Factors affecting internal rotation after reverse shoulder arthroplasty

The limitation or loss of internal rotation (IR) after the surgery is a major problem in reverse shoulder arthroplasty (RSA). The particular factors associated with postoperative IR remain unclear. We aimed to analyze the predictors of IR after RSA.

We included primary RSA patients with the following implants and a minimum of 1-year follow-up: Wright Aequalis (n = 25), DJO Encore (n = 29), Biomet Comprehensive (n = 40), and Exactech Equinoxe (n = 29). Age, sex, dominant hand, primary diagnosis, implant type, preoperative critical shoulder angle, pre- and postoperative acromiohumeral distance, lateral offset of implant, glenosphere inclination, peg-glenoid rim distance (PGRD), glenosphere overhang, scapular notching, subscapularis (SSc) repair, pre- and postoperative ROM, and functional scores were assessed. IR was assessed using a 10-point scale based on the following anatomical levels: from the greater trochanter to the buttocks (2 points), from the sacrum to L4 (4 points), from L3 to L1 (6 points), from T12 to T8 (8 points), and from T7 to T1 (10 points). Univariable and multivariable analyses were performed to identify the factors affecting the IR after RSA.

One hundred twenty-three shoulders (123 patients) with a mean follow-up of 30.59 ± 19.55 (range, 12-83) months were evaluated. In the univariable analysis, preoperative diagnosis [odds ratio (OR) = 0.243, P = 0.001], implant type (P = 0.002), PGRD (OR = 1.187, P = 0.003), and preoperative IR (P < 0.001) were found to be factors associated with postoperative IR. Preoperative IR was the only factor associated with postoperative IR in the multivariable analysis (P < 0.001). Patients with preoperative IR scores of 10 or 8 points showed significantly better IR after RSA than those with preoperative IR scores of 2 or 4 points (10 points vs. 2 or 4 points; OR = 15.433, P = 0.002, 8 points vs. 2 or 4 points; OR = 6.078, P < 0.001).

Preoperative IR was the only independent factor for IR after RSA. Patients with excellent preoperative IR had better postoperative IR than those with poor preoperative IR. Preoperative diagnosis, implant type, SSc repair, implant lateralization, glenosphere inferior placement, and scapular notching were not identified as independent predictors of IR after RSA in our sample size.