Abstract
INTRODUCTION
This study investigates the relationship between direct measurement of outcome and patient report of that outcome via the OKS. The stability of this relationship over time following surgery is also assessed.
METHODS
183 TKA patients were assessed pre-operatively and at 6, 26 and 52 weeks post-op. Oxford Knee Score was obtained along with measures of pain intensity, knee flexion, lower limb power and timed functional assessment. Correlation of performance variables with the OKS was assessed, and regression analysis performed on those that formed significant associations. Significance was accepted at p = 0.05.
RESULTS
Pain intensity and functional ability improved with time post-op. Functional scores correlated modestly with OKS, consistently over the assessment period. Pain correlated poorly pre-op but well post-op. Regression modelling was able to explain differing amounts of the variation in OKS at the 4 time points; 34.8% pre-operatively, 44.0% at 6 weeks, 56.9% at 26 weeks and 62.3% at 52 weeks post-op. Pain was the dominant factor in deriving all of these models.
DISCUSSION
Substantial variation was found in objective assessment of function and patient report of that function through the OKS. This relationship changed with time post-operatively. The explanatory power of the models improved as pain intensity reduced.
CONCLUSION
These results question the accuracy of PROMS in representing the observed ability to perform functional tasks, which has implications for assessing outcome.
