Skip to main content
[Preprint]. 2023 Aug 2:2023.08.01.23293525. [Version 1] doi: 10.1101/2023.08.01.23293525

Table 3:

Validation statistic for every 10-year model. Each model was run 10 times on random 80% derivation samples then assessed on the remaining 20%.

Outcome ASCVD MI Stroke HF All-Cause Mortality Fatal ASCVD Fatal MI Fatal Stroke
C-statistic: Derivation 0.753 0.745 0.794 0.852 0.776 0.836 0.748 0.813
C-statistic: Validation 0.758 0.750 0.802 0.850 0.775 0.834 0.752 0.827
C-statistic: Derivation-Validation Difference −0.005 −0.004 −0.007 0.002 0.001 0.002 −0.004 −0.013
C-statistic: Std Dev of Derivation-Validation Difference 0.061 0.036 0.033 0.034 0.089 0.038 0.012 0.006
Brier Score: Derivation 0.059 0.034 0.032 0.034 0.089 0.037 0.012 0.005
Brier Score: Validation 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.000
Brier Score: Derivation-Validation Difference 0.059 0.034 0.032 0.033 0.089 0.037 0.012 0.005
Calibration in the large: Derivation 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.000
Calibration in the large: validation 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.000
Risk: 25th percentile 0.027 0.015 0.009 0.004 0.035 0.008 0.004 0.001
Risk: 50th percentile 0.047 0.026 0.019 0.007 0.073 0.014 0.008 0.003
Risk: 75th percentile 0.083 0.043 0.040 0.015 0.159 0.027 0.016 0.007