Skip to main content

Table 5 Predictive accuracy of the biomarkers for nonrecovery

From: Derivation and validation of plasma endostatin for predicting renal recovery from acute kidney injury: a prospective validation study

 

Cutoff value

Sensitivity

Specificity

PPV

NPV

Endostatin (ng/ml)

63.7

88.6 (69.8–97.6)

53.7 (40.7–67.5)

52.3 (39.2–66.8)

88.5 (39.2–66.8)

NGAL (ng/ml)

116.2

61.5 (40.6–79.8)

52.7 (38.7–67.5)

41.7 (35.2–59.2)

67.6 (50.1–82.6)

Cystatin C (mg/dl)

4.87

65.4 (44.3–82.8)

50.0 (38.6–64.5)

43.6 (30.8–60.4)

71.0 (52.0–85.8)

Endostatin–clinical risk prediction model

0.279

94.6 (76.8–99.1)

72.7 (57.2–85.0)

66.7 (50.3–81.4)

96.2 (82.3–100.0)

NGAL–clinical risk prediction model

0.266

89.5 (66.9–98.7)

62.5 (43.7–78.9)

58.6 (40.9–76.5)

90.9 (70.8–98.9)

Cystatin C–clinical risk prediction model

0.286

88.4 (66.9–98.7)

57.6 (40.2–74.5)

54.8 (38.8–72.7)

90.2 (69.6–98.3)

  1. Data presented as percentage (95% confidence interval)
  2. NGAL neutrophil gelatinase-associated lipocalin, NPV negative predictive value, PPV positive predictive value
Back to article page

Critical Care

ISSN: 1364-8535