Comparable low revision rates of stemmed and stemless total anatomic shoulder arthroplasties after exclusion of metal backed glenoid components: a collaboration between the Australian and Danish national shoulder arthroplasty registries

Background The stemmed anatomical total shoulder arthroplasty is the gold standard in the treatment of glenohumeral osteoarthritis. However, the use of stemless total shoulder arthroplasties has increased in recent years. The number of revision procedures are relatively low and therefore it has been recommended that national joint replacement registries should collaborate when comparing revision rates. Therefore, we aimed to compare the revision rates of stemmed and stemless TSA used for the diagnosis of glenohumeral osteoarthritis using data from both the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) and the Danish Shoulder Arthroplasty Registry (DSR). Methods We included all patients who were registered in the AOANJRR and the DSR from January 1 2012 to December 2021 with an anatomical total shoulder arthroplasty used for osteoarthritis. Revision for any reason was used as the primary outcome. We used the Kaplan-Meier method to illustrate the cumulative revision rates and a multivariate cox regression model to calculate the hazard ratios. All analyses were performed separately for data from AOANJRR and DSR, and the results were then reported using a qualitative approach. Results 13066 arthroplasties from AOANJRR and 2882 arthroplasties from DSR were included. The hazard ratio for revision of stemmed TSA with stemless TSA as reference, adjusted for age and gender, was 1.67 (95% CI 1.34-2.09, p<0.001) in AOANJRR and 0.57 (95% CI 0.36-0.89, p=0.014) in DSR. When including glenoid type and fixation, surface bearing (only in AOANJRR) and hospital volume in the cox regression model the hazard ratio for revision of stemmed TSA compared to stemless TSA was 1.22 (95% CI 0.85-1.75, p=0.286) in AOANJRR and 1.50 (95% CI 0.91-2.45, p=0.109) in DSR. The adjusted hazard ratio for revision of total shoulder arthroplasties with metal backed glenoid components compared to all-polyethylene glenoid components was 2.54 (95% CI 1.70-3.79, p < 0.001) in AOANJRR and 4.1 (95% CI 1.92-8.58, p<0.001) in DSR. Conclusion Based on data from two national shoulder arthroplasty registries, we found no significant difference in risk of revision between stemmed and stemless total shoulder arthroplasties after adjusting for the type of glenoid component. We advocate that metal-backed glenoid components should be used with caution and not on a routine basis.