多种肾小球滤过率的估算方程在活体供肾者肾功能评估中的适用性研究

The evaluation of glomerular filtration rate estimation equations in potential living related kidney donors

目的比较多种肾小球滤过率（GFR）估算方程在亲属活体供肾功能评估中的准确性，找出适合我国人群的GFR估算方程。方法以44名亲属活体供肾者为对象，以^99mTc-二乙三胺五乙酸（DTPA）肾动态显像测定的GFR为参考标准，并以体表面积（BSA）将其标准化（sGFR）。将以Cockcroft—Gault（C—G）方程估算的肌酐清除率（Qcr），C-G方程、改良C—G方程、肾脏疾病饮食调整研究组（MDRD）方程和改良MDRD方程估算的GFR（eGFR），分别与sGFR进行比较，分析其偏差、相关性、准确性和精确性。结果sGFR为（123±24）ml／min，C—G方程估算的Qcr，以及C-G方程、改良C—G方程、MDRD方程和改良MDRD方程估算的GFR分别为（123±27）ml／min、（104±22）ml／min、（156±28）ml／min、（122±19）ml／min和（138±25）ml／min，其偏差值，MDRD方程最小，两种改良方程的偏差较大；配对t检验及相关性分析，C—G方程的Ccr估算值、MDRD方程的估算值与sGFR的差异无统计学意义，改良CG方程和改良MDRD方程的GFR估算值与sGFR之间的相关性较好；MDRD方程的准确性最高，两种改良方程的准确性较差；改良MDRD方程、改良C—G方程的精确性稍高。结论5个估算方程估算的GFR均有不用程度的误差，相对来说MDRD方程的偏差较小，准确性较高，相关性和精确性尚可，但若应用于临床，有必要对其进行适当修正。

Objective To evaluate the application of glomerular filtration rate estimation equations in living related kidney donors. Methods A total of 44 living related donors were studied. GFR （sGFR） measured by ^99mTc-DTPA was used as the reference standard. The predictive capabilities of Cockcroft-Gauh equation （C-G） for creatinine clearance corrected for body surface area （CC-GrCl）, CG-GrCl corrected for GFR （CG-GFR）,modified C G, MI）RD （simplified MI）RD equation） and modified MDRD were evaluated with sGFR in living related kidney donors. Results The sGFR was （123 24） ml/min. The eGFR of CG-GrCl,CC-GFR,modified C-G,MDRD and modified MDRD was （123 ± 27） ,（104±22）, （156 ±28）, （122 ± 19） and （138 ± 25） ml/min respectively. The eGFR of GC-GrCl and MDRD were not significantly different from the sGFR,and the eGFR of modified C G and modified MI）RD were correlated well with the sGFR. MDRD was the least biased and most accurate, while the more precise formulae were modified CG and modified MDRD. Conclusion Compared with the reference standard,all equations were more or less biased. The modified equations offered no advantage. Compared with other equations, MDRD was least biased and most accurate for predicting GFR in living related kidney donor,and could be the recommended GFR estimating model in this population.