Table 2 Predictive values, sensitivities and specificities for different study populations for APRI and GPR

From: Systematic review and individual-patient-data meta-analysis of non-invasive fibrosis markers for chronic hepatitis B in Africa

Population, diagnostic targeta

Biomarker

Cut-off

PPV (%)

NPV (%)

Sensitivity (%)

Specificity (%)

Asymptomatic screening, LSM > 12.2 kPa

APRI rule-in

0.65

17.0 (13.9–20.2)

98.4 (98.0–98.8)

56.4 (45.9–66.9)

90.6 (89.5–91.6)

GPR rule-in

0.47

20.1 (16.1–24.3)

98.5 (98.1–99.0)

69.7 (48.0–71.4)

91.9 (90.5–93.3)

APRI Youden’s J

0.54

14.0 (11.9–16.0)

98.8 (98.4–99.2)

68.8 (58.6–78.6)

85.6 (84.3–86.8)

GPR Youden’s J

0.37

13.9 (11.3–16.5)

98.6 (98.1–99.0)

63.4 (51.9–74.5)

86.7 (85.0–88.4)

APRI rule-out

0.36

7.4 (6.7–8.1)

99.0 (98.7–99.4)

80.7 (73.7–87.8)

65.5 (64.0–67.1)

GPR rule-out

0.23

7.8 (6.8–8.8)

98.9 (98.5–99.4)

79.4 (69.5–88.5)

68.2 (66.1–70.3)

Suspected liver disease, LSM > 12.2 kPa

APRI rule-in

0.65

70.5 (65.4–75.6)

77.3 (74.6–79.9)

56.6 (49.9–62.8)

86.2 (83.2–89.1)

GPR rule-in

0.47

74.9 (48.9–93.7)

73.0 (68.4–77.2)

42.5 (32.0–52.5)

90.6 (73.9–98.7)

APRI Youden’s J

0.54

66.0 (61.5–70.3)

79.4 (76.6–82.3)

64.0 (58.0–70.0)

80.7 (77.4–84.0)

GPR Youden’s J

0.37

72.0 (51.7–87.8)

76.8 (72.3–81.0)

55.2 (44.2–64.9)

86.3 (70.0–96.3)

APRI rule-out

0.36

51.9 (49.0–54.7)

83.2 (79.3–87.0)

80.3 (75.1–85.4)

56.7 (52.4–60.6)

GPR rule-out

0.23

49.7 (40.8–63.7)

81.2 (73.5–88.6)

80.2 (73.4–86.6)

51.3 (34.4–73.9)

Asymptomatic screening, LSM > 7.9 kPa

APRI rule-in

0.65

10.7 (8.5–12.9)

97.4 (97.3–97.6)

28.8 (23.8–33.8)

91.8 (90.7–92.9)

GPR rule-in

0.40

12.2 (9.8–14.6)

97.7 (97.5–98.0)

38.5 (32.1–44.8)

90.5 (89.0–92.0)

APRI Youden’s J

0.46

8.0 (7.0–9.0)

97.9 (97.7–98.1)

48.8 (43.7–53.9)

80.9 (79.4–82.3)

GPR Youden’s J

0.23

7.0 (6.3–7.7)

98.3 (98.0–98.6)

63.9 (58.1–69.7)

71.0 (68.8–73.2)

APRI rule-out

0.27

4.7 (4.4–5.0)

98.5 (98.2–98.7)

78.9 (75.0–82.9)

45.4 (43.6–47.1)

GPR rule-out

0.17

5.4 (5.0–5.8)

98.5 (98.2–98.8)

76.5 (71.4–81.1)

54.1 (51.7–56.4)

Suspected liver disease, LSM > 7.9 kPa

APRI rule-in

0.65

75.1 (68.8–80.7)

74.9 (72.9–76.9)

47.6 (42.3–53.2)

90.8 (88.0–93.4)

GPR rule-in

0.40

83.6 (69.3–93.2)

79.6 (74.3–83.5)

59.0 (46.0–68.4)

93.1 (85.5–97.6)

APRI Youden’s J

0.46

66.9 (62.5–71.5)

81.6 (79.2–84.1)

69.0 (64.1–73.7)

80.1 (76.2–83.8)

GPR Youden’s J

0.23

60.2 (48.5–70.6)

83.4 (78.6–88.1)

76.2 (67.7–83.5)

69.9 (53.6–82.8)

APRI rule-out

0.27

44.7 (42.5–47.0)

83.8 (79.4–88.4)

87.9 (84.2–91.4)

36.6 (31.8–41.8)

GPR rule-out

0.17

51.8 (43.8–60.3)

84.4 (78.1–89.8)

82.8 (74.0–90.0)

54.5 (38.4–69.9)

  1. Estimates are reported as posterior means and 95% credible intervals from the Bayesian bivariate random-effects models.
  2. aCirrhosis prevalence was 2.5% among people tested for hepatitis B as part of asymptomatic screening, and 26.5% among people tested for hepatitis B due to suspected liver disease (viz with symptoms, clinical signs, or abnormal liver function tests).