Low success rate of closed reductions when treating dislocations after reverse shoulder arthroplasty: a study by the ASES Complications of RSA Multicenter Research Group

Background: Postoperative dislocation is a known complication following reverse shoulder arthroplasty (RSA), but treatment patterns and outcomes remain unclear. The purpose of this study was to identify treatment patterns, rate of successful closed reductions, and factors associated with unsuccessful closed reductions for dislocations after RSA in a large multicenter patient cohort. Methods: A multicenter retrospective review was performed for patients receiving primary or revision RSA from June 2013 to May 2019 across 15 institutions in the United States. Patients who sustained a postoperative shoulder dislocation (defined as complete loss of articulation between the humeral component and glenosphere confirmed on imaging) with a minimum of 3-month follow-up were included. The time from surgery to dislocation, nature of the dislocation, complications associated with the dislocation, initial treatment, success of closed reduction, recurrent dislocations, and subsequent treatments including revision procedures, were recorded. Univariate analysis was performed to identify patient factors associated with failure of an initial closed reduction attempt. Results: A cumulative postoperative dislocation incidence of 2.1% (n = 138) was observed in the 6621 patients undergoing RSA. The median time to dislocation was 7 weeks (interquartile range = 33 weeks), with 61.6% (n = 85) occurring within the first 90 days after surgery. Initial treatment consisted of closed reduction (n = 87; 63.0%), open reduction (n = 1; 0.7%), revision arthroplasty (n = 43; 31.2%), or benign neglect (n = 7; 5.1%). Those treated without an initial closed reduction had higher incidence of associated complications (45.1% vs. 14.9%). Among patients initially treated with a closed reduction, 27 (31.0%) were successful (required no further interventions), 15 (17.2%) subsequently required a revision procedure, and 43 (49.4%) sustained an additional dislocation. The only patient or surgical factor associated with an unsuccessful closed reduction was increased body mass index (BMI; 31.8 ± 6 vs. 28.9 ± 5.2, P = .02). Of the 43 patients who sustained an additional dislocation, 10 received another closed reduction and 30 received revision surgery. Among the 10 patients who received a second closed reduction, 5 remained stable (50.0%). Overall, 92 patients (66.7%) required a revision arthroplasty procedure during the study period, whereas 22 (22.5%) required multiple revision procedures. Ultimately, 18 patients (13.0%) remained unstable (benign neglect) at final follow-up. Conclusion: In this large multicenter series of postoperative dislocations following RSA, a closed reduction was initially attempted in the majority of patients, but only about one-third were successful and required no further intervention. Unsuccessful closed reductions were associated with higher patient BMI. Revision surgery for dislocations was complicated by a high rate of recurrent dislocations and rerevision surgery.