三个估算的肾小球滤过率计算公式在2型糖尿病患者中的应用

Application of three equations for glomerular filtration rate in patients with type 2 diabetes mellitus

目的探讨改良MDRD方程、基于血清CysC方程以及血清CysC与SCr联合方程等三个全国eGFR协作组推荐方程对本地区2型糖尿病患者肾功能评估的适用性,为临床选择提供依据。方法选取我院2014-2015年确诊的2型糖尿病患者76例,分别测定患者血清CysC与SCr浓度,以及2 d内测定Ccr。分别以改良MDRD方程、基于血清CysC方程以及血清CysC与SCr联合方程计算e GFR,比较与Ccr的偏差、精密度;用Spearman相关分析评价与Ccr的相关性;比较各方程e GFR的15%和25%准确性,及以Ccr=60 m L/(min·1.73 m2)为界值时的15%准确性;比较各方程对肾功能评价的一致性。结果 76例2型糖尿病患者Ccr为77.5~111.5 m L/(min·1.73 m2),平均97.5 m L/(min·1.73 m2)。Spearman相关分析显示各方程e GFR均与Ccr显著相关(r值在0.873~0.953之间,P值均<0.01)。联合方程最佳,CysC方程次之。CysC方程在偏差和精密度方面略好于联合方程,改良MDRD方程最差。三个方程15%和25%准确度比较差异均有显著统计学意义(P<0.01),两两比较,联合方程和CysC方程的15%和25%准确度比较差异无统计学意义(P>0.0125),均优于改良MDRD方程(P均<0.012 5)。对各方程的15%准确度分组比较差异亦有统计学意义(P<0.05),两两比较,Ccr<60 m L/(min·1.73 m2)组,CysC方程高于MDRD方程(P<0.015),其余方程之间比较差异均无统计学意义(P均>0.0125);Ccr≥60 m L/(min·1.73 m2)组,联合方程和CysC方程之间比较差异无统计学意义(P>0.012 5),均明显高于改良MDRD方程(P均<0.012 5)。联合方程评价肾功能的一致性最佳(Kappa值为0.808),高于单CysC方程(Kappa值为0.549)和MDRD方程(Kappa值为0.475)。结论 CysC联合SCr方程较之单纯基于SCr或CysC的方程更适合用来评价糖尿病患者的肾功能。但如果被评价者存在影响SCr的因素,或是短期内SCr很不稳定,这时选用单CysC方程可能更好。

Objective To investigate the applicability of improved modification of diet in renal disease(MDRD) equation, the serum cystatin C(CysC)-based equation and the combined serum CysC and creatinine(Scr)based equation, recommended by the national estimation of glomerular filtration rate(e GFR) cooperative group, on the assessment of renal function in patients with type 2 diabetes mellitus(T2DM), and to provide a basis for clinical choice.Methods A total of 76 patients with T2 DM who diagnosed in our hospital from 2014 to 2015 were selected. The serum concentration of CysC and Scr were measured, and the ednogenous creatinine cleanrance(Ccr) was detected in two days.The e GFR were calculated by MDRD equation, CysC-based equation and combined CysC and Scr based equation. The correlation, bias and precision of e GFRs were compared with Ccr by Spearman correlation analysis. The accuracy within15%, 25% and the accuracy of 15% with Ccr=60 m L/(min·1.73 m2) as critical value were compared between the each equation. The consistency of the evaluation of renal function by each equation was compared. Results The Ccr of 76 patients with T2 DM was 77.5~111.5 m L/(min·1.73 m2), with the mean of 97.5 m L/(min·1.73 m2). Spearman correlation analysis showed that the e GFRs equation was significantly correlated with Ccr(r=0.873~0.953, P<0.01). The combined Scr and CysC equation was the best, followed by CysC-based equation. The CysC-based equation was slightly better than the combined equation in terms of deviation and precision, and the improved MDRD equation was the worst. There were statistically significant differences in accuracy of 15% and 25% between the three equations(P<0.01). There was no significant difference in the accuracy of 15% and 25% between the combined equation and CysC-based equation(P>0.012 5),but they were significantly better than improved MDRD equation(all P<0.012 5). There were statistically significant differences in the 15% accuracy in Ccr<60 m L/(min·1.73 m2) group and Ccr≥60 m L/(min·1.73 m2) group(P<0.05). The accuracy in CysC-based equation was higher than that in imprioved MDRD equation in Ccr<60 m L/(min·1.73 m2) group(P<0.015), and the difference was not statistically significant between the other equations(P>0.012 5). There was no signif-icant difference in the values between the combined equation and the CysC-based equation in Ccr≥60 mL /(min·1.73 m2)group(P>0.012 5), but they were significantly higher than that in improved MDRD equation(all P<0.012 5). The consistency of combined equation for estimating renal function was the best(Kappa=0.808), which was significantly higher than that of CysC-based equation(Kappa=0.549) and improved MDRD equation(Kappa=0.475). Conclusion Compared with improved MDRD equation or CysC-based equation, the combined CysC and SCr equation is more suitable for the evaluation of renal function in patients with diabetes. But if there are factors that affect SCr or SCr is very unstable in the short term, the choice of CysC-based equation may be better.