Skip to main content

Table 3 eGFR equation performance in enabling accurate CKD staging (G1, G2, G3a, G3b, G4, G5). (To be placed after ‘Results’ subsection ‘CKD Staging performance by eGFR equations’)

From: Impact of race-independent equations on estimating glomerular filtration rate for the assessment of kidney dysfunction in liver disease

Subgroup

Samples (N)

Equation

Correct CKD Staginga

Error in CKD Staging

Underestimation of CKD severity

Overestimation of CKD severity

Overall

205

eGFRcr(ASR)

98 (47.8%)

64 (31.2%)

43 (21.0%)

eGFRcr(AS)

104 (50.7%)

74 (36.1%)

27 (13.2%)

eGFRcr-cys(ASR)

104 (50.7%)

23 (11.3%)

78 (38.0%)

eGFRcr-cys(AS)

109 (53.2%)

31 (15.1%)

65 (31.7%)

GFRNMR

120 (58.5%)

37 (18.1%)

48 (23.4%)

Preserved liver function

143

eGFRcr(ASR)

78 (54.5%)

26 (18.2%)

39 (27.3%)

eGFRcr(AS)

84 (58.7%)

34 (23.8%)

25 (17.5%)

eGFRcr-cys(ASR)

67 (46.9%)

8 (5.5%)

68 (47.6%)

eGFRcr-cys(AS)

71 (49.7%)

14 (9.7%)

58 (40.6%)

GFRNMR

85 (59.4%)

22 (15.4%)

36 (25.2%)

Reduced liver function

60

eGFRcr(ASR)

20 (33.3%)*

36 (60.0%)

4 (6.7%)

eGFRcr(AS)

20 (33.3%)*

38 (63.3%)

2 (3.4%)

eGFRcr-cys(ASR)

35 (58.3%)

15 (25.0%)

10 (16.7%)

eGFRcr-cys(AS)

36 (60.0%)

17 (28.3%)

7 (11.7%)

GFRNMR

33 (55.0%)

15 (25.0%)

12 (20.0%)

MELD ≤ 15

148

eGFRcr(ASR)

78 (52.7%)

37 (25.0%)

33 (22.3%)

eGFRcr(AS)

80 (54.1%)

46 (31.1%)

22 (14.8%)

eGFRcr-cys(ASR)

73 (49.3%)

12 (8.1%)

63 (42.6%)

eGFRcr-cys(AS)

76 (51.4%)

20 (13.5%)

52 (35.1%)

GFRNMR

87 (58.8%)

26 (17.6%)

35 (23.6%)

MELD > 15

49

eGFRcr(ASR)

18 (36.7%)

21 (42.9%)

10 (20.4%)

eGFRcr(AS)

22 (44.9%)

22 (44.9%)

5 (10.2%)

eGFRcr-cys(ASR)

26 (53.1%)

9 (18.4%)

14 (28.5%)

eGFRcr-cys(AS)

28 (57.1%)

9 (18.4%)

12 (24.5%)

GFRNMR

28 (57.1%)

9 (18.4%)

12 (24.5%)

Ascites not present

163

eGFRcr(ASR)

86 (52.8%)

35 (21.5%)

42 (25.7%)

eGFRcr(AS)

93 (57.1%)

43 (26.4%)

27 (16.5%)

eGFRcr-cys(ASR)

80 (49.1%)

11 (6.7%)

72 (44.2%)

eGFRcr-cys(AS)

84 (51.5%)

18 (11.1%)

61 (37.4%)

GFRNMR

97 (59.5%)

26 (16.0%)

40 (24.5%)

Ascites present

42

eGFRcr(ASR)

12 (28.6%)*

29 (69.0%)

1 (2.4%)

eGFRcr(AS)

11 (26.2%)*

31 (73.8%)

0 (0.0%)

eGFRcr-cys(ASR)

24 (57.1%)

12 (28.6%)

6 (14.3%)

eGFRcr-cys(AS)

25 (59.5%)

13 (31.0%)

4 (9.5%)

GFRNMR

23 (54.8%)

11 (26.2%)

8 (19.0%)

mGFR ≥ 60b

96

eGFRcr(ASR)

56 (58.3%)

21 (21.9%)

19 (19.8%)

eGFRcr(AS)

59 (61.5%)

24 (25.0%)

13 (13.5%)

eGFRcr-cys(ASR)

53 (55.2%)

4 (4.2%)

39 (40.6%)

eGFRcr-cys(AS)

61 (63.5%)

6 (6.2%)

29 (30.3%)

GFRNMR

57 (59.4%)

6 (6.2%)

33 (34.4%)

mGFR < 60b

109

eGFRcr(ASR)

42 (38.5%)*

43 (39.4%)

24 (22.1%)

eGFRcr(AS)

45 (41.3%)*

50 (45.9%)

14 (12.8%)

eGFRcr-cys(ASR)

51 (46.8%)

19 (17.4%)

39 (35.8%)

eGFRcr-cys(AS)

48 (44.0%)

25 (23.0%)

36 (33.0%)

GFRNMR

63 (57.8%)

31 (28.4%)

15 (13.8%)

  1. aCorrect CKD staging marks the number of samples an equation correctly predicted CKD staging (against staging by mGFR), and symbol * indicates statistical significance (any p‐values < 0.05) in pairwise McNemar’s tests of GFRNMR against each CKD-EPI equation; bml/min/1.73 m2. Numbers in parenthesis represent corresponding percentage of total patients within the given subgroup. Additionally, when CKD staging was not correctly predicted, under- and overestimation of CKD severity (corresponding to over- and underestimation of eGFR, respectively) was evaluated. Percentages sum up to 100% over all three columns