External validation of the European and American equations for calculating cardiovascular risk in a Spanish working population.

Abstract Introduction and objective This work aims to externally validate the European and American models for calculating cardiovascular risk in the primary prevention. Methods This is a cross-sectional study of a nation-wide cohort of individuals who are active in the work force. Workers without a medical history cardiovascular disease who attended occupational health check-ups between 2004 and 2007 were included. They were followed-up on until December 2017. Results A total of 244,236 subjects participated. Of them, 24.5% were women and the mean age was 48.10 years (SD 6.26). According to the European SCORE risk chart, the mean risk was 1.70 (SD 1.81) for men and 0.37 (SD 0.53) for women. According to the North American PCE model, the mean risk was 6.98 (SD 5.66) for men and 1.97 (SD 1.96) for women. A total of 1177 events (0.51%) were registered according to the SCORE tool and 2,330 events (1.00%) were registered according to the PCE tool. The Harrell’s C-statistic was 0.746 for SCORE and 0.725 for PCE. Sensitivity and specificity for the SCORE’S 5% cut-off point were 17.59% (95%CI 15.52%–19.87%) and 95.68% (95%CI 95.59%–95.76%). Sensitivity and specificity for the PCE’s 20% cut-off point were 9.06% (95%CI 7.96%–10.29%) and 97.55% (95%CI 97.48%–97.61%), respectively. Conclusions The European SCORE and North American PCE models overestimate the risk in our population but with an acceptable discrimination. SCORE showed better validity indices than the PCE. The population’s risk is continuously changing; therefore, it is important continue updating the equations to include information on current populations.