Evaluation of cystatin C-derived glomerular filtration rate equations in Chinese population

ABTRACT Glomerular filtration rate (GFR) has become the best indicator for assessing renal function. This study aims to validate the existing cystatin C (CysC)-derived estimated glomerular filtration rate (eGFR) equations in Chinese patients to explore whether the reported CysC-derived eGFR formulas could be applied to the Chinese population. This study validated the equations in a population totaling 1816 inpatients. We calculated eGFR by different CysC-derived equations, then compared with the mGFR. Equation performance was assessed by bias (mean difference between mGFR and eGFR), precision (inter-quartile range of difference) and accuracy (mainly, accuracy within 30% [P-30]). All equations expressed poor performance in dialysis patients (n = 345), and the performance for non-dialysis patients (n = 1471) were significantly greater than that in dialysis patients. Feng and Pei equations had higher P-30 (50.82% and 49.73%, respectively) than the widely used CKD-EPICysC (41.10%) and MacIsaac equations (48.23%), and the distribution of eGFR values is more similar to the distribution of mGFR in non-dialysis patients. Similar trends showed in mGFR, sex, age, and BMI subgroups. However, no equation met the guideline standard of P-30 >= 75%. Our results suggest that the published CysC-based eGFR equations are not suitable for dialysis patients, and the accuracy of equations for non-dialysis patients significantly better than the dialysis ones. Moreover, Feng and Pei equation showed better performance in non-dialysis patients.