Table 3. Discrimination parameters of four models for predicting the composite outcome.
Harrell ’s C | 95%CI | NRI | p value | IDI | p value | |||
---|---|---|---|---|---|---|---|---|
Model 1 | Established risk factors | 0.700 | (0.656, 0.744) | Model 1 | ref. | ref. | ref. | ref. |
Model 2 | Model 1 + ABI | 0.703 | (0.658, 0.749) | Model 2 vs. Model 1 | 0.275 | 0.028 | 0.012 | 0.073 |
Model 3 | Model 1 + SPP | 0.721 | (0.677, 0.764) | Model 3 vs. Model 1 | 0.330 | 0.008 | 0.045 | <0.001 |
Model 4 | Model 1 + ABI + SPP | 0.721 | (0.677, 0.765) | Model 4 vs. Model 2 | 0.359 | 0.004 | 0.033 | 0.003 |
Model 4 vs. Model 3 | 0.109 | 0.382 | 0.001 | 0.712 |
We established risk prediction models for the composite event of death and AVEs. Model 1 includes the eight risk factors identified as in Table 2. In Models 2--4, the factors of SPP and/or ABI tertiles are added to Model 1.
Harrell ’s C, Harrell ’s C-statistic; NRI, net reclassification improvement; IDI, integrated discrimination improvement; CI, confidence interval; ref., reference; ABI, ankle-brachial index; SPP, skin perfusion pressure; AVEs, atherosclerotic vascular events.