Skip to main content

Table 4 Diagnostic criteria of validated CKD models for various thresholds: sensitivity, specificity and predictive values and proportions of expected to observed cases with 95%-confidence intervals

From: External validation of six clinical models for prediction of chronic kidney disease in a German population

Model

Sensitivity

Specificity

Predictive values

Positive Negative

Expected/Observed proportion (95% -CI)

SCORED

 4

84.9

44.4

13.4

96.7

6.33 (6.09; 6.59)

 5

61.0

72.5

18.4

94.8

3.32 (3.14; 3.51)

 6

30.6

90.3

24.3

92.8

1.26 (1.15; 1.38)

 7

11.4

97.4

13.5

96.6

0.37 (0.32; 0.44)

Modified SCORED

 4

84.4

45.3

13.5

96.6

6.24 (6.00; 6.50)

 5

60.8

73.4

18.8

94.9

3.24 (3.06; 3.42)

 6

29.1

91.1

24.9

92.7

1.17 (1.06; 1.28)

 7

9.1

97.9

31.0

91.4

0.29 (0.24; 0.33)

Kshirsagar

 3

84.2

44.4

13.3

96.5

6.33 (6.09, 6.59)

 4

60.8

72.8

18.5

94.8

3.29 (3.12, 3.48)

 5

29.9

90.7

24.6

92.7

1.22 (1.11; 1.33)

Kwon

 4

84.7

45.2

13.5

96.7

6.26 (6.01; 6.51)

 5

59.7

73.7

18.7

94.8

3.20 (3.02; 3.38)

 6

27.0

91.6

24.5

92.5

1.10 (1.00; 1.21)

 7

6.2

98.6

30.4

91.2

0.21 (0.16; 0.26)

Thakkinstian

 6

82.9

45.7

13.4

96.3

6.19 (5.94, 6.44)

 7

69.9

58.2

14.5

95.0

4.83 (4.61, 5.05)

 8

43.9

81.0

18.9

93.4

2.32 (2.17, 2.48)

 9

41.8

83.8

20.8

93.4

2.02 (1.88, 2.16)

 10

38.7

866

22.6

93.3

1.71 (1.59; 1.85)

 11

28.8

90.8

24.2

92.7

1.19 (1.09; 1.31)

  1. CKD defined as eGFR < 60 m/min/1.73m2 calculated with CKD-Epi equation. CKD prevalence in validation population 9.2%