Performance of Creatinine-cystatin C Equation for Detection of Hyperfiltration in Chinese Diabetic Patients

赵芳雅,张磊,陈海冰,包玉倩,贾伟平
DOI: https://doi.org/10.11969/j.issn.1673-548X.2015.04.022
2015-01-01
Abstract:Objective Diabetic nephropathy affects 25%-40% of patients with diabetes , which is the common cause of end -stage renal failure .Early detection can ensure timely intervention and improvement of treatment outcome .Since GFR is known to be affect-ed by hyperglycemia, the aim of this study is to compare the performance of GFR estimations calculated by MDRD , cystatin C, and creati-nine-cystatin C equation in Chinese diabetic patients according to glycemic status .Methods The performance of the equations , as com-pared with measured 99mTc-DTPA glomerular filtration rate (mGFR) was analyzed in 2155 diabetic patients.Bias, precision, and accu-racy were compared according to HbA1c square.Receiver-operating characteristic (ROC) curve method was used to evaluate the diag-nostic value of the three equations for detection of moderate renal failure and hyperfiltration .Results The bias of MDRD , cystatin C and creatinine-cystatin C equation was 18.98, 28.74 and 18.04, respectively.The bias was the smallest for creatinine -cystatin C equation ( P<0.05) versus both other two equations .The accuracy within ±15%and 50%of creatinine-cystatin C was significantly higher than that of MDRD and cystatin C equation (P<0.05).For detection of hyperfiltration , the CDK-EPI equation showed the best performance with the greatest combination of sensitivity and specificity (AUC=0.862, P<0.05).Conclusion The creatinine-cystatin C equation appears to be a more generalizable and accurate equation for estimating GFR in comparison with the MDRD and cystatin C equation and could be used as a screening tool for hyperfiltration in diabetes .
What problem does this paper attempt to address?