Comparison of the safety and 24-36-month efficacy of medial unicompartmental knee arthroplasty versus total knee arthroplasty in elderly patients (≥ 75 years) with isolated medial-compartment knee osteoarthritis.

The number of very elderly patients presenting with end-stage knee osteoarthritis (OA) is rising rapidly, yet the comparative merits of medial unicompartmental knee arthroplasty (UKA) and total knee arthroplasty (TKA) in this group remain insufficiently characterised. We compared the perioperative safety and short-term clinical and patient-reported efficacy of the two procedures in patients aged 75 years and older with isolated medial-compartment OA. This single-centre retrospective cohort study compared patients aged ≥ 75 years undergoing primary medial UKA or TKA for isolated medial-compartment OA between January 2018 and December 2022. Among 268 eligible patients (68 UKA and 200 TKA), 1:2 nearest-neighbour propensity-score matching, incorporating operating surgeon together with demographic, comorbidity, radiographic and baseline functional covariates, yielded 60 UKA and 120 TKA patients. Perioperative outcomes, KSS, OKS, pain VAS, ROM, postoperative FJS-12, complications, readmission, reoperation and satisfaction were compared over 24-36 months. Standardised mean differences assessed baseline balance; treatment-effect estimates, risk differences and 95% confidence intervals (CIs) were reported. Matching yielded 180 patients, with all post-match absolute SMDs < 0.10 (maximum 0.087). UKA was associated with shorter operative time (64.3 vs. 84.2 min), lower estimated blood loss (99.9 vs. 183.6 mL), smaller haemoglobin decline (1.5 vs. 2.6 g/dL), shorter stay (3.0 vs. 5.2 days) and earlier independent walking (18.3 vs. 29.6 h). Final KSS components were similar; UKA showed higher OKS and FJS-12, greater knee flexion and lower pain VAS. Complications occurred in 3/60 UKA (5.0%) and 19/120 TKA (15.8%; risk difference - 10.8% points, 95% CI - 19.4 to - 2.3; sensitivity-analysis adjusted OR 0.31, 95% CI 0.09-0.92). Transfusion was infrequent (1.7% vs. 3.3%); readmission and reoperation showed no clear difference. Final satisfaction was 98.3% versus 86.7% (risk difference + 11.7% points, 95% CI 4.8-18.6). In carefully selected patients aged ≥ 75 years with isolated medial-compartment OA, medial UKA was associated with lower measured perioperative burden and more favourable patient-reported outcomes than TKA over 24-36 months. The observed difference in complications should be interpreted cautiously because event counts were small and findings were sensitive to the choice of statistical test. Retrospective allocation, surgeon and implant heterogeneity, potential residual confounding and limited follow-up preclude causal or long-term durability conclusions.