2型糖尿病患者中四种肾小球滤过率评估方法一致性的研究

Consistency Evaluation of Four Estimation Equations for Glomerular Filtration Rate in Type 2 Diabetes Mellitus

目的探讨胱抑素C(Cys-C)评估的肾小球滤过率(GFR)与其他3种方法估测GFR(eGFR)的一致性。方法选择2013年2—11月在中国人民解放军成都军区总医院内分泌科住院的81例成年2型糖尿病患者作为研究对象。通过测定Cys-C、血肌酐和24 h尿肌酐、Cockcroft-Gault方程及MDRD方程4种方法 eGFR,分别记为eGFR1、eGFR2、eGFR3和eGFR4。采用配对t检验、线性相关及回归分析及Bland-Altman作图法统计方法对eGFR2、eGFR3和eGFR4与eGFR1的一致性进行分析。结果 (1)eGFR2和eGFR1的配对差均值为-13.45(P<0.05),相关系数为0.443,95%一致性界限为(-101.2 ml·min-1·1.73 m-2,71.8 ml·min-1·1.73 m-2)。(2)eGFR3和eGFR1的配对差均值为0.86(P>0.05),相关系数为0.588,95%一致性界限为(-71.0 ml·min-1·1.73 m-2,73.0ml·min-1·1.73m-2)。(3)eGFR4和eGFR1的配对差均值为-6.79(P>0.05),相关系数为0.650,95%一致性界限为(-68.1 ml·min-1·1.73 m-2,50.3 ml·min-1·1.73 m-2)。(4)eGFR2、eGFR3和eGFR4落在eGFR1±15%的比例分别为27.2%、34.6%和42.0%;落在eGFR1±30%的比例分别为54.3%、71.6%和75.3%;落在eGFR1±50%的比例分别为74.1%、91.4%和88.9%。结论在2型糖尿病患者中,Cys-C估测的eGFR与24 h尿肌酐估测的eGFR一致性较差,与Cockcroft-Gault方程估测的eGFR固定误差较小,与MDRD方程估测的eGFR准确度最高。

Objective To evaluate consistency of four estimation equations for glomerular filtration rate （GFR） in type 2 diabetes methods, Methods 81 adults of type 2 diabetes mellitus, who were hospitalized in our hospital between February and November, 2013, were selected as the research subjects. GFR （eGFR） was estimated by measuring Cys -C, serum creati- nine and 24 hour urinary creatinine, Cockcroft - Gault equation and MDRD equation and was named eGFR1, eGFR2, eGFR3 and eGFR4 respectively. Pair t test, linear correlation and regression analysis, Bland -Altman analysis chart method were adopt- ed to evaluate consistency. Results （ 1 ） Paired difference average value between eGFR2 and eGFR1 was - 13.45 （ P 〈 0.05）. Correlation index was 0.443. Limit of 95% consistency was - 101.2 ml · min-2 · 1.73 m-2 to 71.8 ml · min-1 · 1.73 m-2. （2） Paired difference average value between eGFR3 and eGFR1 was 0. 86 （ P 〉 0.05 ） . Correlation index was 0. 588. Limit of 95% consistency was -71.0 ml · min-1· 1.73 m-2 to 73.0 ml · min-1 · 1.73 m-2. （3） Paired difference average value between eGFR4 and eGFR1 was -6. 79 （P 〉 0. 05 ） . Correlation index was 0. 650. Limit of 95% con- sistency was -68. 1 ml · min-1 · 1.73 m-2 to 50. 3 ml · min-1 · 1.73 m-2. （4） Ratio of cases in eGFR1 ± 15% of eGFR2, eGFR3 and eGFR4 was respectively 27.2%, 34. 6% and 42. 0%. Ratio of cases in eGFR1 ± 30% was 54. 3%, 71.6% and 75.3%. Ratio of cases in eGFR1 ± 50% was 74. 1%, 91.4% and 88.9%. fiontlusion There is less eonsistency betteen eGFR estimated by cystatin C and eGFR estimated by 24 hours urinary creatinine. There is less bias between eGFR estimated by cystatin C and Cockcroft - Gault equation. The consistency between eGFR estimated by cystatin Cand MDRD equation is better.