Table 3. Concordance and Reclassification Indices.
| Model | Wolbers C index (95%CI)a | Comparison to reference models | |
|---|---|---|---|
| Wolbers C index increase (95% CI)a,b | 10-y Risk integrated discrimination improvement (95% CI) | ||
| Primary event population: No. of participants/events, 4018/465 | |||
| Age + sex | 0.719 (0.698 to 0.741) | ||
| Reference model: age + sex | |||
| Age + sex + polygenic risk scores | 0.737 (0.716 to 0.759) | 0.018 (0.009 to 0.028) | 0.014 (0.008 to 0.020) |
| Age + sex + protein risk score | 0.747 (0.727 to 0.768) | 0.028 (0.013 to 0.045) | 0.056 (0.037 to 0.074) |
| Age + sex + protein risk score + polygenic risk scores | 0.756 (0.737 to 0.777) | 0.038 (0.021 to 0.056) | 0.066 (0.046 to 0.084) |
| Clinical risk factorsc | 0.736 (0.713 to 0.759) | 0.017 (0.007 to 0.027) | 0.015 (0.008 to 0.021) |
| Reference model: age + sex + polygenic risk scores | |||
| Age + sex + protein risk score + polygenic risk scores | 0.756 (0.737 to 0.777) | 0.019 (0.005 to 0.034) | 0.052 (0.034 to 0.069) |
| Reference model: age + sex + protein risk score | |||
| Age + sex + protein risk score + polygenic risk scores | 0.756 (0.737 to 0.777) | 0.009 (0.003 to 0.015) | 0.010 (0.005 to 0.015) |
| Reference model: clinical risk factors | |||
| Age + sex + protein risk score | 0.747 (0.727 to 0.768) | 0.011 (−0.002 to 0.025) | 0.041 (0.023 to 0.057) |
| Clinical risk factors + polygenic risk scores | 0.749 (0.728 to 0.770) | 0.013 (0.005 to 0.021) | 0.009 (0.004 to 0.015) |
| Clinical risk factors + protein risk score | 0.750 (0.730 to 0.771) | 0.014 (0.002 to 0.028) | 0.045 (0.028 to 0.062) |
| Clinical risk factors + protein risk score + polygenic risk scores | 0.758 (0.738 to 0.778) | 0.022 (0.007 to 0.038) | 0.053 (0.034 to 0.071) |
| Secondary event population | |||
| No. of participants/events, 6307/432 | 2-y Risk integrated discrimination improvement (95% CI) | ||
| Age + sex | 0.570 (0.541 to 0.602) | ||
| Reference model: age + sex | |||
| Age + sex + protein risk score | 0.667 (0.634 to 0.702) | 0.096 (0.062 to 0.128) | 0.015 (0.011 to 0.019) |
| Refitted SMART2 scored | 0.651 (0.617 to 0.689) | 0.081 (0.042 to 0.119) | 0.017 (0.013 to 0.022) |
| Reference model: refitted SMART2 scored | |||
| Age + sex + protein risk score | 0.667 (0.634 to 0.702) | 0.015 (−0.014 to 0.045) | −0.002 (−0.007 to 0.002) |
| Refitted SMART2 scored + protein risk score | 0.678 (0.646 to 0.710) | 0.026 (0.011 to 0.042) | 0.004 (0.002 to 0.007) |
Abbreviation: SMART, secondary manifestations of arterial disease.
a
This table shows Wolbers C indices that account for the competing risk of non–athersclerotic cardiovascular disease death (Harrell C indices are shown in eTable 4 in Supplement 2).
b
C index increase corresponds to difference relative to the reference model.
c
In the primary event population, the clinical risk factors are as follows: age, sex, smoking, body mass index (calculated as weight in kilograms divided by height in meters squared), type 2 diabetes, hypertension treatment, and statin use.
d
Indicates a refitted (to our data) SMART2 risk score using a training set in the secondary event population.