Introduction: Unicompartmental knee arthroplasty (UKA) is a proven for treatment of knee osteoarthritis (OA). Survival rates have been found comparable with total knee arthroplasty (TKA) in specialty hospitals series, but registry based studies show worse results of survival of UKA. High BMI, age of the patient, patellofemoral arthritis or learning curve have been found to have only mild consequences to the survival rates. Original indications for Oxford UKA in OA are severe pain and full thickness cartilage loss with bone-on-bone contact in the medial side. After widespread use of UKA surgeons are broadening their indications. Purpose of this study was to evaluate the influence of preoperative degree of OA on survival rate of UKA.

Material and Methods: 113 knees in 103 patients were operated with Oxford phase 3 UKA. We evaluated all the patient data retrospectively and patient age, body mass index (BMI), sex, earlier arthroscopies, operation time, follow-up time, preoperative medial joint space widths, reoperations and survival of UKA was recorded

Results: The mean age of the patients was 58 years (38–81) and mean follow-up time was 47 months (3–114). 22 UKAs were revised and the overall survival rate was 80.5%. 68% of revised knees have had undergone arthroscopy before UKA to confirm existence of arthritis. Odds ratio for female gender was statistically non-significant 1.59 (95% CI 0.57–4.45, p=0.46,). For BMI and patient’s age, the association remained non-significant with odds ratios of 1.07 (95% CI 0.98–1.17, p=0.14) and 0.96 (95% CI 0.90–1.02, p=0.19). Patients were divided four sub-groups according medial joint space width (medial joint space width ≤2 mm and > 2 mm) and Lateral/medial joint space width ratio (L/M-ratio ≤2.5 and > 2.5). Over 2 mm medial joint space width or L/M-ratio less than 2.5 were found significant risk factors for revisions, odds ratios being 6.00 (95% CI 2.12–17.00, p< 0.01) and 7.88 (95% CI 2.76–22.54, p< 0.01), respectively.

Discussion: Nowadays UKAs are performed on patients with mild OA against the original indications. In more severe OA varus alignment of the knee causes mechanical overload to the medial compartment, which is well corrected by UKA. Also it is possible that in the cases of prolonged knee pain caution is focused incorrectly to mild OA, which is typical radiological finding even in asymptomatic middle aged and elderly patients. Also in the early phase of OA it is impossible to estimate progression of cartilage damage in other two compartments. In conclusion we suggest that not to extend original indication of UKA and patient should have true medial bone-on-bone OA in preoperative radiographs. Performing UKA for patients with medial joint space width over 2 mm or L/M-ratio less than 2.5 should be concerned particularly careful.