Kidney & Blood Pressure Research (Jun 2019)

Revised Equations to Estimate Glomerular Filtration Rate from Serum Creatinine and Cystatin C in China

  • Min Yang,
  • Yonghua Zou,
  • Tong Lu,
  • Yule Nan,
  • Jianying Niu,
  • Xin Du,
  • Yong Gu

DOI
https://doi.org/10.1159/000500460

Abstract

Read online

Aim: Our previous study demonstrated that the cystatin C-based chronic kidney disease (CKD)-EPI equation and combined by serum creatinine (CKD-EPIscr-cys) had better capability to accurately evaluate glomerular filtration rate in the CKD participants. Considering that the accuracy of estimated glomerular filtration rate (eGFR) remains less ideally, it is essential to modify the equation by including the Chinese eGFR racial factor in order to improve its performance. Methods: Two prospective cohorts were enrolled in 2 medical centers. New equations were developed in 529 participants and validated in 313 participants. Reference glomerular filtration rate (rGFR) was taken by 99mTc-DTPA renal dynamic imaging method (Gates method). The primary outcomes of this study were bias, precision (interquartile range of difference [IQR]), and accuracy (the proportion of eGFR within 30% of rGFR [P30] and root mean square error [RMSE]) of eGFR versus rGFR. Results: In a development data set, Chinese coefficients for CKD-EPIscr (C-CKD-EPIscr), CKD-EPIcys (C-CKD-EPIcys), and CKD-EPIscr-cys (C-CKD-EPIscr-cys) were 0.871, 0.879, and 0.891, respectively. In a validation data set, C-CKD-EPIcys was the most accurate with highest P30 value (62.3%), relative lowest IQR (15.45), and RMSE (0.80) among 6 equations, though the bias of C-CKD-EPIcys was not better than CKD-EPIcys. C-CKD-EPIscr and C-CKD-EPIscr-cys equations were improved in bias (p < 0.001), ­precision, and accuracy (p = 0.004 and <0.001 for P30) compared with CKD-EPIscr and CKD-EPIscr-cys. Conclusion: C-CKD-EPIcys was the most accurate with the highest P30 value, relative lowest IQR, and RMSE among 6 equations. C-CKD-EPIscr and C-CKD-EPIscr-cys equations were improved in bias, precision, and accuracy. Other external validation of these equations is needed.

Keywords