Costs and Outcomes after Primary and Revision Hip Replacements by Implant Bearing Materials: Analysis of 550,760 linked Patient Records from English Routinely Collected Databases

Objective
We investigated 10-year revision risks, hospital admission costs, and quality-of-life for patients in the year after elective primary and revision total hip replacement (THR), overall and by implant bearing materials and fixation combination.

Methods
We analysed linked National Joint Registry, Hospital Episode Statistics, and Patient Reported Outcome Measures databases for adults undergoing elective primary THR in England (2009-2018). Implants were classified by bearing material combinations (cobalt-chrome and stainless-steel [CC/SS], delta or alumina ceramics, and highly or non-highly crosslinked polyethylene [HCLPE]) and fixation (cemented, uncemented, and hybrid). We estimated hazard ratios with flexible parametric survival models, and generalised linear models for costs and quality-of-life. Missing quality-of-life data was imputed using multiple imputation.

Results
We analysed 550,760 elective primary (mean age=69, SD=10.7) and 9,590 subsequent revisions THRs. On average, a primary THR costed, £10,365 (95% confidence interval £10,350,£10,381); quality-of-life after primary was 0.786 (0.785,0.786), and 10-year revision risk was 2.4% (2.08%,2.78%). First and second revisions costed, on average, £20,387 and £24,290; with mean quality-of-life scores of 0.609 and 0.502 , respectively. Compared with cemented CC/SS-on-HCLPE, cemented delta-on-HCLPE, uncemented delta-on-HCLPE, and hybrid delta-on-HCLPE, alumina-on-HCLPE, delta-on-delta and alumina-on-alumina were associated with lower one-year costs, improved quality-of-life, and lower 10-year revision risks.

Conclusions
Revision surgery costs double that of an elective primary with drastically lower quality-of-life. Bearing materials and fixation were associated with varying one-year costs, quality-of-life scores and revision risks. Although observational data may be affected by unmeasured confounding, our findings could help guide procurement decisions given changing trends in implant choice.