The survival of the glenoid implant in shoulder arthroplasty mostly depends on its positioning on the glenoid, which is made complex by glenoid deformities. These glenoid morphotypes have been classified by Walch and Favard, Patient-specific instrumentation (PSI) systems were developed to improve implant positioning. They are created from CT scans and preoperative planning, serving as a guide for the placement of the guidewire. To date, the systematic use of PSI remains controversial, and no study has evaluated its relevance according to glenoid morphologies. The aim of this study was to compare the benefit of using PSI according to glenoid deformities, relative to the freehand traditional technique.

We hypothesized that patient-specific targeting systems do not improve the accuracy of guidewire positioning compared to the traditional technique in cases of minimal deformities.

For this in vitro experimental study, 108 three-dimensional (3D) printed glenoids were created from 18 patient records eligible for reverse shoulder arthroplasty. These glenoids were classified according to Walch and Favard classifications and their degree of retroversion. The preoperative planning software used was Biomet Surgical Planning (Zimmer®), from which patient-specific instruments (PSI) were designed. Two surgeons positioned guidewires, with and without PSI, in these 3D models. Guidewire positioning was defined according to entry point, version, and inclination. These parameters were analyzed and measured by superimposing photographic images using GIMP software. Linear regression analysis was performed to determine for which deformities, according to Walch, Favard, or glenoid retroversion, PSI was necessary. Results are reported with beta coefficients and 95% confidence intervals. A p-value < 0.05 was considered statistically significant (power 80%, alpha risk 5%).

Across all glenoids, the comparison between PSI use and the standard freehand technique was statistically significant for entry point (p = 0.0006) and inclination (p = 0.011), but not for version (p = 0.25). No statistically significant differences were observed with or without PSI (0.1 < p-value < 0.8) for Walch type A1, A2, and C deformities, Favard type E0 and E4, and retroversion <10 °. The traditional technique demonstrated excellent intra-observer reproducibility, but poor inter-observer reproducibility (entry point, p = 0.43; 95% CI: −0.03–0.74).

Our study confirms that PSI should not be used systematically but rather adapted to glenoid morphology. Preoperative planning is essential to identify these deformities. PSI appears relevant in glenoids at higher risk of malposition—namely deformed glenoids (Walch B2, B3, D and Favard E1, E3, retroversion > 10) but does not provide benefit for minimally deformed glenoids (A2, C, or E4).

V; laboratory study.