Intravenous Fluid Volume Is Not Associated With Worsening Left Ventricular Function in Septic Patients With or Without Heart Failure: A Bayesian Analysis

Volume resuscitation of septic patients is challenging and must balance harms and benefits. Many clinicians are concerned that adhering to guideline-recommended intravenous fluid (IVF) volumes may result in fluid overload, particularly in patients with a history of heart failure (HF). The goal of this study was to compare the association between IVF volume and left ventricular function over the first 24-hours of admission in septic patients with, versus without, HF history.

Secondary analysis of a prospective, observational cohort of adults presenting to an urban Emergency Department with sepsis (treatment for suspected infection plus either lactate >2mmol/L or systolic blood pressure ≤90mm/hg). Patients had an echocardiogram performed at 0, 3, and 24-hours after ED arrival, with assessment of left ventricular ejection fraction (LVEF), e’ velocity, and E/e’. The primary outcome was the association of 24-hour IVF volume and LV function (modeled as a tensor product spline interaction between IVF volume and ED SOFA score) assessed using multilevel Bayesian location-scale models.

Seventy-three patients were enrolled, with mean (SD) age, 24-hour IVF volume, and ED SOFA score of 60.7 years (15.6), 3.8L (2.1), and 3.8 (2.8), respectively. In-hospital mortality was 23% and 19% had a history of HF. For all measures of LV function, after adjusting for age, biologic sex, history of heart failure, and inferior vena cava collapsibility, each SD increase in IVF volume was associated with small improvements in LV function in patients with, or without, HF history: average marginal effect 2.4% (95% credible interval(CI) -2.6% to 10%) versus 1.3% (CI -1.2 to 4.7%) for LVEF; 1.2 cm/s (CI -0.6 to 3.1 cm/s) versus 0.8 cm/s (CI -0.004 to 1.2 cm/s) for e’ velocity; -1.72 (CI -6.9 to 2.5) versus -0.92 (CI -3.1 to 1.0) for E/e’.

While a small, positive association between IVF volume and LV function was found, CIs included the possibility of no change or slightly worsened function; results were similar for those with or without HF history. In most cases, the magnitude of change (regardless of direction) was small and likely not clinically significant. Thus, restrictive approaches to volume expansion in septic patients with HF are not supported. Further studies, with focus on clinical outcomes, are needed to confirm our findings.