论文部分内容阅读
目的:比较三种肾小球滤过率(GFR)计算方法(e GFRcre、e GFRcys与e GFRcre-cys)评价2型糖尿病肾病(T2DN)患者肾功能的准确性及判断预后的临床价值。方法:纳入随访登记数据库中T2DN患者501例,通过横断面分析,比较0、12、24月时用e GFRcre、e GFRcys与e GFRcre-cys公式计算得到e GFR的分布及偏移状态,并随访观察12月和24月时不同e GFR下降幅度(20%、30%、40%和57%)对判断患者肾脏终点事件的敏感性和特异性。结果:将e GFRcre-cys设为参考,e GFRcre公式会高估GFR水平,在e GFRcre-cys>120 ml/(min·1.73m~2),90~120 ml/(min·1.73m~2),60~89 ml/(min·1.73m~2),30~59 ml/(min·1.73m~2),<30 ml/(min·1.73m~2)的T2DN患者中,中位偏移分别为-0.21 ml/(min·1.73m~2),6.17 ml/(min·1.73m~2),12.69 ml/(min·1.73m~2),9.55 ml/(min·1.73m~2),5.27 ml/(min·1.73m~2)。低龄、高收缩压、低体质指数(BMI)和高血糖增加偏移度。基于e GFRcre-cys公式24月的GFR下降速率较e GFRcre和e GFRcys公式更好的判断肾脏终点,曲线下面积(ROC)分别为0.857±0.029 7,0.839±0.031 9和0.796±0.035 7。24月的e GFRcre-cys下降速率联合基线e GFRcrecys的GFR水平可有效判断肾脏预后,ROC达0.917±0.017 8。基线e GFRcre-cys≥90 ml/(min·1.73m~2)的患者,以e GFRcre-cys 24月下降57%作为判断指标,敏感性和特异性均为100%;基线e GFRcre-cys<90 ml/(min·1.73m~2)的患者,以e GFRcre-cys 24月下降40%作为判断预后指标,其准确性更好。结论:e GFRcre公式可能高估T2DN患者的GFR,尤其在e GFR 60~120 ml/(min·1.73m~2)的患者。以e GFRcre-cys公式计算的24月GFR下降57%和40%分别用于判断基线e GFRcre-cys≥90 ml/(min·1.73m~2)和<90 ml/(min·1.73m~2)的T2DN患者肾脏预后的敏感度和特异度较好。
OBJECTIVE: To evaluate the accuracy and prognosis of renal function in patients with type 2 diabetic nephropathy (T2DN) by comparing three GFR calculation methods (e GFRcre, e GFRcys and e GFRcre-cys). METHODS: 501 patients with T2DM in the follow-up enrolled database were enrolled in this study. The eGFR distribution and migration status were calculated by eGFRcre, eGFRcys and eGFRcre-cys formula at 0, 12, 24 months compared with 0, 12, and 24 months To observe the sensitivity and specificity of different eGFR declines (20%, 30%, 40% and 57%) in determining the end point of renal disease in patients at December and 24 months. Results: Using e GFRcre-cys as a reference, the e GFRcre formula would overestimate the GFR level, with e GFRcre-cys> 120 ml / (min · 1.73 m 2), 90-120 ml / (min · 1.73 m 2 ), 60 ~ 89 ml / (min · 1.73m ~ 2), 30 ~ 59ml / (min · 1.73m ~ 2) and <30ml / (min · 1.73m ~ 2) (Min · 1.73m ~ 2), 6.17ml / (min · 1.73m ~ 2), 12.69ml / (min · 1.73m ~ 2) and 9.55ml / (min · 1.73m ~ 2 ), 5.27 ml / (min · 1.73 m ~ 2). Older age, high systolic blood pressure, low body mass index (BMI) and hyperglycemia increased bias. Based on the eGFRcre-cys formula, the GFR decline rate in 24 months was better than the eGFRcre and eGFRcys formulas in determining renal end points. The area under the curve (ROC) were 0.857 ± 0.029 7, 0.839 ± 0.031 9 and 0.796 ± 0.035 7.24, respectively The monthly decline of eGFRcre-cys combined with the baseline GFRcrecys GFR level can effectively determine the renal prognosis, ROC 0.917 ± 0.0178. The baseline eGFRcre-cys≥90 ml / (min · 1.73m ~ 2) patients, eGFRcre-cys decreased by 57% in 24 months as a diagnostic indicator, the sensitivity and specificity were 100%; baseline e GFRcre-cys < 90 ml / (min · 1.73m ~ 2) in patients with e GFRcre-cys decreased by 40% in 24 months as a prognostic indicator, the accuracy is better. Conclusion: The eGFRcre formula may overestimate the GFR in patients with T2DM, especially in patients with e GFR of 60-120 ml / (min · 1.73 m 2). GFRcre-cys calculated in accordance with eGFRcre-cys formula of 57% and 40% decreased GFR were used to determine baseline e GFRcre-cys ≥ 90 ml / (min · 1.73m ~ 2) and <90 ml / (min · 1.73m ~ 2 ) T2DM patients with better prognosis of the kidney sensitivity and specificity.