Unit of Geriatric Pharmacoepidemiology, Italian National Research Center on Aging, Cosenza, Italy.
Unit of Geriatric Medicine, University Campus Biomedico, Rome, Italy.
Geriatr Gerontol Int. 2018 Apr;18(4):607-614. doi: 10.1111/ggi.13225. Epub 2018 Jan 22.
There is uncertainty about which estimated glomerular filtration rate eGFR equation to use in older people with respect to the prediction of prognosis. Our aim was: (i) to compare the discriminative capacity of eGFR estimated by different equations with respect to all-cause mortality; and (ii) to identify the eGFR threshold at which the risk of mortality starts to increase for each equation.
We used data from 828 community-dwelling older adults aged >65 years enrolled in the InCHIANTI study. The outcome measure was all-cause mortality at 9 years. GFR was estimated by five different equations: Chronic Kidney Disease Epidemiological Collaboration (creatinine equation [CKD-EPI ], and creatinine and cystatin C equation [CKD-EPI ]), Berlin Initiative Study (BIS and BIS ) and full age spectrum. Sensitivity, specificity, areas under receiver operating curve (AUC) and C-statistics were used to compare their predictive capacity.
The best mix of sensitivity, specificity, AUC and C-statistic value in predicting mortality was observed with BIS equations. BIS (AUC 0.65, 95% CI 0.61-0.69) outperformed both CKD-EPI (AUC 0.60, 95% CI 0.56-0.64; P = 0.005) and full age spectrum (AUC 0.63, 95% CI 0.59-0.67; P = 0.002) in terms of predictivity. Similarly, BIS (AUC 0.67, 95% CI 0.63-0.71) outperformed CKD-EPI (AUC 0.63, 95% CI 0.59-0.67; P = 0.01). AUC obtained with equations also including cystatin C were not significantly different compared with their creatinine-based counterparts. The risk of long-term mortality began to increase at under 65.6 mL/min/1.73 m for CKD-EPI , 60.5 for CKD-EPI , 60 for BIS , 56.3 for BIS and 55.2 for full age spectrum.
The BIS equation discriminates the risk of all-cause mortality better than other equations in older community-dwelling individuals. The eGFR threshold under which mortality starts to increase could change as a function of the equation used. Geriatr Gerontol Int 2018; 18: 607-614.
在预测预后方面,对于老年人来说,哪种肾小球滤过率估计(eGFR)方程存在不确定性。我们的目的是:(i)比较不同方程估计的 eGFR 对全因死亡率的区分能力;(ii)确定每个方程中 eGFR 阈值开始增加的死亡率风险。
我们使用了来自 828 名年龄> 65 岁的社区居住老年人的 INCHIANTI 研究数据。主要结局是 9 年的全因死亡率。通过五种不同的方程估计 GFR:慢性肾脏病流行病学合作研究(肌酐方程[CKD-EPI]和肌酐和胱抑素 C 方程[CKD-EPI])、柏林倡议研究(BIS 和 BIS )和全年龄谱。使用敏感性、特异性、受试者工作特征曲线下面积(AUC)和 C 统计量来比较其预测能力。
在预测死亡率方面,BIS 方程的敏感性、特异性、AUC 和 C 统计量的最佳组合。BIS(AUC 0.65,95%CI 0.61-0.69)在预测能力方面优于 CKD-EPI(AUC 0.60,95%CI 0.56-0.64;P = 0.005)和全年龄谱(AUC 0.63,95%CI 0.59-0.67;P = 0.002)。同样,BIS(AUC 0.67,95%CI 0.63-0.71)优于 CKD-EPI(AUC 0.63,95%CI 0.59-0.67;P = 0.01)。与基于肌酐的方程相比,包含胱抑素 C 的方程的 AUC 没有显著差异。CKD-EPI 为 65.6 mL/min/1.73 m,CKD-EPI 为 60.5,BIS 为 60,BIS 为 56.3,全年龄谱为 55.2,全因死亡率开始增加的风险开始增加。
在社区居住的老年人中,BIS 方程比其他方程更能区分全因死亡率的风险。死亡率开始增加的 eGFR 阈值可能会随着所使用的方程而变化。老年医学与老年病学国际 2018;18:607-614。