Table 2.
The IDI and NRI were used to assess reclassification performance and improvement in discrimination of our novel prediction model
| IDIa | P Value | NRIb | P value | |
|---|---|---|---|---|
| For derivation cohort | ||||
| Our model vs APA model | 0.138 | < 0.001* | 0.283 | < 0.001* |
| Our model vs RPA model | 0.163 | < 0.001* | 0.283 | < 0.001* |
| Our model vs GPA model | 0.156 | < 0.001* | 0.388 | < 0.001* |
| For validation cohort | ||||
| Our model vs APA model | 0.139 | 0.158 | 0.272 | 0.178 |
| Our model vs RPA mode | 0.235 | < 0.001* | 0.460 | 0.020* |
| Our model vs GPA model | 0.108 | 0.198 | 0.246 | 0.099 |
IDI integrated discrimination improvement index, NRI net reclassification improvement index. *P < 0.05
a,bPositive velue represents better accuracy, negative velue represents worse accuracy