Table 3.
Test characteristics of endothelial dysfunction marker-based multivariable logistic regression model to estimate risk of D3 SA-AKI in the entire cohort and across predefined groups based on PERSEVERE-II mortality risk
Training set | Test set | |
---|---|---|
Entire cohort | ||
AUROC | 0.78 | 0.77 |
True positive, n | 52 | 51 |
False negative, n | 51 | 52 |
False positive, n | 29 | 30 |
True negative, n | 165 | 164 |
Sensitivity % | 50.5% | 49.5% |
Specificity % | 85.1% | 84.5% |
False positive rate (%) | 14.9% | 15.5% |
False negative rate | 49.5% | 50.5% |
High and intermediate risk P-II strata | ||
AUROC | 0.88 | 0.85 |
True positive, n | 37 | 35 |
False negative, n | 11 | 13 |
False positive, n | 9 | 9 |
True negative, n | 40 | 40 |
Sensitivity % | 77.1% | 72.9% |
Specificity % | 81.6% | 81.6% |
False positive rate (%) | 18.4% | 18.4% |
False negative rate (%) | 22.9% | 27.1% |
Low-risk P-II Strata | ||
AUROC | 0.77 | 0.73 |
True positive, n | 16 | 17 |
False negative, n | 39 | 38 |
False positive, n | 10 | 14 |
True negative, n | 135 | 131 |
Sensitivity % | 29.1% | 30.9% |
Specificity % | 93.1% | 90.3% |
False positive rate (%) | 6.9% | 9.7% |
False negative rate (%) | 70.9% | 69.1% |
Incorporated D1 SA-AKI, sTM, Angpt-2, Tie-2, and Angpt-2/Tie-2 ratio