Abstract
Background and purpose - The long-term effects of computer-assisted surgery in total knee replacement (CAS) compared to conventionally operated knee replacement (CON) are still not clear. We compared survivorship and relative risk of revision in CAS and CON based on data from the Norwegian Arthroplasty Register. Patients and methods - We assessed primary total knee replacements without patellar resurfacing reported to the Norwegian Arthroplasty Register from 2005 through 2014. The 5 most used implants and the 3 most common navigation systems were included. The groups (CAS, n = 3,665; CON, n = 20,019) were compared using a Cox regression analysis adjusted for age, sex, ASA category, prosthesis brand, fixation method, previous surgery, and diagnosis with the risk of revision for any reason as endpoint. Secondary outcomes were reasons for revision and effects of prosthesis brand, fixation method, age (± 65 years), and hospital volume. Results - Prosthesis survival and risk of revision were similar for CAS and CON. CAS had significantly fewer revisions due to malalignment. Otherwise, no statistically significant difference was found between the groups in analyses of secondary outcomes. Mean operating time was 13 minutes longer in CAS. Interpretation - At 8 years of follow-up, CAS and CON had similar rates of overall revision, but CAS had fewer revisions due to malalignment. According to our findings, the benefits of CAS at medium-term follow-up are limited. Further research may identify subgroups that benefit from CAS, and it should also emphasize patient-reported outcomes.