Comparison of PREVENT with the pooled cohort equation for estimating atherosclerotic cardiovascular disease risk among patients with rheumatoid arthritis.

Objective: To determine the performance of the PREVENT equation for estimating atherosclerotic cardiovascular disease (ASCVD) risk in rheumatoid arthritis (RA) compared to the American College of Cardiology/American Heart Association ASCVD risk estimator, i.e., the pooled cohort equation (PCE). Methods: We studied patients from an RA longitudinal registry with annual blood samples. After identifying the earliest sample for each patient (index date), we excluded patients with prevalent cardiovascular disease (CVD) or on statin/lipid lowering therapy through the index date. Medical records were reviewed for ASCVD events including coronary heart disease and stroke up to 10 years after index date. PREVENT-ASCVD and PCE were calculated, and patients were categorized into low, borderline, intermediate, and high ASCVD risk categories. Discrimination and calibration of PREVENT and PCE for ASCVD events were determined. Results: We studied 267 RA patients mean age 57 years, 83 % female; 27 patients (10.1 %) had an ASCVD event. The AUC for PREVENT was 0.71 and 0.70 for PCE. Compared to observed ASCVD events, PREVENT underestimated ASCVD risk, to a larger degree compared to PCE; PREVENT either classified patients in the same risk category or reclassified to lower risk categories compared to PCE; 28 % of patients would have a definitive recommendation for statin initiation with PCE vs 14 % with PREVENT. Conclusion: Among a cohort of RA patients, PREVENT and PCE underestimated 10-year ASCVD risk compared to observed ASCVD events, with overall lower risk estimates in PREVENT compared to PCE. Adoption of PREVENT may lead to less aggressive primary CV prevention in RA.