Identification of early prognostic factors for knee and hip arthroplasty; a long-term follow-up of the CHECK cohort.

BACKGROUND: Patients with the clinical symptoms of knee or hip osteoarthritis without solid X-ray features present a therapeutic dilemma. The question arises whether the decision for a surgical treatment should be based on the clinical presentation or the X-ray. OBJECTIVE: To determine prognostic patient factors for knee and hip arthroplasty when the X-ray does only show Kellgren and Lawrence grade 0-2 osteoarthritis. STUDY DESIGN: Nationwide prospective cohort study. METHODS: Participants of the Cohort Hip and Cohort Knee (CHECK) with KL 0-2 osteoarthritis on the X-ray were contacted to determine whether any knee or hip arthroplasty had taken place. A Cox proportional hazards regression analysis was performed to find baseline patient factors predicting the decision for arthroplasty. RESULTS: Regarding the knee, sex HR 0.207 P = 0.030, BMI HR 1.081 P = 0.018 and WOMAC total sum score HR 1.022 P = 0.017 were statistically significant predictors of the outcome arthroplasty. Age was not a significant predictor (P = 0.079). Concerning the hip, sex HR 2.103 P = 0.012, age HR 1.062 P = 0.022 and WOMAC total sum score HR 1.019 P = 0.029 were found to be statistically significant predictors for arthroplasty. BMI (P = 0.576), contralateral pain (P = 0.877) and health perception (P = 0.405) did not predict the end point hip arthroplasty. CONCLUSION: Predictors for knee arthroplasty were being female, having a higher BMI and a higher WOMAC total sum score. Predictors for hip arthroplasty were being male, having a higher age and a higher WOMAC total sum score. The incidence of arthroplasty was 5.1% (10.2 years) for the knee and 10.2% (9.7 years) for the hip.