[In the identification of cardiovascular risk with the SCORE model, could we recommend its calculation interchangeably with total cholesterol or atherogenic index? Concordance between total cholesterol and atherogenic index in the SCORE table]

Abstract

The SCORE table indiscriminately recommends the use of total cholesterol (SCORE-TC) or atherogenic index (SCORE-AI) for calculating cardiovascular (CV) risk. We evaluated reliability and agreement between both methods and the clinical implications for the identification of high CV risk. Observational study (n=8942) in a 40- to 65-year-old population. Spearman's Rho correlation was 0.987 (P<.001), the agreement intraclass correlation coefficient was 0.671 (IC 95% 0.413-0.796; with Bland-Altman's method, the average of the differences between models was 0.74. Kappa index was poor, 0.297 (P<.001) and positive specific agreement was 0.31. Discrepancies fitted individuals with high CV risk with SCORE-TC and not-high with SCORE-AI (4.7%) and 5.8% (n=518) of individuals were classified as high-risk according to SCORE-TC versus 1.1% (n=95) according to SCORE-AI. Poor agreement was found between SCORE-TC and SCORE-IA for identification of high cardiovascular risk individuals.