[Agreement between REGICOR and SCORE scales in identifying high cardiovascular risk in the Spanish population].

INTRODUCTION The aims of this study were to evaluate the consistency between the SCORE (Systematic Coronary Risk Evaluation) and REGICOR (Registre Gironí del cor) scales in identifying high cardiovascular risk and to describe the characteristics of those individuals for whom scale results were discrepant. METHODS This cross-sectional study involved 8942 subjects aged 40-65 years who had an indication for a complete lipid profile. The agreement between SCORE (for low-risk countries) and Framingham-REGICOR (with a high risk threshold of 10%) scales in classifying patients as high risk was evaluated using the kappa statistic. Subjects for whom there was a discrepancy between classifications were identified and variables associated with this discrepancy were determined by multivariate analysis involving binary logistic regression. RESULTS The REGICOR scale classified 6.7% of subjects (95% confidence interval [CI], 6.2%-7.3%) as high-risk, while SCORE classified 12.5% (95% CI 11.8%-13.2%) as high-risk. Discrepant findings were observed in 10.2% of the total population (8% had a high risk on SCORE but not REGICOR, and 2.2% had a high risk on REGICOR but not SCORE; kappa=0.420; P< .001). The best agreement was observed between SCORE and REGICOR with a high-risk threshold of 8% (kappa=0.463). Multivariate analysis showed that a high risk on SCORE but not REGICOR was associated with lower age, female sex, a high fasting glucose level, and raised diastolic blood pressure, and a high risk on REGICOR but not SCORE, with male sex, smoking, and a low high-density lipoprotein (HDL) cholesterol level. These variables accounted for the extent of the discrepancy in 93.2% of cases. CONCLUSIONS The SCORE and REGICOR (threshold 10%) scales identified different populations as being at a high risk, though the agreement between them was reasonably good. The concurrence of a number of factors (e.g., male sex, low HDL-cholesterol, and smoking) in a subject with a low risk on the SCORE scale should be regarded as increasing the cardiovascular risk.