Table 2 Performance of the equations in the development sample using 10-fold cross-validation and temporary validation sample.

From: Development and validation of a new equation to estimate glomerular filtration rate in Argentinian adults

Equations

RMSE

r

P15

P30

Median bias (Q1, Q3)

CC%

Development sample using 10-fold cross-validation (n = 583)

L6

18.4

0.86

58.0

83.2

-2.5 (-13.1, 7.9)

72

G6

18.2

0.86

58.5

82.5

-0.7 (-11.5, 9.6)

73.2

Q6

18.0

0.87

60.7

83.2

-0.3 (-10.1, 9.0)

73.8

MDRD-4

21.8

0.83

44.1

81.3

-8.8 (-18.9, 4.2)

63.8

MDRD-6

20.7

0.84

50.3

82.8

-7.0 (-17.3, 5.0)

65.9

MCQ

21.7

0.82

47.2

77.9

6.9 (-5.7, 17.6)

71.2

CKD-09

18.3

0.87

58.0

84.4

-4.3 (-13.8, 6.0)

73.6

CKD-21

17.9

0.87

61.2

83.7

-0.6 (-10.1, 9.5)

73.9

EKFC

19.7

0.87

50.4

84.9

-7.1 (-18.6, 3.3)

69.8

Temporary validation sample (n = 78)

AE (Q6)

15.0

0.89

60.2

89.7

-1.1 (-12.1, 6.9)

79.5

MDRD-4

21.8

0.88

39.7

83.3

-13.4 (-25.2, -1.4)

61.5

MDRD-6

18.8

0.89

46.2

89.7

-9.8 (-21.2, 0.4)

61.5

MCQ

17.2

0.89

51.3

83.3

7.5 (-1.5, 19.8)

66.7

CKD-09

16.8

0.88

57.7

88.5

-4.6 (-16.2, 2.4)

70.5

CKD-21

15.4

0.89

61.5

85.9

-1.1 (-12.4, 6.4)

74.4

EKFC

18.2

0.88

52.6

87.2

-7.6 (-21.8, 0.7)

70.5

  1. RMSE = root mean squared error; r = Pearson’s correlation coefficient; Q1 = first quartile; Q3 = third quartile; CC%= correct classification percentage; L6 = linear model using 6 predictive variables; G6 = generalized linear model using 6 predictive variables; Q6 = quasi-likelihood model using 6 predictive variables; MDRD-4 and MDRD-6 = Modification of Diet in Renal Disease with 4 and 6 variables; MCQ = Mayo Clinic Quadratic; CKD-09 = Chronic Kidney Disease Epidemiology Collaboration 2009 (CKD-EPI 2009); CKD-21 = Chronic Kidney Disease Epidemiology Collaboration 2021 (CKD-EPI 2021); EKFC = European Kidney Function Consortium; AE = Argentinian Equation.
  2. The values of the new equations with equal or better performance than the previous ones are highlighted in bold.