Table 9.
NRI and IDI for Final CTA Models Versus Model Including UKPDS and log CAC Score Alone
| Total Cohort (N=630, 41 Events) | Patients With Coronary Plaque (N=500, 41 Events) | |
|---|---|---|
| Categorical NRI | ||
| Patients with events correctly reclassified to higher risk | 11 | 22 |
| Patients with events incorrectly reclassified to lower risk | 5 | 2 |
| Event net reclassification improvement | 14.6% | 48.8% |
| Patients without events correctly reclassified to lower risk | 102 | 54 |
| Patients without events incorrectly reclassified to higher risk | 45 | 150 |
| Nonevent net reclassification improvement | 9.6% | −20.9% |
| Overall net reclassification improvement | 0.24 | 0.28 |
| Continuous NRI | ||
| Patients with events correctly reclassified to higher risk | 26 | 33 |
| Patients with events incorrectly reclassified to lower risk | 15 | 8 |
| Event net reclassification improvement | 26.8% | 61.0% |
| Patients without events correctly reclassified to lower risk | 402 | 249 |
| Patients without events incorrectly reclassified to higher risk | 187 | 210 |
| Nonevent net reclassification improvement | 36.5% | 8.5% |
| Overall net reclassification improvement | 0.632 | 0.70 |
| Absolute IDI | 0.037 | 0.025 |
| Relative IDI | 64.7% | 30.3% |
CAC indicates coronary artery calcium; CTA, computed tomography angiography; UKPDS, United Kingdom Prospective Diabetes Study; NRI, net Reclassification Improvement; IDI, integrated Discrimination Improvement .