Does the etiology of a failed hemiarthroplasty affect outcomes when revised to a reverse shoulder arthroplasty?

The purposes of this study were to evaluate patient outcomes after revision of hemiarthroplasty to reverse shoulder arthroplasty (RSA) based on initial pathology, to determine the re-revision rate, and to identify characteristics that may predict subsequent re-revision. A total of 207 shoulder hemiarthroplasty, bipolar prosthesis, and humeral resurfacing cases revised to RSA between January 2004 and January 2017 were reviewed. Outcome measures included shoulder motion and American Shoulder and Elbow Surgeons and Simple Shoulder Test (SST) scores. Sixteen RSAs underwent re-revision. A case-control study with each revised RSA matched to 4 controls based on age, sex, and minimum 2-year follow-up was performed to evaluate for factors predicting re-revision. The mean time from initial hemiarthroplasty to RSA was 3.6 years (range, 0.1-20 years). There were 114 patients with a minimum of 2 years' follow-up (mean, 57 months; range, 24-144 months). The most common initial diagnoses for hemiarthroplasty were fracture (n = 72), cuff tear arthropathy (CTA) (n = 22), and osteoarthritis (OA) (n = 20). Overall mean scores and range-of-motion values were as follows: American Shoulder and Elbow Surgeons score, 59 (95% confidence interval [CI], 54-64); SST score, 4 (95% CI, 4-5); forward flexion, 106° (95% CI, 96°-116°); and abduction, 95° (95% CI, 85°-105°). Compared with fracture cases, CTA cases had better forward flexion (P = .01) and abduction (P = .006) and OA cases had better SST scores (P = .02) and abduction (P = .04). The re-revision rate was 7.7% at a mean of 31 months (range, 0-116 months), with the most common diagnosis being fracture (10 of 16 cases). Humeral loosening (8 of 16 cases) was the most common failure mechanism, and larger glenosphere sizes were more likely to be revised. Functional outcome scores of hemiarthroplasty cases revised to RSA were better for patients with OA than for patients with CTA or fracture. Cases of hemiarthroplasty for fracture had decreased motion after revision to RSA compared with CTA and OA. Humeral loosening was the most common failure mechanism.