Revision rates and patient-reported outcomes after shoulder hemiarthroplasty vs. anatomic total shoulder arthroplasty in patients younger than 60 years: a cohort study with data from the Dutch Arthroplasty Register (LROI).

Abstract

Although anatomic total shoulder arthroplasty (aTSA) is standard of care in older patients with an intact rotator cuff undergoing shoulder arthroplasty for osteoarthritis (OA), rheumatoid arthritis (RA), late post-traumatic sequelae (LPTS), or avascular necrosis (AVN), superiority of aTSA vs. shoulder hemiarthroplasty (HA) remains controversial in younger patients. This study aimed to compare these 2 procedures in patients aged <60 years regarding (1) overall and diagnosis-specific revision rates and (2) overall 1-year change in pain and function. A retrospective cohort study was performed with data from the Dutch Arthroplasty Register (LROI) from 2014-2022. All patients with OA, RA, LPTS, or AVN aged <60 years undergoing HA or aTSA were included. Cox proportional hazards regression analysis was performed to compare the incidence of revision in these 9 years, adjusting for diagnosis, Walch classification, and year of surgery. Also, the 1-year postoperative change on the numeric rating scale for pain during activity (NRS, range 0-10) and the Oxford Shoulder Score (OSS, range 0-48) were compared between these 2 procedures using multivariate linear regression analysis. Multiple testing was adjusted for as proposed by Benjamini and Hochberg. In total, 521 HAs and 1056 aTSAs were included (median age 54.0 years, 73.6% OA, 4.4% RA, 12.0% LPTS, and 10.0% AVN). The overall unadjusted 8-year revision rates were 21.1% (95% confidence interval [CI] 16.9%-26.4%) and 11.3% (95% CI 8.8%-14.5%) in the HA and aTSA group, respectively. In the total study population, HAs had a 1.77 (95% CI 1.24-2.53) times higher hazard on revision after adjustment for potential confounders. In the subgroup of OA or RA, the hazard ratio was 2.21 (95% CI 1.47-3.34) in favor of aTSA, and in the subgroup of LPTS or AVN, this ratio was 1.13 (95% CI 0.59-2.18). In the whole study population, 16.5% of patients provided patient-reported outcome measure data, and although there was no statistically significant difference for the adjusted 1-year change on the NRS between the groups, the OSS showed a 5.0 (95% CI 0.5-9.5) points greater improvement in the aTSA group. In younger patients with an intact rotator cuff, aTSA is superior to HA regarding medium-term revision rates when performed for OA or RA, whereas revision rates are comparable in case of LPTS or AVN. Also, although the survey response rate was limited, overall patient-reported shoulder function may undergo a relatively greater 1-year improvement after aTSA than after HA.