Randomised controlled trials that compared total hip arthroplasty with hemiarthroplasty for treating displaced femoral neck fractures (Garden III, IV-type) in participants aged 65 or over were eligible for inclusion. Patients with severe cognitive dysfunction, previous artificial joint replacement or previous osteosynthesis were excluded. Outcomes of interest were: mortality (30 days); complications from dislocation, infection (30 days), reoperation (one year), mobility, pain; and operation details.
Mean patient ages ranged from 69 to 81 years. In most studies patients had displaced fracture of the femoral neck; one study was of patients with femur fractures. Where stated, types of hemiarthroplasty were bipolar or Moore. More than half of the studies were conducted in the UK.
Three reviewers independently selected studies and resolved disagreements by discussion.