肾小球滤过率估算公式在老年人群中的应用

    Application of the formula for estimating glomerular filtration rate in the elderly population

    • 摘要: 目的 评估肾小球滤过率对老年人群的重要性,探究需要监测肾小球滤过率的老年人群的临床特征,并验证现行常用的估算肾小球滤过率(estimated glomerular filtration rate,eGFR)公式的性能。方法 收集2012年12月1日至2021年12月31日在华中科技大学同济医学院附属同济医院进行肾动态显像99m锝-二乙三胺五乙酸测量肾小球滤过率(measured glomerular filtration rate,mGFR)检查且年龄≥60岁的313例患者,分析老年患者的特征,分析常用的eGFR公式与mGFR的一致性、精度与准确度。用布兰德-奥尔特曼散点图比较各公式计算得出的eGFR值与mGFR值之间的偏差,以eGFR在mGFR(1 ±10%)内的比例(P10)、(1 ±30%)内的比例(P30)、平均百分比误差(mean percent error,MPE)、平均绝对百分比误差(mean absolute percent error,MAPE)评估各公式的准确度。采用t检验比较各公式间的偏差,McNemar检验比较各公式P10和P30间的差异。结果 本研究的老年人群年龄为68.3(63.0,72.0)岁。现行常用的eGFR公式均不同程度地高估mGFR,使用残差(eGFR-mGFR)的四分位间距(inter quartile range,IQR)、MPE、MAPE、均方根误差(root mean square error,RMSE)评估公式eGFR的精度,科克罗夫特-高尔特(Cockroft-Gault)公式和欧洲肾功能联盟(European kidney function consortium,EKFC)公式均表现了较高的精度,其中Cockroft-Gault公式的MPE和IQR最低,EKFC公式的MAPE和RMSE最低。使用绝对百分比误差在10%和30%以内的患者百分比作为评估eGFR公式准确度性能的指标,EKFC公式和柏林倡议研究(Berlin initiative study,BIS1)公式均表现出较高的准确度,其中EKFC公式的P10最高(26.8%),BIS1公式的P30最高(67.4%)。中国改良简化肾病膳食改良实验(Chinese modified modification of diet in renal disease,C-MDRD)公式在精度及准确度中均表现最差。结论 现行常用的eGFR公式均不同程度地高估老年人mGFR,综合来看,EKFC公式在精度和准确度中都有着较好的表现,且EKFC公式的准确度P30高于Cockroft-Gault公式,而C-MDRD公式在精度及准确度中均表现最差。

       

      Abstract: Objective To accurately assess the importance of glomerular filtration rate in the elderly population, explore the clinical characteristics of the elderly population who need to monitor glomerular filtration rate, and verify the performance of the currently commonly used estimated glomerular filtration rate (eGFR) formula.Methods A total of 313 patients aged 60 years and older who underwent 99mTc-DTPA renal dynamic imaging for measured glomerular filtration rate (mGFR) in Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology from December 1, 2012 to December 31, 2021 were enrolled. The characteristics of elderly patients were analyzed. The consistency, precision and accuracy of commonly used eGFR formulas and mGFR were analyzed. Bland-Altman plots were used to compare the deviations between eGFR values calculated by each formula and mGFR values. The accuracy of each formula was evaluated by the proportion of eGFR within mGFR (1 ±10%) (P10), the proportion within (1 ±30%) (P30), mean percent error (MPE), and mean absolute percent error (MAPE). The deviations between the formulas were compared using the t-test, and the differences in P10 and P30 between the formulas were compared using the McNemar test.Results Results The age of the elderly population in this study was 68.3(63.0,72.0) years. All the commonly used eGFR formulas overestimated mGFR to varying degrees. The interquartile range (IQR), MPE, MAPE, and root mean square error (RMSE) of the residual (eGFR-mGFR) were used to evaluate the accuracy of the formula eGFR. Both the Cockroft-Gault formula and the EKFC formula showed higher accuracy, among which the Cockroft-Gault formula had the lowest MPE and IQR, and the EKFC formula had the lowest MAPE and RMSE. The percentage of patients with absolute percentage errors within 10% and 30% was used as an indicator to evaluate the accuracy performance of the eGFR formula. Both the EKFC formula and the BIS1 formula showed higher accuracy, among which the EKFC formula had the highest P10 (26.8%), and the BIS1 formula had the highest P30 (67.4%). The C-MDRD formula performed the worst in both accuracy and precision.Conclusion All commonly used eGFR formulas overestimate mGFR in the elderly to varying degrees. In general, the EKFC formula has better performance in accuracy and precision, and the accuracy P30 of the EKFC formula is higher than that of the Cockroft-Gault formula, while the C-MDRD formula performs the worst in both accuracy and precision.

       

    /

    返回文章
    返回