Development and Validation of a Nomogram Based on Noninvasive Liver Reserve and Fibrosis (PALBI and FIB-4) Model to Predict Posthepatectomy Liver Failure Grade B-C in Patients with Hepatocellular Carcinoma

Table 4

Accuracy of the prediction score of the nomogram for evaluating the risk of PHLF grade B/C incidence.

Variable	Training set (n = 383)		Validation set (n = 191)
	Value	95% CI	Value	95% CI
Area under ROC curve	0.832	0.777–0886	0.803	0.723–0.883
Cut-off score	2.109	—	1.66	—
C-index	0.832	—	0.808	—
Sensitivity (%)	0.813	0.690–0.903	0.884	0.698–0.975
Specificity (%)	0.731	0.679–0.778	0.618	0.539–0.692
Positive predictive value (%)	35.5	30.0–54.0	26.7	20.8–65.5
Negative predictive value (%)	95.5	91.6–96.5	97.1	91.1–97.9
Positive likelihood ratio	3.02	2.43–3.76	2.31	1.82–2.94
Negative likelihood ratio	0.25	0.14–0.43	0.18	0.06–0.54