Shoulder hemiarthroplasty: outcomes and long-term survival analysis according to etiology.

BACKGROUND: The indications for hemiarthroplasty in glenohumeral joint diseases remain controversial and depend mainly on the original underlying diagnosis. Our objective was to investigate the influence of the primitive aetiology on long-term prosthesis survival and on the Constant-Murley score.
MATERIALS AND METHODS: We studied 272 shoulders with the following diagnoses: fracture sequelae (n=73), primary osteoarthritis (n=67), cuff tear arthropathy (n=43), avascular necrosis (n=40), rheumatoid arthritis (n=31), and other (n=18). Of the 272 shoulders, 139 were evaluated after at least 8 years (mean follow-up, 134 months). In all, 30 prostheses required removal. Functional status was evaluated using the Constant-Murley score and survival rate using the Kaplan-Meier method with prosthesis removal or conversion to total arthroplasty as the endpoint.
RESULTS: Ten-year prosthesis survival was 88.13% overall, 100% in the rheumatoid arthritis group, 94.9% in the avascular necrosis group, 94.2% in the primary osteoarthritis group, 81.5% in the cuff tear arthropathy group, and 76.8% in the fracture sequelae (P=0.05). The mean Constant-Murley score after 8 years or more was 70.1 in avascular necrosis, 60.7 in primary osteoarthritis, 57.7 in fracture sequelae, 55.3 in rheumatoid arthritis, and 46.2 in cuff tear arthropathy (P=0.0006). The complication rate with the initial population as the denominator was 24.7% in fracture sequelae, 18.6% in cuff tear arthropathy, 15% in avascular necrosis, 8.9% in primary osteoarthritis, and 3.2% in rheumatoid arthritis.
CONCLUSIONS: The best indication for shoulder hemiarthroplasty is avascular necrosis and the worst indications are cuff tear and post-traumatic fracture sequellae. Rheumatoid arthritis and primary glenohumeral osteoarthritis are good indications in patients younger than 50 years of age. LEVEL OF EVIDENCE: Level IV, retrospective study.