Table 2.
Comparison of model discrimination
|
Variables
|
Training cohort, n = 500
|
Validation cohort, n = 62
|
||||
|
Model
|
PM
|
ALBI
|
MELD
|
PM
|
ALBI
|
MELD
|
| AUC (95%CI) | 0.833 (0.792-0.873) | 0.651 (0.598-0.703) | 0.508 (0.436-0.548) | 0.802 (0.684-0.920) | 0.658 (0.536-0.774) | 0.631 (0.499-0.750) |
| Sensitivity, % | 83.100 (118/142) | 43.700 (62/142) | 62.000 (88/142) | 95.500 (21/22) | 72.300 (13/22) | 59.100 (13/22) |
| Specificity, % | 73.500 (263/358) | 80.200 (287/358) | 53.100 (190/358) | 52.500 (21/40) | 57.500 (29/40) | 72.500 (29/40) |
| P value | - | < 0.0011 | < 0.0012 | - | 0.0403 | 0.0484 |
1
Area under receiver operating characteristic curve (AUC) values of the albumin-bilirubin (ALBI) score compared to that of the posthepatectomy liver failure model (PM) in the training cohort.
2
AUC values of the model for end-stage liver disease (MELD) score compared to that of the PM in the training cohort.
3
AUC values of the ALBI score compared to that of the PM in the validation cohort.
4
AUC values of the model for end-stage liver disease score compared to that of the PM in the validation cohort.
CI: Confidence interval.