A comparison of Framingham and SCORE-based cardiovascular risk estimates in participants of the German National Health Interview and Examination Survey 1998

Abstract

Overestimation of risk by Framingham risk functions not only in southern but also in northern European populations including Germany, has led to the development of the SCORE risk estimation model. Data of the German National Health Interview and Examination Survey 1998 was used to determine whether SCORE leads to lower estimates of the 10-year absolute risk of fatal cardiovascular disease and fatal coronary heart disease than a Framingham model. Predicted numbers of events were compared with approximations based on national mortality statistics. Inclusion criteria followed the recommendations for the use of SCORE: age 30 to 69 years, no previous history of cardiovascular disease and no markedly raised levels of single risk factors (leaving 1811 men and 1955 women for analysis). The SCORE model for high-risk regions (SCORE-HIGH, which is recommended for Germany pending calibration with national data) predicted the highest number of events, followed by the estimations with mortality statistics, the Framingham model and SCORE-LOW (87 fatal cardiovascular disease events versus 77, 62 and 47; fatal coronary heart disease events 62 versus 46, 46 and 30). Agreement on high-risk status, defined as the 10-year risk of fatal cardiovascular disease of 5% or higher now or if extrapolated to age 60, was moderate for both men and women (kappa 0.52 and 0.42 for Framingham and SCORE-HIGH). Our results suggest that SCORE-HIGH may overestimate absolute risk of fatal coronary heart disease and cardiovascular disease in Germany and may need calibration. Furthermore, the limitations of current risk prediction tools emphasize the ongoing need for comprehensive, high-quality and timely European cohort data.