Table 3. Accuracy and precision of five estimated glomerular filtration rate equations.
| Group | CKD-EPIcr | Asian CKD-EPI | Revised LM | FAS | Korean FAS |
|---|---|---|---|---|---|
| All (N = 1,312) | |||||
| P10, N (%) | 429 (32.7) | 439 (33.5) | 376 (28.7)* | 404 (30.8) | 444 (33.8) |
| P30, N (%) | 1,005 (76.6) | 980 (74.7)* | 994 (75.8) | 997 (76.0) | 995 (75.8) |
| RMSE | 15.6 | 15.6 | 17.9 | 16.3 | 15.8 |
| mGFR ≥ 60 mL·min−1·1.73 m−2 (N = 674) | |||||
| P10, N (%) | 274 (40.7) | 290 (43.0) | 206 (30.6)† | 239 (35.5)* | 272 (40.4) |
| P30, N (%) | 618 (91.7) | 624 (92.6) | 572 (84.9)† | 607 (90.1) | 623 (92.4) |
| RMSE | 18.1 | 17.0 | 22.7 | 19.6 | 18.1 |
| mGFR < 60 mL·min−1·1.73 m−2 (N = 638) | |||||
| P10, N (%) | 155 (24.3) | 149 (23.4) | 170 (26.6) | 165 (25.9) | 172 (27.0) |
| P30, N (%) | 387 (60.7) | 356 (55.8)† | 422 (66.1)† | 390 (61.1) | 372 (58.3) |
| RMSE | 12.5 | 13.9 | 10.7 | 12.0 | 13.0 |
| Age ≥ 70 yr (N = 392) | |||||
| P10, N (%) | 123 (31.4) | 118 (30.1) | 121 (30.9) | 125 (31.9) | 147 (37.5)* |
| P30, N (%) | 288 (73.5) | 275 (70.2)* | 300 (76.5) | 298 (76.0) | 295 (75.3) |
| RMSE | 11.7 | 12.5 | 12.5 | 12.5 | 12.2 |
| Age < 70 yr (N = 920) | |||||
| P10, N (%) | 306 (33.3) | 321 (34.9) | 255 (27.7)† | 279 (30.3)* | 297 (32.3) |
| P30, N (%) | 717 (77.9) | 705 (76.6) | 694 (75.4)* | 699 (76.0) | 700 (76.1) |
| RMSE | 17.0 | 16.8 | 19.7 | 17.7 | 17.1 |
P10 and P30 represent the eGFR within the range of ±10% and ±30% of the mGFR, respectively.
*P<0.05 for McNemar test (CKD-EPIcr equation vs each equation); †P<0.001 for McNemar test (CKD-EPIcr equation vs each equation).
Abbreviations: CKD-EPIcr, creatinine-based Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; FAS, full age spectrum; mGFR, measured glomerular filtration rate; Revised LM, revised Lund-Malmö; RMSE, root mean square error.