The Estimated Lifetime Risk of Revision After Primary Knee Arthroplasty Is Influenced by Implant Design and Patient Age: Data From the National Joint Registries

The aim of this study was to determine the lifetime risk of revision surgery after primary knee arthroplasty (KA) according to implant choice and patient age. The risk of revision according to the implant type (unicondylar, unconstrained, semiconstrained, and fully constrained) was obtained from the National Joint Registry of England and Wales. Mortality risk according to age was estimated from the Scottish Arthroplasty Project (1998 to 2019). The cumulative incidence of revision and death was calculated up to 20 years. The lifetime risk was calculated as a percentage using a multiple decrement life-table methodology. The lifetime risk of revision varied according to the implant, with unconstrained having the lowest risk and fully constrained and unicondylar having the highest risks, which increased with younger age. For an “average” patient aged between 65 and 69 years, the revision risks for an unconstrained (3.6%, 95% confidence interval [CI] 3.4 to 3.9), semiconstrained (7.2%, 95% CI 3.9 to 13.6), unicondylar (13.7%, 95% CI 12.4 to 15.2), and fully constrained (15.4%, 95% CI 9.2 to 25.7) total KA gradually increased. Relative to an unconstrained KA, the risk of revision for a semiconstrained implant was similar in patients aged < 65 years (relative risk < 2), whereas when this was employed in those ≥ 65 years, the relative risk was more than double (relative risk ≥ 2). Relative to an unconstrained KA, the risk of revision for a fully constrained or a unicondylar KA was more than double in patients aged < 55 years (relative risk > 2) and more than triple (relative risk ≥ 3) in patients ≥ 55 years The estimated lifetime risk of revision following KA was dependent on patient age and implant type, with fully constrained and unicondylar KA being associated with a higher lifetime revision risk. There are limitations in the data, and the indication for surgery and the reasons for revision were not assessed. III.