Outcomes After Anatomic and Reverse Shoulder Arthroplasty for the Treatment of Glenohumeral Osteoarthritis A Propensity Score-Matched Analysis

Background: Reverse shoulder arthroplasty (RSA) is increasingly being utilized for the treatment of primary osteoarthritis. However, limited data are available regarding the outcomes of RSA as compared with anatomic total shoulder arthroplasty (TSA) in the setting of osteoarthritis. Methods: We performed a retrospective matched-cohort study of patients who had undergone TSA and RSA for the treatment of primary osteoarthritis and who had a minimum of 2 years of follow-up. Patients were propensity score-matched by age, sex, body mass index (BMI), preoperative American Shoulder and Elbow Surgeons (ASES) score, preoperative active forward elevation, and Walch glenoid morphology. Baseline patient demographics and clinical outcomes, including active range of motion, ASES score, Single Assessment Numerical Evaluation (SANE), and visual analog scale (VAS) for pain, were collected. Clinical and radiographic complications were evaluated. Results: One hundred and thirty-four patients (67 patients per group) were included; the mean duration of follow-up (and standard deviation) was 30 +/- 10.7 months. No significant differences were found between the TSA and RSA groups in terms of the baseline or final VAS pain score (p = 0.99 and p = 0.99, respectively), ASES scores (p = 0.99 and p = 0.49, respectively), or SANE scores (p = 0.22 and p = 0.73, respectively). TSA was associated with significantly better postoperative active forward elevation (149 degrees +/- 13 degrees versus 142 degrees +/- 15 degrees; p = 0.003), external rotation (63 degrees +/- 14 degrees versus 57 degrees +/- 18 degrees; p = 0.02), and internal rotation (>= L3) (68.7% versus 37.3%; p < 0.001); however, there were only significant baseline-to-postoperative improvements in internal rotation (gain of >= 4 levels in 53.7% versus 31.3%; p = 0.009). The overall complication rate was 4.5% (6 of 134), with no significant difference between TSA and RSA (p = 0.99). Radiolucent lines were observed in association with 14.9% of TSAs, with no gross glenoid loosening. One TSA (1.5%) was revised to RSA for the treatment of a rotator cuff tear. No loosening or revision was encountered in the RSA group. Conclusions: When performed for the treatment of osteoarthritis, TSA and RSA resulted in similar short-term patient-reported outcomes, with better postoperative range of motion after TSA. Longer follow-up is needed to determine the ultimate value of RSA in the setting of osteoarthritis.