Bedside echocardiography for predicting early hemodynamic instability during continuous venovenous hemodiafiltration: a prospective observational study
摘要
Early hemodynamic instability and intradialytic hypotension are frequent and clinically significant complications at the initiation of continuous venovenous hemodiafiltration (CVVHDF) in intensive care unit (ICU) patients. Reliable bedside predictors before dialysis remain limited. This study evaluated whether pre-dialysis transthoracic echocardiographic parameters, particularly left ventricular outflow tract velocity–time integral (LVOT VTI), can predict early hemodynamic instability after CVVHDF initiation. In this prospective observational cohort study, adult ICU patients scheduled for CVVHDF underwent standardized transthoracic echocardiography immediately before dialysis initiation. LVOT VTI, mitral annular plane systolic excursion (MAPSE), tricuspid annular plane systolic excursion (TAPSE), systolic myocardial velocity (S′), and the E/e′ ratio were recorded. Patients were monitored for 60 min after CVVHDF initiation. Early hemodynamic instability was defined as hypotension, significant mean arterial pressure reduction, need for fluid resuscitation, or initiation/escalation of vasoactive support. Receiver operating characteristic analyses were performed. Forty patients were included; 19 developed early hemodynamic instability. Baseline LVOT VTI, MAPSE, and TAPSE were significantly lower in unstable patients. ROC analysis showed an AUC of 1.00 for LVOT VTI in relation to early hemodynamic instability. A cohort-derived LVOT VTI cut-off of 17.0 cm identified unstable patients with 100% sensitivity and 100% specificity in the present dataset; however, this exploratory threshold requires external validation. Mitral E/e′ also showed strong predictive performance, whereas S′ velocities were not discriminatory. Pre-dialysis LVOT VTI was strongly associated with early hemodynamic instability following CVVHDF initiation and may represent a promising bedside echocardiographic risk-stratification marker. However, the observed 17.0 cm threshold requires validation in larger multicenter cohorts before clinical implementation.
Trial registration: ClinicalTrials.gov, NCT07019051. Registered June 23, 2025.