Table 3.
Performance of the models for post-hepatectomy liver failure prediction in training and test cohorts.
| Clin-model | Rad-model | Combined model | ||
|---|---|---|---|---|
| Training cohort | Cut-off value | 0.27 | 0.29 | 0.28 |
| AUC (95% CI) | 0.74 (0.65–0.83) | 0.79 (0.72–0.86) | 0.84 (0.77–0.90) | |
| Sensitivity | 0.70 | 0.70 | 0.78 | |
| Specificity | 0.74 | 0.76 | 0.81 | |
| PPV | 0.49 | 0.51 | 0.59 | |
| NPV | 0.88 | 0.88 | 0.91 | |
| Accuracy | 0.73 | 0.75 | 0.80 | |
| Test cohort | AUC (95% CI) | 0.71 (0.57–0.84) | 0.79 (0.69–0.89) | 0.82 (0.72–0.91) |
| Sensitivity | 0.87 | 0.87 | 0.93 | |
| Specificity | 0.55 | 0.70 | 0.67 | |
| PPV | 0.30 | 0.39 | 0.39 | |
| NPV | 0.95 | 0.96 | 0.98 | |
| Accuracy | 0.61 | 0.73 | 0.72 | |
AUC, area under the receiver operating characteristics curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value.