Background and objectives: The aim of our study is to validate the registration of knee arthroplasty revisions in the Lithuanian Arthroplasty Register (LAR) and thus give an indication of the accuracy of the published revision rates. Materials and methods: A total of 4269 primary total knee arthroplasties (TKAs) registered in the LAR between 2013 and 2015 were included. Two years after surgery the patients were contacted by phone in order to inquire if they had been subject to revision. The information from the patients was then cross checked against what had been registered in the LAR, and in case of a revision not having been registered hospital charts were investigated. Thus, the patients were followed up with regarding revision and/or death until 2017. A true revision was defined as an addition, exchange, or removal of one or all components. Results: Out of 4269 primary TKAs, we managed to contact and interview 2769 patients. Nine small hospitals were not able to provide contact details (telephone numbers) for 533 patients (549 knees). Sixty-seven patients (67 knees) were deceased (data from the Lithuanian National Census Register) and a further 438 patients (565 knees) appeared to have a wrong or non-valid telephone number, leaving 3031 (3091 knees) patients being contacted. Of those, 262 patients (266 knees) refused to participate in the study which left 2769 responders (2825 knees). Sixty-one patients said that reoperation had been performed on the index knee within two years of their primary surgery. After checking with the clinics, 10 were surgical procedures on the knee but not true revisions by our criteria. Out of the 51 true revisions we found that 46 were registered to the LAR as revised, while five (9.8%) revisions were missing. Conclusions: We conclude that the Lithuanian Arthroplasty Register has a good completeness of registered revision TKAs as only 9.8% of revisions were missing.
Keywords: arthroplasty; interview; knee; register; revision.