Risk factors for total knee arthroplasty in rheumatoid arthritis.

We conducted a study to assess the predictive factors for total knee arthroplasty (TKA) in a cohort of rheumatoid arthritis (RA) patients recruited and followed prospectively for 5 years. A linked registry study using information from a large observational cohort of RA patients followed at the Institute of Rheumatology, Tokyo Women's Medical University (IORRA) was done. Baseline routine clinical and laboratory assessments were recorded. The data were analyzed using the multivariate piecewise-linear Cox (PL-Cox) regression model; the model initially included variables such as gender, age, duration of the disease, visual analog scale (VAS) generated by physicians (VAS-physician), patient-reported VAS for pain (VAS-pain), VAS for general health (VAS-GH), disability level using the Japanese version of the Health Assessment Questionnaire (J-HAQ), C-reactive protein, erythrocyte sedimentation rate, rheumatoid factor (RF), and hemoglobin. Of the 3945 patients registered at baseline, 955 (24.2%) had pain or tenderness in their knee joints, and 114 (11.9%) had TKA surgery in one or both knee joints. On PL-Cox regression, the variables with positive coefficients were J-HAQ, VAS-pain, VAS-physician, and RF positive; advanced age was associated with a reduced risk of TKA. The hazard ratios were: 0.920 for age >60 years; 2.64 for J-HAQ <1.5; 1.01 for J-HAQ >1.5; 1.47 for VAS-pain >6 (cm); 1.20 for VAS-physician >4 (cm); and 2.08 for RF positive. The consistently predictive factors for TKA in RA were age, J-HAQ, VAS-pain, VAS-physician, and RF positive. Age greater than 60 years was associated with a decreased risk of TKA, while J-HAQ from 0 to 1.5, VAS-pain >6 (cm), and VAS-physician >4 (cm) were associated with an increased risk for TKA surgery. These results suggest that, when treating RA patients, physicians should pay particular attention to pain complaints, the patient's daily activity level, and the RF factor status.