Four RCTs were included in the review. Three were included in meta-analysis (n=407, range 89 to 180). The fourth RCT reported data on a cohort (n=278) that overlapped with one of the other studies and was excluded from the meta-analysis. Only one RCT used an adequate randomisation procedure; the others were quasi-randomised. None used adequate blinding. Only one reported use of intention to treat analysis.
Two of the RCTs included in the meta-analysis reported the primary outcome: one found a significantly lower risk of revision hip surgery in the total hip arthroplasty group over 13-year follow-up (RR 0.28, 95% CI 0.12 to 0.66; n=180); and the other found no statistically significant difference between the groups at three years (n=89).
Results were inconsistent for pain outcomes: analgesia requirements were significantly lower in the total hip arthroplasty group at one year (RR 0.53, 95% CI 0.37 to 0.77; two RCTs) and 13 years (RR 0.12, 95% CI 0.05 to 0.30; one RCT); there was significant heterogeneity (I2=93%) for the one-year analysis.
There was no statistically significant difference between the groups in analgesia requirements at two years (one RCT). The total hip arthroplasty group had a greater likelihood of being ambulant at 13 years than the hemiarthroplasty group (RR 1.32, 95% CI 1.04 to 1.68; one RCT); the groups did not differ significantly at shorter durations of follow-up (two RCTs).
There was no statistically significant difference between the groups at any time-point in overall need for second anaesthetic over follow-up (three RCTs), mortality rate (two RCTS) and dislocation rate (two RCTs).