The Laparoscopic Feasibility Index: A New Approach to Formally Classifying the Difficulty of a Robotic Case

Abstract

Study Objective To introduce a new method for formally classifying the difficulty of a robotic case. Design Retrospective chart review Setting Las Vegas, Nevada Patients or Participants Female patients undergoing robotic gynecologic surgery from 2005-2019 by a single gynecologic oncologist (n=1,591). Interventions N/A Measurements and Main Results A high volume gynecologic oncologist developed an index, the Laparoscopic Feasibility Index(LFI). The index had a score from one to five. A score of 1 meant the robotic case could have been done just as easily using conventional laparoscopy. A score of 5 meant the surgeon would not even attempt the case using conventional laparoscopy due to extreme challenges and a high perceived risk of conversion to open. Scores were assigned at the end of each case. We examined 1,591 robotic cases performed between 2005 and 2019. The distribution of scores were: 1(n=463, 29.1%); 2(n=313,19.7%); 3(n=345, 21.7); 4(n=231, 14.5%); 5(n=239, 15.0%). Mean console times were: 0.53hrs (score=1); 0.78hrs (score=2); 0.89hrs (score=3); 1.02hrs (score=4); 1.43hrs (score=5). Use of the 4th arm was: 8.9%(score=1); 26.6%(score=2); 37.4%(score=3); 52.6%(score=4); 89.4(score=5). Mean EBL was: 30.1cc(score=1); 45.9cc(score=2); 57.1cc(score=3); 81.1cc(score=4); and 94.5cc(score=5). All three of these trends were statistically significant. We created a regression model to predict the probability that a case had an LFI score of 5 based on EBL, use of 4th arm, and operative time. The 1591 cases were divided into 2 halves using a random number generator. The model was developed on half of the sample and coefficients applied to the other half of the sample(the validation half). The correlation between predicted and observed probabilities in the validation half of the sample was 0.98. Conclusion The Laparoscopic Feasibility Index appears to have excellent internal validity based on our analysis of over 1500 robotic cases. This index could potentially be used to monitor a surgeon's learning curve and how they progress with the complexity of their robotic cases over time.