Journal of Shoulder and Elbow Arthroplasty · Nov 2026
Complication risks, implant survival, and functional outcomes after reverse total shoulder arthroplasty in elderly trauma patients: a competing risk analysis
By Lebe M, Unterfrauner I, Spross C et al.
Background Reverse total shoulder arthroplasty (RTSA) is increasingly used to treat displaced three- and four-part proximal humerus fractures (PHF) in elderly patients, but data on implant-related complications are limited and often confounded by the high competing risk of death. This study aimed to report functional outcomes after RTSA for trauma in older patients and to evaluate implant-related failure during the observed follow-up using competing risk analysis. Methods A retrospective cohort study using prospectively collected registry data was conducted at a tertiary centre. Consecutive patients aged ≥65 years undergoing primary RTSA for acute three- or four-part PHF or fracture-dislocation were included, with minimum 2-year follow-up. Standardised surgical technique and rehabilitation were used. Complications requiring intervention (dislocation, infection, periprosthetic fracture, aseptic loosening) and mortality were recorded. Functional outcomes were assessed using Constant Score (CS), relative Constant Score (RCS) and Subjective Shoulder Value (SSV) at 1, 2 and 5 years. Changes over time were analysed with one-way ANOVA. A competing risk analysis was performed, considering implant-related failure as the event of interest and death as a competing event. Results Two hundred and eight patients (51 males, mean age 79 years, SD 7 years; and 157 females, mean age 79.8 years, SD 7.0 years) were included, with up to 14 years follow-up; 113 (54.3%) were observed for >5 years. Twelve intervention-requiring complications (5.8%) occurred: four infections, three dislocations, four periprosthetic fractures and one aseptic loosening. Seventy-five patients (36.1%) died during follow-up; 90-day mortality was 1.4%. CS, RCS and SSV were in the good–excellent range at all time points, with values of approximately 62 and 70 points (CS), 90% and 99% (RCS), and 77% and 87% (SSV) at 1 and 5 years, respectively. The reported overall ANOVA p-values were p=0.25 for CS, p<0.01 for RCS, and p<0.001 for SSV. During the observed follow-up, implant-related failure was less frequent than death. Conclusion In this selected cohort of elderly patients treated with RTSA for complex PHF, implant-related failures were uncommon during the observed follow-up, while mortality was substantially more frequent. Functional outcomes were assessed through 5 years. When competing mortality is considered, these findings suggest that most patients are more likely to die with a functioning RTSA without requiring revision surgery than to undergo revision during the observed follow-up.