Table 1.
Baseline characteristics of the study population.
Table 2.
Distribution of 10-year CVD risk as estimated for individual persons in the study population using competing risk regression stratified by age and gender.
Table 3.
Concordance between the risks predicted by the different risk scores and those estimated by competing risk regression in CPRD.
Table 4.
Differences in individual risk predictions between risk scores and risks estimated using competing risk regression stratified by risk factors (selected results for larger differences).
Figure 1.
Absolute differences in individual 10-year CVD risk prediction between the Framingham, ASSIGN and QRISK2 risk scores and the individual risks estimated in CPRD using competing risk regression stratified by vigintiles of predicted risk.
X-axis: Vigintiles of predicted risk. Y-axis: Percentage of persons.
Table 5.
Consistency in risk predictions with risk scores or competing risk regression in predicting high and low CVD risk.
Table 6.
10-year CVD risk at index date for persons who developed CVD after the index date (cases).