Supplementary Material for: The Degree of the Predischarge Pulmonary Congestion in Patients Hospitalized for Worsening Heart Failure Predicts Readmission and Mortality

Background: Prediction of readmission and death after hospitalization for heart failure (HF) is an unmet need. Aim: We evaluated the ability of clinical parameters, NT-proBNP level and noninvasive lung impedance (LI), to predict time to readmission (TTR) and time to death (TTD). Methods and Results: The present study is a post hoc analysis of the IMPEDANCE-HF extended trial comprising 290 patients with LVEF ≤45% and New York Heart Association functional class II–IV, randomized 1:1 to LI-guided or conventional therapy. Of all patients, 206 were admitted 766 times for HF during a follow-up of 57 ± 39 months. The normal LI (NLI), representing the “dry” lung status, was calculated for each patient at study entry. The current degree of pulmonary congestion (PC) compared with its dry status was represented by ΔLIR = ([measured LI/NLI] – 1) × 100%. Twenty-six parameters recorded during HF admission were used to predict TTR and TTD. To determine the parameter which mainly impacted TTR and TTD, variables were standardized, and effect size (ES) was calculated. Multivariate analysis by the Andersen-Gill model demonstrated that ΔLIRadmission (ES = 0.72), ΔLIRdischarge (ES = –3.14), group assignment (ES = 0.2), maximal troponin during HF admission (ES = 0.19), LVEF related to admission (ES = –0.22) and arterial hypertension (ES = 0.12) are independent predictors of TTR (p χ2 = 1,206). Analysis of ES showed that residual PC assessed by ∆LIRdischarge was the most prominent predictor of TTR. One percent improvement in predischarge PC, assessed by ∆LIRdischarge, was associated with a likelihood of TTR increase by 14% (hazard ratio [HR] 1.14, 95% confidence interval [CI] 1.13–1.15, p p Conclusion: The degree of predischarge PC assessed by ∆LIR is the most dominant predictor of TTR and TTD.

研究背景：心力衰竭（Heart Failure，HF）患者住院后再入院与死亡的预测仍是尚未满足的临床需求。研究目的：本研究评估了临床参数、N末端B型利钠肽前体（NT-proBNP）水平以及无创肺阻抗（Noninvasive Lung Impedance，LI）对再入院时间（Time to Readmission，TTR）与死亡时间（Time to Death，TTD）的预测能力。方法与结果：本研究为针对IMPEDANCE-HF扩展试验的事后分析，共纳入290例左心室射血分数（Left Ventricular Ejection Fraction，LVEF）≤45%且纽约心脏协会（New York Heart Association，NYHA）心功能II~IV级的患者，按1:1比例随机分配至LI指导治疗组与常规治疗组。所有患者的随访时长为57±39个月，期间共有206例患者因HF累计入院766次。研究入组时，为每位患者计算代表肺部"干燥"状态（即无肺淤血状态）的基础肺阻抗（Normal Lung Impedance，NLI）。当前肺淤血（Pulmonary Congestion，PC）程度相较于其干燥状态的差值，以ΔLIR = ([实测LI/NLI] – 1) × 100%进行计算表示。本研究选取HF入院期间记录的26项参数，用于预测TTR与TTD。为明确对TTR与TTD影响最为显著的参数，研究对变量进行了标准化处理，并计算了效应量（Effect Size，ES）。采用Andersen-Gill模型进行多因素分析结果显示：入院时ΔLIR（ES = 0.72）、出院时ΔLIR（ES = –3.14）、分组情况（ES = 0.2）、HF入院期间肌钙蛋白峰值（ES = 0.19）、入院时相关LVEF（ES = –0.22）以及动脉高血压（ES = 0.12）均为TTR的独立预测因素（χ²检验P = 1206）。效应量分析显示，通过出院时ΔLIR评估的残余肺淤血是TTR最显著的预测因素。经出院时ΔLIR评估的出院前肺淤血每改善1%，与TTR增加14%的风险相关（风险比（Hazard Ratio，HR）1.14，95%置信区间（Confidence Interval，CI）1.13–1.15，P值）。研究结论：通过ΔLIR评估的出院前肺淤血程度是TTR与TTD最主要的预测因素。