Chronic pain at 1-year following knee arthroplasty is associated with a worse joint-specific function and health-related quality of life.

Abstract

The purpose of this study was to identify independent variables associated with chronic pain (CP) at 1 year following knee arthroplasty (KA) and whether this influenced functional outcomes. This retrospective study was conducted over a 2-year period and included 2588 patients with completed Oxford knee score (OKS) and EuroQol (EQ)-five domains (5D) preoperatively and at 1 and 2 years postoperatively. The OKS pain component score was used to define patients with CP (≤14 points). The mean age was 70.0 (range 34-94) years and there were 1553 (60.0%) females. There were 322 (12.4%) patients with CP at 1 year. A worse preoperative EQ-5D (p = 0.025), EQ-visual analogue scale (VAS) (p = 0.005) and OKS questions relating to washing (p = 0.010), limping (p = 0.007), kneeling (p = 0.003) and night pain (p = 0.004) were independently associated with risk of CP. However, the preoperative OKS (area under the curve [AUC]: 72.0, p < 0.001) and EQ-5D score (AUC: 70.1, p < 0.001) were the most reliable predictors, with threshold values of <18-points and <0.300 being predictive of CP, respectively. Of those with CP at 1 year, 231 were followed up at 2 years, of which 92 (39.8%) had resolution of their CP. A worse response to OKS question 11 (ability to shop) and EQ-5D (p = 0.028) at 1 year was independently associated with persistent CP. Patients with CP had significantly (p < 0.001) worse OKS, EQ-5D and EQ-VAS at 1 year compared to those without. However, for those that had resolution of their CP at 2 years, their outcomes were clinically equal to those that did not have CP at 1 year. One in eight patients had CP at 1 year following surgery, which was associated with clinically worse knee-specific outcomes and quality of life. However, by 2 years, two in five patients had resolution of their CP and had functional outcomes clinically equal to those without CP at 1 year. The risk factors identified could be used to inform patients of their risk for CP and the potential for resolution. Level III retrospective study.