Abstract
Revision is a key negative outcome of joint replacements. The purpose of this abstract is to present revision risk curves for hip and knee replacements based on the most recently available national data sources. Having a better understanding of determinants of revision risk can help inform clinical and health care system improvements.
We explored revision risk of primary joint replacement stratified by key clinical, prosthesis, and surgeon-level factors using data from three databases managed by CIHI: the Canadian Joint Replacement Registry (CJRR), the Discharge Abstract Database, and the National Ambulatory Care Reporting System. To investigate early revisions, we used Kaplan-Meier analysis stratified by demographic factors to determine the risk of revision within up to five years of primary surgery. This analysis identified the primary cohort from the CJRR from April 1, 2012 to March 31, 2017 and was limited to mandatory reporting provinces (British Columbia, Manitoba and Ontario) to ensure maximal coverage of prosthesis information. Bearing surface was obtained from the International Prosthesis Library maintained by the International Consortium of Orthopaedic Registries (ICOR) and the International Society of Arthroplasty Registers (ISAR).
The total revision risk cohort contained 283,620 primary surgeries, of which 5,765 (2%) had at least one revision. For total hip arthroplasties, revision risk at the end of the follow-up period did not differ by age, by sex or across bearing surface (metal on cross-linked polyethylene [XLPE], ceramic on XLPE, ceramic on ceramic, metal on non-cross-linked polyethylene). For hemiarthroplasties in hip fracture patients, cement fixation was significantly associated with decreased revision risk. Surgeon volume had a positive effect on survivorship with cementless fixation (2.7% at one year [95% CI 2.3%, 3.1%] for high-volume vs 3.2% [2.7%, 3.7%] for low-volume). However, surgeon volume did not have an effect on survivorship with cemented fixation (2% at one year [95% CI 1.3%, 2.6%] for high-volume vs. 2% [1.4%, 2.6%] for low-volume). For total knee replacements, revision risk increased with decreasing age and male sex. For patients aged 75 and older, four-year revision risks were 1.5% (95% CI 1.3%, 1.7%) for women and 2.0% (1.7%, 2.3%) for men, but for patients under 55 years old, they were 4.3% (3.7%, 5%) for women and 5.9% (4.9%, 6.9%) for men. Additional results from the upcoming 2019 CJRR annual report, including data up to March 31, 2018, will be presented.
Revisions represent a key failure of the primary replacement, they are costly to the health care system and negatively affect patients' quality of life. CJRR's coverage is currently 72%, increased coverage and follow-up time will allow increasingly comprehensive reporting on hip and knee prostheses in Canada. Future work in this area involves exploring additional prosthesis attributes for stratification of revision risk curves and calculation of hazards ratios adjusted by age and sex.
