Retrospective Cohort Study
Copyright ©The Author(s) 2022.
World J Gastroenterol. Nov 14, 2022; 28(42): 6045-6055
Published online Nov 14, 2022. doi: 10.3748/wjg.v28.i42.6045
Table 1 Comparison of demographic and clinical characteristics of chronic liver disease patients in the training cohort and validation cohort
Validation cohort, n = 132Training cohort, n = 360P value
Gender (male/female, n)90/42260/1000.381
Age (yr)54.84 ± 27.7048.71 ± 13.340.001
Etiology0.097
HBV111278
Others2182
HCC (n)42630.001
HB (g/dL)14.28 ± 10.8613.71 ± 2.160.346
PLT (× 109/L)174.96 ± 83.93170.95 ± 74.690.610
PT (s)13.21 ± 1.6612.84 ± 1.670.030
APTT (s)33.28 ± 7.7332.54 ± 6.140.273
INR1.16 ± 0.161.09 ± 0.160.000
AST (U/L)72.14 ± 78.1986.16 ± 135.810.264
ALT (U/L)106.35 ± 166.60136.73 ± 247.370.192
ALB (g/L)38.03 ± 5.2240.03 ± 5.080.000
TBIL (μmol/L)28.29 ± 31.0519.94 ± 16.990.000
CHE (U/L)6163.11 ± 2647.196748.44 ± 2127.440.015
ALP (U/L)118.60 ± 95.93179.64 ± 721.940.334
GGT (U/L)128.91 ± 201.79116.89 ± 188.750.539
GLO (g/L)28.70 ± 4.8129.02 ± 5.190.534
Cr (μmol/L)65.36 ± 23.2365.68 ± 15.380.860
MELD8.14 ± 3.467.80 ± 2.560.241
ALBI-2.36 ± 0.54-2.60 ± 0.480.000
PTAR0.31 ± 0.070.28 ± 0.070.000
ICGR15 (%)11.55 ± 11.828.16 ± 8.560.000
LSM (kPa)19.54 ± 18.2816.34 ± 16.620.066
Table 2 Multivariate logistic stepwise regression analysis of indocyanine green retention rate at 15 min ≥ 10% in the training cohort
VariableBSEWaldP valueOR (95%CI)
LSM (kPa)1.4720.33819.008< 0.0014.357 (2.248-8.445)
PTAR1.2650.33514.260< 0.0013.544 (1.838-6.835)
Age (yr)0.0470.01215.329< 0.0011.048 (1.024-1.073)
MELD0.2910.07813.844< 0.0011.337 (1.147-1.558)
Constant-7.6001.02255.302< 0.0010.007
Table 3 Multivariate logistic stepwise regression analysis of indocyanine green retention rate at 15 min ≥ 20% in the training cohort
VariableBSEWaldP valueOR (95%CI)
LSM (kPa)1.1380.5204.7780.0293.120 (1.125-8.656)
PTAR1.2600.5215.8250.0163.524 (1.267-9.801)
Age (yr)0.0580.01711.2260.0011.059 (1.024-1.096)
MELD0.3200.09411.6520.0011.377 (1.146-1.655)
Constant-9.750-1.45444.963< 0.0010.001
Table 4 Comparison of the predictive performance of the new constructed models (mLPaM and sLPaM) and other models in the assessment of impaired liver reserve function in the training cohort
AUC (95%CI)Optimal cut-offSensitivity (%)Specificity (%)PPV (%)NPV (%)Accuracy (%)
mLPaM0.855 (0.809-0.901)0.13591.366.436.0997.3570.68
MELD0.752 (0.688-0.817)7.66280.061.431.2593.3354.75
ALBI0.776 (0.717-0.835)-2.55776.367.937.6791.8569.90
PTAR0.728 (0.664-0.791)0.15073.871.842.1190.7972.24
LSM (kPa)0.733 (0.672-0.794)1.5078.867.937.6792.8670.05
sLPaM0.872 (0.823-0.921)0.04696.864.614.8399.6966.53
MELD0.786 (0.687-0.886)9.38071.085.235.4596.2583.74
ALBI0.798 (0.706-0.890)-2.22064.587.439.8195.0184.78
PTAR0.731 (0.644-0.818)0.15080.665.515.3397.7666.59
LSM (kPa)0.706 (0.618-0.795)1.5080.660.612.7997.7661.94
Table 5 Comparison of the predictive performance of the new constructed models (mLPaM and sLPaM) and other models in the assessment of impaired liver reserve function in the prospective validation cohort
AUC (95%CI)Optimal cut-offSensitivity (%)Specificity (%)PPV (%)NPV (%)Accuracy (%)
mLPaM0.869 (0.810-0.929)0.24089.174.460.8593.8678.94
MELD0.729 (0.633-0.824)9.74343.596.593.6559.0167.74
ALBI0.824 (0.749-0.900)-2.31578.376.763.7887.0977.25
PTAR0.672 (0.580-0.765)1.50089.145.330.7093.8654.66
LSM (kPa)0.782 (0.702-0.862)1.50091.365.149.9395.1572.33
sLPaM0.876 (0.812-0.940)0.07392.968.349.9495.1572.33
MELD0.803 (0.701-0.904)9.18764.385.661.5487.0079.98
ALBI0.836 (0.743-0.929)-1.89771.488.567.4490.2784.22
PTAR0.666 (0.566-0.767)1.50092.959.628.4397.9864.50
LSM (kPa)0.743 (0.653-0.833)1.50092.955.825.3797.9860.96