Li Yiou, Chen Jiajia, Bian Jianye, Chen Fangyuan, Wan Qianli, Yuan Fang
Department of Critical Care Medicine, Tongren Hospital Shanghai Jiao Tong University School of Medicine, 200336 Shanghai, China.
Rev Cardiovasc Med. 2025 May 22;26(5):27100. doi: 10.31083/RCM27100. eCollection 2025 May.
Ultrafiltration (UF) is an alternative approach to diuretic therapy for the treatment of acute heart failure (AHF), but its optimal endpoint is unclear. This study explores using non-invasive ultrasonic cardiac output monitor (USCOM) to determine UF endpoints based on hemodynamic changes.
In this single-anonymized, randomized controlled trial, acute decompensated heart failure patients were randomly assigned to UF (U, n = 20) and USCOM+UF (UU, n = 20) groups at a ratio of 1:1. A mixed linear model was utilized to analyze repeated measurement data of hemodynamic indicators (primary endpoint) in the U and UU groups. A 30% or 50% decrease in B-type natriuretic peptide (BNP) concentrations relative to the baseline was established as the criteria for the UF endpoint success. Multivariate logistic regression was used to identify potential indicators within the USCOM that could have influenced the UF endpoint success. Receiver operating characteristic (ROC) curves were used to evaluate the value of the predictive model. Economic benefits, including treatment costs and hospitalization duration, were also assessed.
Change rates in mean arterial pressure, heart rate (HR), urine output, hematocrit, and BNP concentrations were similar between the U and UU groups over 7 days (all > 0.05). On day 4, significant correlations were found between various USCOM parameters, including inotropy (INO), systemic vascular resistance index (SVRI), systemic vascular resistance, corrected flow time (FTc), velocity time integral, and the BNP of the UF parameters. Multivariate logistic regression revealed that INO and SVRI were correlated with a 30% reduction in BNP on day 4 compared to baseline, while FTc and HR were found to be independently associated with a 50% reduction in BNP on day 4 compared to baseline. The UF endpoint prediction formula for a 30% reduction in BNP was -2.462 + 0.028 × INO - 0.069 × SVRI, with sensitivities, specificities, and accuracies of 70%, 83%, and 75%, respectively. The UF endpoint prediction formula for a 50% reduction of BNP was -2.640 - 0.088 × FTc - 0.036 × HR, with sensitivities, specificities, and accuracies of 83%, 63.0%, and 72.5%, respectively. The addition of the USCOM significantly reduced treatment costs and hospitalization stay lengths.
Observing the USCOM using probability formulas served to determine appropriate UF endpoints during AHF treatments. UF combined with the USCOM can reduce the costs of UF and hospitalization.
NCT06533124, https://clinicaltrials.gov/study/NCT06533124?term=NCT06533124&rank=1.
超滤（UF）是治疗急性心力衰竭（AHF）的一种替代利尿治疗的方法，但其最佳终点尚不清楚。本研究探讨使用无创超声心输出量监测仪（USCOM）根据血流动力学变化来确定超滤终点。
在这项单盲、随机对照试验中，急性失代偿性心力衰竭患者按1:1的比例随机分为超滤组（U组，n = 20）和USCOM + 超滤组（UU组，n = 20）。采用混合线性模型分析U组和UU组血流动力学指标（主要终点）的重复测量数据。将B型利钠肽（BNP）浓度相对于基线降低30%或50%确定为超滤终点成功的标准。使用多因素逻辑回归来识别USCOM中可能影响超滤终点成功的潜在指标。采用受试者工作特征（ROC）曲线评估预测模型的价值。还评估了包括治疗成本和住院时间在内的经济效益。
U组和UU组在7天内平均动脉压、心率（HR）、尿量、血细胞比容和BNP浓度的变化率相似（均P>0.05）。在第4天，发现各种USCOM参数之间存在显著相关性，包括心肌收缩力（INO）、全身血管阻力指数（SVRI）、全身血管阻力、校正血流时间（FTc）、速度时间积分以及超滤参数中的BNP。多因素逻辑回归显示，与基线相比，第4天INO和SVRI与BNP降低30%相关，而FTc和HR被发现与第4天BNP降低50%独立相关。BNP降低30%的超滤终点预测公式为 -2.462 + 0.028×INO - 0.069×SVRI，敏感性为70%，特异性为83%，准确性为75%。BNP降低50%的超滤终点预测公式为 -2.640 - 0.088×FTc - 0.036×HR，敏感性为83%，特异性为63.0%，准确性为72.
