Surgeon acceptance of an initial 3D glenoid preoperative plan: rates and risk factors

Although the effect of 3-dimensional (3D) planning for total shoulder arthroplasty (TSA) on component positioning and patient outcomes has been increasingly studied, the effect of 3D planning on surgeon decision making has not been well studied.

A retrospective review was performed of a database containing TSA cases for which the glenoid component was planned with a commercially available 3D computed tomography software program (Virtual Implant Positioning; Arthrex, Inc.) from 2016 to 2019. A total of 6483 cases planned by 417 surgeons were included. The glenoid version (V_tech) and inclination (I_tech) of the Virtual Implant Positioning technician plan as well as the surgeon's final plan for version (V_surg) and inclination (I_surg) were extracted. When the version and/or inclination of the surgeon plan matched that of the technician, that variable was defined as "accepted." The rates of acceptance of V_tech and I_tech were calculated and analyzed for association with implant type, native version and inclination, and running case count. A subgroup analysis of high-volume users (n > 30 cases) was analyzed to determine if any of the variables independently was associated with surgeon acceptance.

There was a very high rate of matching of version (66%), inclination (72%), or both (55%) and a low rate (18%) where neither parameter of the glenoid plan matched that of the technician. In univariate analysis, as the case count and retroversion increased the rate of accepting of version dropped noticeably (70%-50% and 47%, respectively [ P< .0001]). The rate of accepting the plan for inclination did not vary much as case count changed. In the multivariate analysis, 23 of 56 high-volume surgeons had at least 1 independent factor associated with accepting the technician-planned glenoid version, and 5 surgeons had 2 independent factors. In the multivariate analysis of matching glenoid inclination, 27 of 56 high-volume surgeons had at least 1 independent factor associated with accepting the technician-planned glenoid version, and 9 surgeons had 2 or more independent factors.

In a large database of TSAs with 3D-planned glenoids, there were high rates of cases with surgeon agreement with an initial plan provided by an industry technician: 66% in version, 72% in inclination, 55% for both version and inclination. Surgeon acceptance of the initial plan decreased as pathoanatomy increased and case count increased. Shoulder surgeons should be aware that an initial 3D preoperative plan provided by industry represents a potential source of cognitive bias in shoulder arthroplasty planning.