老年慢性心力衰竭恶化的风险模型构建及NT-proBNP的预测价值探讨

A nomogram to predict the risk of worsening chronic heart failure in the elderly and the predictive value of NT-proBNP

目的:构建并验证老年慢性心力衰竭恶化(WHF)的风险预测列线图,并探讨列线图与N末端B型利钠肽前体(NT-proBNP)联合预测WHF的价值。方法:回顾性纳入2018年1月—2022年6月在我院连续就诊的老年WHF患者(WHF组,572例),另纳入同期于我院门诊规律随访的老年慢性心力衰竭稳定期患者561例(Non-WHF组)。比较WHF组与Non-WHF组临床指标的差异。运用LASSO回归联合多因素logistic回归模型构建老年WHF风险预测列线图。由内部验证集评估列线图的可行性,采用受试者工作特征曲线(ROC曲线)、校准曲线和决策曲线(DCA)评价列线图的区分度、校准度及临床实用性。最后通过对比ROC曲线下面积(AUC)、整体鉴别指数(IDI)及净重新分类指数(NRI)来评估列线图、NT-proBNP及其联合使用在老年WHF人群中的预测价值。结果:入院心率、合并心房颤动、左心室射血分数(LVEF)、红细胞分布宽度(RDW)、肌酐(Scr)、血清尿素氮/白蛋白比值(BAR)是WHF发生的独立危险因素(均P<0.05);根据以上6个变量绘制相应的可视化列线图,列线图在训练集和验证集中的AUC分别为0.830(95%CI:0.802~0.858)、0.834(95%CI:0.792~0.876)。DCA提示该列线图的校准度好,临床实用性强。经ROC曲线分析,列线图、NT-proBNP及其联合预测老年WHF的AUC值分别为0.830、0.890和0.924。相对于NT-proBNP,列线图和NT-proBNP联合预测的IDI、NRI分别为0.073、0.065。结论:包含6个预测变量(入院心率、合并心房颤动、LVEF、RDW、Scr和BAR)的列线图可用于预测老年慢性心力衰竭患者WHF发生风险。此列线图与NT-proBNP联合评估更有助于识别高风险的老年WHF人群,协助医务工作者制定临床决策。

Objective:To establish and verify a nomogram for predicting the risk of worsening chronic heart failure(WHF) in elderly patients,as well as to investigate the value of combining the nomogram with N-terminal B-type natriuretic peptide(NT-proBNP) in predicting WHF.Methods:We retrospectively reviewed and analyzed the clinical characteristics of 572 elderly patients with WHF(WHF group) and 561 elderly patients with stable chronic heart failure who underwent regular follow-up(Non-WHF group) in our medical center from January 2018 to June 2022.Clinical indexes between the two groups were compared.The elderly WHF nomogram was constructed by LASSO regression combined with multivariate logistic regression analysis.The internal validation set was used to evaluate the feasibility of the nomogram.The receiver operating characteristic curve(ROC),calibration curve,and decision curve analysis(DCA) were adopted to evaluate the discrimination,calibration,and clinical validity of the nomogram.The area under the receiver operating characteristic curve(AUC),integrated discrimination index(IDI),and net reclassification index(NRI) were applied to evaluate the predictive value of the nomogram,NT-proBNP,and their combination in elderly patients with WHF.Results:Admission heart rate,concomitant atrial fibrillation,left ventricular ejection fraction,red cell distribution width,creatinine,and blood urea nitrogen/albumin ratio were independent predictors of WHF in elderly patients(P<0.05).The visual nomogram was drawn based on the six variables,and the AUCs of the nomogram in the training set and validation set were 0.830(95%CI:0.802-0.858) and 0.834(95%CI:0.792-0.876).Calibration curves and DCA curves demonstrated perfect calibration and clinical practicability of the nomogram.This study found that the NT-proBNP levels of patients in the WHF group were significantly higher than those in the Non-WHF group(P<0.001).Meanwhile,the AUCs of the nomogram,NT-proBNP,and their combination were 0.830,0.890,and 0.924,respectively.Compared with NT-proBNP,the IDI and NRI of their combination were 0.073 and 0.065.Conclusion:The nomogram containing six predictors could be used to predict the risk of WHF in elderly patients.The combined assessment of nomogram and NT-proBNP contributes to recognizing the high risk of WHF in elderly patients and helping medical staff to make clinical strategies.