Abstract
The purpose of this paper was to investigate the predictability of outcome of a consecutive series of cemented unicompartmental or total knee replacements in a single surgeon series.
Between September 2006 and February 2009, ninety-nine cemented, fixed bearing TKR were performed with patellar resurfacing. 52 cemented Miller Galante (Zimmer) Tibio-femoral UKR were performed in the same time interval. The minimum follow up was 6 months. Oxford and AKSS knee scores were collected prospectively at pre-operative and at routine follow up appointments.
Pre-operative mean AKSS Knee score for TKR group was 33.9 and improved to 84.2 at 1 year. Mean scores for Tibiofemoral UKR were 40.4 improving to 84.3 at 1 year.
Pre-operative mean Oxford knee score for TKR group was 34.6 (28%) and improved to 16.6 (65%) at 1 year. Mean scores for UKR were 28.5 (41%) improving to 14.0 (71%) at 1 year. These data would suggest that unicompartmental replacement performs as well as TKR.
However, in the TKR group, 59% achieved a knee score >85 and 23% an Oxford score >80%. In the UKR group, 67% achieved knee score >85 and 45% an Oxford score >80%. Conversely, only 5% of TKR achieved knee score <50 and 20% Oxford score <50% whilst 10% of UKR had a knee score <50 and 26% and Oxford score <50%.
These data show that whilst mean outcomes for TKR and UKR look similar, TKR offers a more predictable outcome with fewer clinical failures but also fewer excellent results. UKR offers a more polarised set of outcomes with far more clinically excellent results but also more clinical failures. These data can inform the ongoing debate regarding the role of unicompartmental arthroplasty. Patient selection is clearly critical but remains an inexact process.
