Predictive accuracy of the SCORE risk function for cardiovascular disease in clinical practice: a prospective evaluation of 44 649 Austrian men and women

Abstract

Background In 2003, a new risk function for cardiovascular risk in clinical practice was developed by the SCORE project group. The aim of this paper was to evaluate the predictive accuracy of the SCORE in a large Austrian population. Design A prospective, multicentre, longitudinal linkage project. Methods Using the ‘SCORE risk function for low-risk regions', we calculated the risk of death from cardiovascular and coronary heart disease events over a 10-year period for 44 649 participants aged 40-65 years in the Vorarlberg Health Monitoring and Promotion Programme (VHM&PP). The predicted risks were compared with the 95% confidence intervals (CI) of the observed events. Results We observed a total of 487 deaths (1.1%; 95% CI 1.0-1.2) for all cardiovascular disease within 10 years, 371 (1.8%; 95% CI 1.6-2.0) in men and 116 (0.5%; 95% CI 0.4-0.6) in women. The SCORE function overestimated cardiovascular mortality and predicted 666 (1.5%) events, 444 (2.2%) in men and 222 (0.9%) in women. Receiver operating characteristics analyses revealed area under the curve values of 0.76 (95% CI 0.74-0.79) for men and 0.78 (95% CI 0.74-0.82) for women. Multivariable analyses showed that obesity (in men only) increased levels of glucose, gamma-glutamyl transferase, triglycerides (in women only), and blue-collar job status (in women only) significantly contributed to the SCORE as additional independent risk factors. Conclusion Although the SCORE overpredicted the mortality pattern in the cohort as a whole, its predictive ability at the individual level still demonstrates a potentially widespread utility in clinical practice.