2型糖尿病患者中不同肾小球滤过率评估方程的临床应用评价

发布时间：2018-03-21 11:57

本文选题：估算肾小球滤过率评估方程　切入点：肌酐　出处：《上海交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：背景与目的:糖尿病肾病是糖尿病患者常见的并发症之一,并且也是终末期肾病最常见的单一原因。故准确评价这类人群的肾功能具有重要的意义。肾小球滤过率(GFR)是衡量肾功能的重要指标,采用99mTc标记的二乙三胺五乙酸(99mTc-DTPA)肾动态显像法测定GFR开展较广泛。但因其价格昂贵,不利于广泛筛查以及基层医院的推广。目前基于普通人群所开发的通用估算肾小球滤过率(eGFR)方程在临床广泛使用,但一些特定人群由于其自身特点的影响,这些eGFR方程在这类人群中的性能亦会与普通人群有所不同。本研究对于不同肾小球滤过率评估方程在2型糖尿病患者中的适用性进行评价,以筛选出适用于中国2型糖尿病患者肾功能评估的计算方程。方法:选取2010至2014年住院的326名2型糖尿病患者。提取信息包括年龄、性别、身高、体重、肌酐、半胱氨酸蛋白酶抑制剂C(Cys C)。99mTc-DTPA测定的GFR(mGFR)作为参考标准。用Cockcroft-Gauh(C-G)方程(eGFR1方程)、肾脏病膳食改良试验-4(MDRD-4)方程(eGFR2方程)、中国方程(eGFR3方程)、慢性肾脏病流行病学协作组(CKD-EPI)肌酐方程(eGFR4方程)、CKD-EPI Cys C方程(eGFR5方程)、CKD-EPI Cys C肌酐联合方程(eGFR6方程)、Feng Cys C方程(eGFR7方程)以及Feng Cys C肌酐联合方程(eGFR8方程)来估测GFR。各eGFR方程与mGFR之间的比较使用配对t检验、Pearson相关性分析。各eGFR方程间的比较使用Bland-Altman分析、偏差、精确度、准确率和受试者工作曲线分析(ROC)。使用Bland-Altman方法评价一致性限度专业值界定为60mL/(min·1.73 m2)。偏差=eGFR与mGFR差值的平均值,精确度=偏差的标准差,准确率=eGFR数值落在mGFR 10%、30%、50%范围的病例百分比。所有统计学分析在SPSS 19.0和MedCalc 12.7中完成。结果:总人群mGFR 60.46±27.20mL/(min·1.73 m2)。八个eGFR方程中只有eGFR5方程[61.29±30.39 mL/(min·1.73 m2),P=0.450]与mGFR无统计学差异,其余七个方程均高估了患者的GFR水平。Bland-Altman分析结果显示八个方程的一致性限度均超过了规定的专业值,其中eGFR6[66.1(-41.1~25.0)mL/(min·1.73 m2)]一致性限度最小。eGFR5方程在八个方程中偏差最小[-0.83mL/(min·1.73 m2)]。在精确度表现上,eGFR4方程[18.35m L/(min·1.73 m2)]、eGFR5方程[19.75m L/(min·1.73 m2)]、eGFR6方程[16.84m L/(min·1.73 m2)]、eGFR7方程[21.67mL/(min·1.73 m2)]性能较好。在准确率表现上,eGFR5方程和eGFR6方程有较高的10%符合率(分别为24.54%和26.99%)、30%符合率(分别为68.71%和65.64%)和50%符合率(分别为87.42%和84.36%)。而在ROC分析中,八个方程中eGFR6方程(0.905)和eGFR4方程(0.891)ROC曲线下面积最大。结论:在中国2型糖尿病患者中eGFR5和eGFR6的诊断效能较高,优于其余六个方程。所有方程的一致性表现均超过了专业界值,故这两个方程在常规应用于临床前仍需进一步修正。
[Abstract]:Background and objective: diabetic nephropathy is one of the common complications of diabetes, and is also the single cause of end-stage renal disease is the most common. It has important significance in renal function evaluation of this kind of people. The glomerular filtration rate (GFR) is an important indicator of renal function, using 99mTc labeled two B three (five amino acetic acid 99mTc-DTPA) renal dynamic imaging method for the determination of GFR carried out widely. But because the price is expensive, is not conducive to the promotion of extensive screening and primary hospital. The general is currently based on the estimated glomerular filtration rate of the general population development equation (eGFR) is widely used in clinic, but the specific population due to its own characteristics, the performance of the eGFR equation in this population will also be different from ordinary people. This study for different GFR estimation equations in applicability in patients with type 2 diabetes were evaluated by sieve Choose the suitable calculation equation on renal function in patients with type 2 diabetes China evaluation. Methods: 326 patients with type 2 diabetes hospitalized from 2010 to 2014. The extraction information including age, gender, height, body weight, serum creatinine, Cystatin C (Cys C.99mTc-DTPA) determination of GFR (mGFR) as the reference standard. Cockcroft-Gauh (C-G) equation (eGFR1 equation), MDRD -4 (MDRD-4) equation (eGFR2 equation), China equation (eGFR3 equation), chronic kidney disease epidemiology collaboration group (CKD-EPI) creatinine equation (eGFR4 equation), the CKD-EPI Cys C equation (eGFR5 equation), CKD-EPI Cys C (creatinine combined equation eGFR6, Feng Cys) equation C equation (eGFR7 equation) and the Feng Cys C equation (eGFR8 equation) were combined to estimate the comparison between GFR. and mGFR by eGFR equation matching t test, Pearson correlation analysis. The comparison between the Bland eGFR equation -Altman analysis, bias, precision, accuracy analysis and receiver operating curve (ROC). The method of Bland-Altman to evaluate the consistency of the limit value is defined as the use of professional 60mL/ (min 1.73 m2). The average deviation of the =eGFR and mGFR difference value, accuracy = deviation standard deviation, the accurate rate of fall in mGFR 10% = eGFR 30%, 50% cases, percentage range. All the statistical analysis in SPSS 19 and MedCalc 12.7 in complete. Results: the total population of mGFR 60.46 + 27.20mL/ (min 1.73 m2). Eight eGFR equations only eGFR5 equation [61.29 + 30.39 mL/ (min 1.73 m2), no significant difference between P= 0.450] and mGFR, the rest the seven equations are overestimated in the level of GFR.Bland-Altman analysis of the patients showed consistency limit of eight equations are more than the required professional values, including eGFR6[66.1 (-41.1~25.0) mL/ (min - 1.73 m2)] consistency limit minimum.EGFR5 equation in the eight equation the deviation Small [-0.83mL/ (min 1.73 m2). The accuracy of performance, the eGFR4 equation [18.35m L/ (min - 1.73 m2)], eGFR5 L/ equation [19.75m (min - 1.73 m2)], eGFR6 L/ equation [16.84m (min - 1.73 m2)], eGFR7 equation [21.67mL/ (min - 1.73 m2)] in good performance. The accuracy of the performance, the eGFR5 equation and the eGFR6 equation has a higher coincidence rate of 10% (24.54% and 26.99%), the coincidence rate of 30% (68.71% and 65.64%) and the coincidence rate of 50% (87.42% and 84.36%). In the ROC analysis, the eGFR6 equation of the eight equation and the eGFR4 equation (0.905) (0.891) the largest area under the ROC curve. Conclusion: in Chinese eGFR5 and eGFR6 in patients with type 2 diabetes and the diagnostic efficiency is higher, better than the other six equations. The consistency of all equations are more than professional values, so these two equations in conventional before clinical application needs further modification.

【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R587.2;R692.9

【参考文献】