Comparison of 3D computer-assisted planning with and without patient-specific instrumentation for severe bone defects in reverse total shoulder arthroplasty.

Background: Preoperative planning is an integral aspect of managing complex deformity in reverse shoulder arthroplasty (RSA). The purpose of this study was to compare the success of patient-specific instrumentation (PSI) and 3D computer-assisted planning with standard instrumentation (non-PSI) in achieving planned corrections of the glenoid among patients undergoing RSA with severe bony deformity requiring glenoid bone grafts.

Methods: A retrospective case-control study was performed, including all patients that underwent RSA with combined bone grafting procedures (BIO-RSA or structural bone grafting) for severe glenoid deformity by a single study between June 2016 and July 2023. Patients were required to have preoperative and postoperative CT scans as well as preoperative 3D planning performed for inclusion. Patients were divided into two groups based on the use of 3D computer-assisted planning with or without PSI (PSI vs. non-PSI). The corrected inclination and version were measured by two separate reviewers on preoperative and postoperative 2D CT scans and compared to their corresponding preoperative planning goals utilizing bivariate analyses.

Results: We identified 45 patients that met our inclusion criteria (22 PSI and 23 non-PSI). Preoperative inclination (mean ± SD) (PSI 10.12° ± 15.86°, non-PSI 9.43° ± 10.64°; P = .864) and version (PSI -18.78° ± 18.3°, non-PSI -17.82° ± 11.49°; P = .835) measurements were similar between groups. No significant differences in the mean deviation (error) between the postoperative and planned inclination (PSI 5.49° ± 3.72; non-PSI 6.91° ± 5.05; P = .437) and version (PSI 8.37° ± 5.7; non-PSI 5.37° ± 4.43; P = .054) were found between groups. No difference in the rate of outliers (>10° error) was noted in inclination (P = .135) or version (P = .445) between groups. Greater planned version correction was correlated with greater error when PSI was utilized (PSI r = 0.519, P = .013; non-PSI r = 0.362, P = .089).

Conclusion: Both PSI and 3D computer-assisted planning without PSI (non-PSI) appear to be useful techniques to achieve version and inclination correction among patients undergoing RSA with severe glenoid deformity required glenoid bone grafting with no clear superiority of one method over the other. Surgeons should be aware that when utilizing PSI, slightly greater error in achieving version goals may occur as version correction is increased.