Skip to main content
. 2017 Sep 29;17:650. doi: 10.1186/s12879-017-2728-7

Table 3.

Predictive accuracy of ACI and other index for diagnosing significant fibrosis in total patients

Index Patients AUC (95% CI) Cut-off Sensiti -vity (%) Speci-ficit (%) PPV (%) NPV (%) LR
ACI ALT <2 × ULN
(n = 460)
0.813
(95% CI: 0.772, 0.854)
≥0.2 88.0 50.9 51.3 90.8 1.6
≥0.3 76.1 74.4 64.0 84.0 2.97
≥0.5 50.9 90.0 66.7 74.9 4.80
Normal ALT
(n = 196)
0.824
(95% CI: 0.763, 0.884)
≥0.2 81.8 63.7 54.0 87.5 2.25
≥0.3 65.2 86.3 66.1 84.0 4.75
≥0.5 40.0 94.4 75.7 75.9 6.98
APRI ALT <2 × ULN
(n = 460)
0.742
(95%CI: 0.695, 0.789)
≥0.5 70.0 63.6 53.9 78.8 2.07
≥1.0 26.4 94.4 73.8 68.9 4.71
Normal ALT (n = 196) 0.780
(95% CI: 0.710, 0.849)
≥0.5 55.2 85.8 58.9 78.3 3.90
≥1.0 16.4 98.4 84.6 69.1 10.43
FIB-4 ALT <2 ULN
(n = 460)
0.741
(95% CI: 0.693, 0.788)
≥1.0 89.6 34.2 43.9 66.4 1.42
≥1.5 17.7 96.4 75.6 67.1 4.97
Normal ALT
(n = 196)
0.763
(95% CI: 0.689, 0.830)
≥1.0 90.8 31.8 40.4 86.7 1.33
≥1.5 21.5 96.8 77.8 70.5 6.78
Forns’ index ALT <2 × ULN
(n = 460)
0.766
(95% CI: 0.717, 0.816)
≥6.0 90.1 37.5 44.9 86.8 1.44
≥8.0 54.6 83.5 64.7 75.8 3.3
Normal ALT
(n = 196)
0.762
(95% CI: 0.689, 0.847)
≥6.0 86.54 40.38 41.3 85.1 1.43
≥8.0 50.0 87.5 67.5 78.4 4.00
Hui Model ALT <2 × ULN
(n = 460)
0.736
(95% CI: 0.686, 0.787)
≥0.15 60.1 74.4 58.1 76.5 2.55
≥0.3 38.5 90.9 71.3 71.7 4.25
Normal ALT
(n = 196)
0.745
(95% CI: 0.664, 0.825)
≥0.15 51.8 81.8 58.8 77.4 2.85
≥0.3 33.9 91.8 66.7 72.7 4.15
≥0.3 33.9 91.8 66.7 72.7 4.15