Background Carotid artery plaque as identified by ultrasound represents very high risk atherosclerosis.
Objectives We determined the performance of cardiovascular disease (CVD) risk equations including the Systematic COronary Risk Evaluation (SCORE) and Framingham score in predicting plaque presence in patients with rheumatoid arthritis (RA).
Methods A cohort of 330 cases was assembled amongst 451 Spanish patients that underwent CVD risk screening and carotid ultrasound. Applied exclusion criteria were established CVD, diabetes and moderate or severe chronic kidney disease. The findings were validated in 90 black and 97 white African RA patients.
Results Carotid plaque was present in 162 (49.1%) of the Spanish patients. The SCORE and Framingham score were each strongly associated with plaque (P <0.0001). In predicting plaque presence, the area under the curve (AUC) (SE) of the receiver operator characteristic (ROC) curve for the Framingham score was larger than for the SCORE (0.799 (0.024) versus 0.747 (0.027), P =0.003). The optimal cut-off value and corresponding sensitivity and specificity for the Framingham score and SCORE were 11.0, 64% and 81% and 0.5, 86% and 58%, respectively. Based on optimal cut-off values, a high Framingham score but not SCORE was associated with carotid plaque independent of age, sex, erythrocyte sedimentation rate and C-reactive protein concentrations. Whereas a conventional Framingham score value of ≥20 correctly classified only 25% as being at high CVD risk, this proportion increased to 64% in those with a Framingham score of >11; the percentage of patients without plaque incorrectly classified as being at high CVD risk increased from 17% in those with a Framingham score or ≥20 to only 23% in those with a Framingham score of >11.0. External validation produced similar results in white but not black Africans with RA.
Conclusions In this study, the Framingham score outperformed the SCORE in predicting plaque presence in RA. Upon CVD risk stratification, a Framingham score of >11.0 comprises a measure that can reclassify about 40% of white RA patients without established CVD, chronic kidney disease or/and diabetes more accurately. Delineation of effective population specific CVD risk assessment strategies is needed in black African RA patients.
Disclosure of Interest None declared