Estimating Performance Of CG And Re-expressed MDRD Equations In Patients With Chronic Kidney Disease

Objective The Cockcorft-Gault equation (CG equation) and re-expressed MDRD (Modification of Diet in Renal Disease) equation have been widely used to estimate glomerular filtration rate(GFR) in patients with chronic kidney disease. The objective of this study was to evaluate the performance of these equations in estimating GFR of Chinese patients with CKD.Methods The study investigated 581 patients with CKD, including 22 cases of single kidney. Their gender, age, height and body weight were recorded, and their serum creatinine and 99mTc-DPTA based TGFR were measured. All patients were devided into several kind of groups based on their gender, age or different side GFR. The CG equation, calibrated CG equation, abbreviated MDRD equation(aMDRD equation), re-expressed MDRD equation(rMDRD equation) and modified MDRD equation for Chinese CKD patients(cMDRD equation) were used to estimate GFR of all patients. The estimated GFR and measured GFR (standard GFR) were compared in different patient groups.Results①In 559 cases of CKD patients, the predictive ability of calibrated CG equation was better than other equations.30% accuracy of rMDRD equation(59.75%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.6541) was better than others;②In 371 cases of male CKD patients, the predictive ability of rMDRD equation was better than other equations.30% accuracy of rMDRD equation(59.03%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.6294) was better than others. In 188 cases of female CKD patients, the predictive ability of calibrated CG equation was better than other equations.30% accuracy of aMDRD equation(62.23%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.7018) was better than others;③In 206 cases of male elderly CKD patients, the predictive ability of CG equation was better than other equations.30% accuracy of CG equation(56.80%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.4960) was better than others; In 100 cases of female elderly CKD patients, the predictive ability of calibrated CG equation was better than other equations.30% accuracy of aMDRD equation and rMDRD equation(57%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.6244) was better than others;④In 45 cases of single kidney patients and CKD patients with distinct difference between two sides'GFR(one side is only 20 percent of the other side or below), the predictive ability of calibrated CG equation was better than other equations.30% accuracy of calibrated CG equation and aMDRD equation(60%) was better than others. Linear regression showed R2 value of calibrated CG equation(0.4698) was better than others;⑤In elderly CKD patients over 75 years old, the predictive ability of CG equation was better than other equations.30% accuracy of rMDRD equation(53.40%) was better than others. Linear regression showed R2 value of aMDRD equation and rMDRD equation(0.4993) was better than others.Conclusions①In two kidney CKD patients and different gender and age subgroups, the calibrated CG equation underestimated TGFR.②Generally speaking, in two kidney CKD patients, calibrated CG equation and rMDRD equation performed better than other equations. In male CKD patients, rMDRD equation performed better than other equations. In female CKD patients, calibrated CG equation and rMDRD equation performed better than other equations. In male elderly CKD patients, CG equation and rMDRD equation performed better than other equations. In female elderly CKD patients, the rMDRD equation performed better than other equations. In single kidney patients and CKD patients with distinct difference between two sides' GFR(one side is only 20 percent of the other side or below), the calibrated CG equation performed better than other equations. In elderly CKD patients over 75 years old, the rMDRD equation performed better than other equations.③After calibration of estimating bias, in two kidney group, different gender subgroups and the elderly subgroup, the calibrated CG equation performed better than other equations.