices. METHODS: 28 novices (no prior surgical experience) were randomized to gestures training (GT) or non‐gestures training (NGT) for 7 sessions (S1‐7). All participants were trained on synthetic VUA models (3DMed) and were provided with stitch location and needle driving direction (urethra/bladder, in/out). GT was further instructed on needle grip and wrist rotation direction according to the VUA tutorial. Automated performance metrics (APMs) (system events, Endowrist manipulation and instrument kinematic metrics) were captured using a systems data recorder (Intuitive Surgical). Cognitive workload was determined using task evoked pupillary response (TEPR, Tobii Pro Glasses 2) and measured by index of cognitive activity (ICA, Eyeworks). RESULTS: During S1, GT showed less instrument movement efficiency due to increased distance traveled (L arm: 52 vs 48 cm, p<0.012; R arm: 59 vs 48 cm, p<0.001) and greater instrument articulation (p<0.007). No differences in active instrument time, task completion time and ICA were observed. During S7, GT demonstrated faster completion time (33.4 vs 39.6 min, p<0.001) and superior instrument movement efficiency due to less distance traveled (L arm: 33 cm vs 39 cm, p<0.001) and decreased total active instrument time (p<0.001). Less left instrument articulation (p<0.016) was noted. GT also showed decreased ICA in the left eye, indicating a lower level of cognitive workload for spatial tasks (p≤0.05). From S1 to S7, both groups completed the task faster (GT: D35.8 min, p<0.001; NGT: D30.6 min, p<0.001) and more efficiently with less total distance traveled (p<0.001) and decreased total active instrument time (p<0.001). GT had a significantly greater improvement in movement efficiency, as measured by less distance traveled (D25 cm vs D13 cm, p<0.027). Decreased wrist articulation (p<0.001) was noted in both groups. CONCLUSIONS: Our study demonstrates that novices trained using a standardized robotic VUA tutorial (GT) complete the task faster, more efficiently and with less cognitive workload. While there is a decrease in efficiency as gestures are introduced, our findings suggest that overall efficiency is increased and demonstrates a more robust improvement over time.