Abstract
Background
The purpose of this study was to assess the overall clinical and radiographic outcomes of unicompartmental knee arthroplasty (UKA) in the 2–10 year postoperative period. The secondary goal was to compare outcomes between fixed- (FB) and mobile-bearing (MB) implant designs.
Methods
We performed a retrospective analysis of 237 consecutive primary medial UKAs from a single academic center. All cases were performed by high-volume fellowship-trained arthroplasty surgeons, though UKA comprised <10% of their overall knee arthroplasty practice (<20 medial UKAs per surgeon per year). Clinical outcomes included the Oxford Knee Scores (OKS) and revision rates. Femoral and tibial coronal and sagittal angles (FCA, FSA, TCA, TSA) were radiographically measured. FCA (>±10º deviation from the neutral axis), FSA (>15º flexion), TCA (>±5º deviation from the neutral axis), and TSA (>±5º deviation from 7º) outliers were defined. Far outliers were defined as measurements that fell an additional >±2º outside of these ranges. Outcomes were compared between the FB and MB groups.
Results
Overall, OKS scores improved significantly from 18.6 to 34.2 (p<0.0001) following UKA. The overall revision rate at an average 5.5-year follow-up was 14.3%. Only 48.9% and 46.4% of knees simultaneously fell within coronal and sagittal alignment targets for femoral and tibial alignment, respectively. Only 24.1% of all UKAs fell within target alignment in all four measurements. When comparing FB and MB knees, there was no difference in the overall revision rate (12.5% vs. 17.6%, p=0.280), nor were there differences in postoperative OKS (33.6 vs. 35.4, p=0.239) or outlier risk.
Conclusions
The proportion of UKA revisions and alignment outliers is greater than expected, even among high-volume surgeons. In general, implant design does not appear to significantly impact clinical outcomes, revision rates, or implant alignment. There was a trend for far outliers to have a higher rate of revision and lower OKS.
For figures, tables, or references, please contact authors directly.
