Learning curve for robotic assisted total knee arthroplasty: our experience with imageless hand-held Navio system.

Abstract

The main purpose of this study was to determine the learning curve of Robotic assisted Total Knee Arthroplasty surgery through assessment of operative time and comparison with that of conventional jig based Total Knee Arthroplasty. The study included our first 75 Robotic assisted Total Knee Arthroplasty and 25 randomly selected conventional jig-based Total knee arthroplasty from June 2017 to December 2017. The 75 cases were divided into 3 groups of 25 consecutive cases. The mean of operative time for each phase and total time was compared between the 3 groups and with the mean of total time for conventional jig based group. In our experience, Robotic assisted Total Knee arthroplasty was associated with a learning curve of approximately 25 cases. The mean for Registration phase of Group A (1st set of 25 cases) was 6.12min (SD 1.8min), group B (2nd set of 25 cases) was 4.46min (SD 0.79min) and group C (3rd set of 25 cases) was 4.17min (SD 0.59min). The mean for Planning phase of group A was 5.08min (SD 1.01min), group B was 4.04min (SD 0.37min) and group C was 4.01min (SD 0.35min). The mean for Cutting Phase of group A was 28.22min (SD 6.24min), group B was 22.49min (SD 0.79min) and group C was 22.36min (SD 0.88min). The mean for total time of group A was 39.42min (SD 8.02), group B was 31min (SD 1.22min), group C was 30.53min (SD 1.14min) and conventional group was 30.54min (SD 1.14min). On comparing the Registration phase (Group A vs B, p < 0.001; Group B vs C, p 0.14; Group A vs C, p < 0.001), Planning phase (Group A vs B, p < 0.001; Group B vs C, p 0.75; Group A vs C, p < 0.001), Cutting phase (Group A vs B, p < 0.001; Group B vs C, p 0.58; Group A vs C, p < 0.001) and Total time (Group A vs B, p < 0.001; Group B vs C, p 0.74; Group A vs C, p < 0.001; Group A vs Conventional, p < 0.001; Group B vs Conventional, p 0.17, Group C vs Conventional, p 0.99), the results showed that the inflection point for learning curve in our hands was 25 cases. The learning curve and increased operation theatre time are likely to be major barrier in widespread acceptance of robotic technology amongst arthroplasty surgeons. We, in our experience can say that the learning curve was approximately 25 cases. The results of this study will help the arthroplasty surgeons in accepting this technology and achieve better outcomes.