Predictors of acromial and scapular stress fracture after reverse shoulder arthroplasty: a study by the ASES...

Acromial (ASF) and scapular spine (SSF) stress fractures are well-recognized complications of reverse shoulder arthroplasty (RSA), but much of the current data are derived from single-center or single-implant studies with limited generalizability. This study from the American Shoulder and Elbow Surgeons (ASES) Complications of Reverse Shoulder Arthroplasty Multicenter Research Group determined the incidence of ASF/SSF after RSA and identified preoperative patient characteristics associated with their occurrence.

Fifteen institutions including 21 ASES members across the United States participated in this study. Patients undergoing either primary or revision RSA between January 2013 and June 2019 with a minimum 3-month follow-up were included. All definitions and inclusion criteria were determined using the Delphi method, an iterative survey process involving all primary investigators. Consensus was achieved when at least 75% of investigators agreed on each aspect of the study protocol. Only symptomatic ASF/SSF diagnosed by radiograph or computed tomography were considered. Multivariable logistic regression was performed to identify factors associated with ASF/SSF development.

We identified 6755 RSAs with an average follow-up of 19.8 months (range, 3-94). The total stress fracture incidence rate was 3.9% (n = 264), of which 3.0% (n = 200) were ASF and 0.9% (n = 64) were SSF. Fractures occurred at an average 8.2 months (0-64) following RSA with 21.2% (n = 56) following a trauma. Patient-related factors independently predictive of ASF were chronic dislocation (odds ratio [OR] 3.67, P = .04), massive rotator cuff tear without arthritis (OR 2.51, P < .01), rotator cuff arthropathy (OR 2.14, P < .01), self-reported osteoporosis (OR 2.21, P < .01), inflammatory arthritis (OR 2.18, P < .01), female sex (OR 1.51, P = .02), and older age (OR 1.02 per 1-year increase, P = .02). Factors independently associated with the development of SSF included osteoporosis (OR 2.63, P < .01), female sex (OR 2.34, P = .01), rotator cuff arthropathy (OR 2.12, P = .03), and inflammatory arthritis (OR 2.05, P = .03).

About 1 in 26 patients undergoing RSA will develop a symptomatic ASF or SSF, more frequently within the first year of surgery. Our results indicate that severe rotator cuff disease may play an important role in the occurrence of stress fractures following RSA. This information can be used to counsel patients about potential setbacks in recovery, especially among older women with suboptimal bone health. Strategies for prevention of ASF and SSF in these at-risk patients warrant further study. A follow-up study evaluating the impact of prosthetic factors on the incidence rates of ASF and SSF may prove highly valuable in the decision-making process.