Table 3.
Statistic | 5% ≤HS<10% (n=33) | 10%≤HS<20% (n=26) | HS ≥20% (n=35) |
| |||
Significant fibrosis assessment (F≥2) | |||
AUC | 0.78 (0.60, 0.90) | 0.79 (0.59, 0.93) | 0.93 (0.79, 0.99) |
Standard error | 0.083 | 0.092 | 0.041 |
Cut-off value | 6.6 | 6.8 | 8.4 |
Sensitivity (%) | 83.3 (58.6, 96.4) | 66.7 (34.9, 90.1) | 76.5 (50.1, 93.2) |
Specificity (%) | 66.7 (38.4, 88.2) | 85.7 (57.2, 98.2) | 100.0 (81.5, 100.0) |
PPV (%) | 75.0 (50.9, 91.3) | 80.0 (44.4, 97.5) | 100.0 (75.3, 100.0) |
NPV (%) | 76.9 (46.2, 95.0) | 75.0 (47.6, 92.7) | 81.8 (59.7, 94.8) |
Comparison of AUCs | |||
5% ≤HS<10% | - | P=0.94 | P=0.11 |
10%≤HS<20% | P=0.94 | - | P=0.16 |
HS ≥20% | P=0.11 | P=0.16 | - |
Advanced fibrosis assessment (F≥3) | |||
AUC | 0.96 (0.83, 0.99) | 0.94 (0.77, 0.99) | 0.95 (0.82, 0.99) |
Standard error | 0.037 | 0.054 | 0.039 |
Cut-off value | 8.6 | 7.2 | 8.4 |
Sensitivity (%) | 90.0 (55.5, 99.7) | 85.7 (42.1, 99.6) | 85.7 (57.2, 98.2) |
Specificity (%) | 95.7 (78.1, 99.9) | 89.5 (66.9, 98.7) | 95.2 (76.2, 99.9) |
PPV (%) | 90.0 (55.5, 99.7) | 75.0 (34.9, 96.8) | 92.3 (64.0, 99.8) |
NPV (%) | 95.7 (78.1, 99.9) | 94.4 (72.7, 99.9) | 90.9 (70.8, 98.9) |
Comparison of AUCs | |||
5% ≤HS<10% | - | P=0.76 | P=0.85 |
10%≤HS<20% | P=0.76 | - | P=0.88 |
HS ≥20% | P=0.85 | P=0.88 | - |
Cirrhosis assessment (F=4) | |||
AUC | 1.00 (0.89, 1.00) | 0.98 (0.79, 1.00) | 0.90 (0.75, 0.98) |
Standard error | 0.000 | 0.029 | 0.076 |
Cut-off value | 10.7 | 10.0 | 10.9 |
Sensitivity (%) | 100.0 (59.0, 100.0) | 100.0 (15.8, 100.0) | 80.0 (28.4, 99.5) |
Specificity (%) | 100.0 (86.8, 100.0) | 95.8 (78.9, 99.9) | 86.7 (69.3, 96.2) |
PPV (%) | 100.0 (59.0, 100.0) | 66.7 (9.4, 99.2) | 50.0 (15.7, 84.3) |
NPV (%) | 100.0 (86.8, 100.0) | 100.0 (85.2, 100.0) | 96.3 (81.0, 99.9) |
Comparison of AUCs | |||
5% ≤HS<10% | - | P=0.49 | P=0.19 |
10%≤HS<20% | P=0.49 | - | P=0.33 |
HS ≥20% | P=0.19 | P=0.33 | - |
Data in parentheses were 95% confidence interval. AUC=area under the ROC curve. HS=hepatic steatosis. PPV=positive predictive value. NPV=negative predictive value.