Objectives: The objective of this study was to explore the extent of the differences in definitions of composite end points and assess how these differences influence estimates of cardiovascular disease (CVD) burden.
Study Design and Settings: Data from a Dutch cohort study (n = 19,484) was used to calculate 10-year risks according to four CVD risk prediction models: Adult Treatment Panel (ATP) III, Framingham Global Risk Score (FRS), Pooled Cohort Equations (PCE), and SCORE. Health loss was estimated based on the impact of event types included in the corresponding composite end points. Finally, each prediction model was used to estimate the expected CVD burden in high-risk individuals, expressed as Quality-Adjusted Life Years (QALYs) lost.
Results: The definition of the composite end points varied widely across the four models. FRS predicted the highest CVD risks, and the composite end point used in SCORE was associated with the highest health burden. The predicted CVD burden in high-risk individuals was 0.23, 0.74, 0.43, and 0.39 QALYs lost per individual when using ATP, FRS, PCE, and SCORE, respectively.
Conclusion: The investigated CVD risk prediction models showed huge variation in definition of composite end points and associated health burden. Therefore, health consequences related to predicted risks cannot be readily compared across prediction models, and estimates of burden of disease depend crucially on the prediction model used.
- Burden of disease
- Cardiovascular disease
- Composite end point
- Prediction model