Metal-on-metal hip resurfacing compared with contemporary total hip arthroplasty : an analysis of National Joint Registry data.

Abstract

Our aim was to estimate the survival of hip resurfacing arthroplasty (HRA) currently available for implantation (males only, head size > 48 mm, MatOrtho Adept or Smith & Nephew Birmingham Hip Resurfacing (BHR)), performed by all surgeons, compared with total hip arthroplasty (THA) using contemporary bearings and fixation. We performed a retrospective analysis of the National Joint Registry. We included all males who had undergone BHR or Adept with a head size greater than 48 mm. Propensity score matching was used to produce two comparable groups of patients for HRA or contemporary THA. We matched in a 3:1 ratio (THA:HRA) using sex, American Society of Anesthesiologists grade, BMI group, age at primary procedure, surgeon volume, diagnosis, and surgeon grade as covariates. The primary analysis was survival at 18 years. Time to revision was assessed using Kaplan-Meier curves. Cox's proportional hazard models were used to investigate between-group differences with adjustment for matching factors. A total of 9,574 HRAs were available for analysis. After matching, the HRA and THA groups were well balanced in terms of covariates. Survival at 18 years was 92.7% (95% CI 91.9 to 93.3) in the HRA group and 94.2% (93.5 to 94.8) in the THA group. Despite these relatively similar estimates, the adjusted hazard ratio (HR) was 1.22 (95% CI 1.09 to 1.36; p < 0.001) for all-cause revision over the entire study period, indicating that the HRA group experienced higher probability of revision than the THA group. Survival of HRA was 92.7% at 18 years. However, the HR for all-cause revision for HRA was higher compared with THA.