Knee replacement revision: an international comparison

AbstractBackground and purpose: The need for knee replacement revision arises from a combination of patient, prosthesis and surgeon factors. Registry data can help study these relationships. Pooling data from multiple registries may increase both reliability and generalizability. The study aim was to gain a multi-national overview of knee replacement revision, to inform best-practice and improve outcomes.Patients and methods: Data was obtained from the Swedish Knee Arthroplasty Register (SKAR), the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR), and the Kaiser Permanente Joint Replacement Registry (KPJRR) from the US. Procedure numbers, demographic characteristics, prosthesis factors, revision diagnoses and revision procedure information were used. Equivalent diagnosis groups were created to allow analysis. Similarities and differences between registries were determined, as were time-related trends, and meta-analytic techniques were used to estimate the influence of patient or prosthesis factors on revision. More detailed study of revision for instability was carried out.Results: Primary knee replacement incidence had increased and revision procedures too, but by a smaller amount. Most common reasons for revision were infection, loosening and instability. Revision for infection had increased. Practice variations were seen between registries, particularly with prosthesis constraint and patella component usage, and also over time, with bearing mobility and polyethylene type. All-cause revision rates were higher with posterior stabilized, cementless and mobile-bearing components. Posterior stabilized prostheses had increased revision for infection, fracture and loosening, mobile-bearing components had increased revision for arthrofibrosis, instability and patella reasons, cementless fixation revisions for wear and procedures where patella components were not used had increased revision for patella reasons. Further prosthesis factors related to revision for instability were non-cruciate retaining components and inserts made of non-cross-linked polyethylene and those >14mm thick. Insert exchange was the most common revision procedure, but fewer 2nd revisions were seen with a major revision using more constrained implants.Interpretation: Practice variation can partially explain between-registry differences in incidence, and variability in reasons for revision. Understanding interactions between prosthesis factors and revision can help inform prosthesis choices. Use of lower risk prostheses can improve all-cause revision, revision for specific reasons, and additionally, revision and 2nd revision for instability