Skip to main content

Table 4 Diagnostic Performance of Biomarkers Alone and in Combination with Clinical Factors in the Development Cohort (JHU/UTSMC/WVC) and Model 5 in the Validation Cohort (PUMCH)

From: Validation of the novel GLAS algorithm as an aid in the detection of liver fibrosis and cirrhosis based on GP73, LG2m, age, and sex

 

Model

Predictor Variables

N

Events (n)

AUC

AUC 95% CI

SE (%)

SP (%)

SE (SP = 90%)

SP (SE = 90%)

SP (SE = 75%)

Development Cohort

1

GP73

393

260

0.86

(0.82, 0.89)

82.7

64.7

68.8

47.4

78.9

2

LG2m

393

260

0.83

(0.79, 0.87)

81.9

60.9

66.5

32.3

79.7

3

GP73 + Age + Sex

393

260

0.91

(0.89, 0.94)

89.6

73.7

76.2

73.7

91.7

4

LG2m + Age + Sex

393

260

0.88

(0.85, 0.92)

88.5

75.9

61.9

72.9

84.2

5

GP73 + LG2m + Age + Sex (GLAS)

393

260

0.92

(0.90, 0.95)

88.8

75.9

79.2

73.7

94.7

 

Model

Etiology

N

Predicted Events (n)

AUC

AUC 95% CI

SE (%)

SP (%)

SE (SP = 90%)

SP (SE = 90%)

SP (SE = 75%)

Validation Cohort

5 (GLAS)*

All Fibrosis/Cirrhosis

354

248

0.93

(0.90, 0.95)

91.1

80.2

81.0

82.1

93.4

Viral

158

52

0.91

(0.86, 0.96)

88.5

80.2

75.0

81.1

91.5

Non-viral

252

146

0.94

(0.91, 0.97)

93.2

80.2

86.3

84.9

93.4

Unknown Etiology

156

50

0.91

(0.86, 0.96)

88.0

80.2

74.0

73.6

89.6

Chronic Liver Disease

253

147

0.65

(0.58, 0.71)

42.9

80.2

26.5

15.1

39.6

All Liver Disease

501

395

0.82

(0.78, 0.86)

73.2

80.2

61.0

36.8

76.4

  1. AUC, area under the ROC curve; SE, sensitivity; SP, specificity
  2. *Comparison group for all analyses is healthy controls (n = 106)