The recommended surgical treatment of osteoarthritis (OA) of the shoulder remains controversial.

Recent published and presented trials evaluating hemiarthroplasty (HA) vs total shoulder arthroplasty (TSA) have been underpowered to detect a clinically relevant difference between the treatments.

A meta-analysis was carried out using methodology as described by the Cochrane Collaboration. Comprehensive search strategy was used including Medline, Science Citation Index, a review of references of relevant papers and abstracts of recent orthopaedic meetings. All articles were reviewed in a blinded fashion to select qualified articles based on population, intervention, outcome and methodological rigor. 4 RCT’s were identified. One was excluded due to selection bias introduced by the timing of randomization. The data of the 3 remaining trials was combined for analysis.

In each trial multiple tools were used to assess patient outcome however, the UCLA score was the only one consistent to all 3 studies.

The results demonstrated a statistically significant difference in change scores from baseline to 1 year (TSA 15.6; HA 11.2, p=0.012) in favour of the total shoulder arthroplasty group. This meta analysis demonstrated that in this selected group of patients, (primary osteoarthritis with intact rotator cuff and without excessive glenoid erosion), total shoulder arthroplasty provided a more predictable improvement in pain and function than hemiarthroplasty at 2 years.