Early Fluid Resuscitation Volume as an Independent Determinant of Mortality in Sepsis: A Multicenter Real-World ICU Cohort Study

Authors

  • Amina El-Sayed
    Department of Infectious Diseases, General Hospital, Cairo, Egypt
  • Omar Hassan
    Department of Infectious Diseases, General Hospital, Cairo, Egypt
  • Youssef Khaled
    Department of Infectious Diseases, General Hospital, Cairo, Egypt

Abstract

Introduction: Early intravenous fluid resuscitation is central to sepsis management; however, the optimal volume during the initial resuscitation window remains uncertain. Although current guidelines recommend at least 30 mL/kg of crystalloids within 3 h, this fixed-volume threshold may not capture the heterogeneity of septic ICU care unit. This study evaluated the association between early fluid resuscitation volume and mortality using multicenter, real-world ICU data.


Methods: We conducted a retrospective multicenter cohort study using MIMIC-IV and eICU-CRD. Adult ICU patients who fulfilled the Sepsis-3 criteria were included. The primary exposure was the cumulative crystalloid volume administered within the first 3 h after sepsis onset. Fluid volume was analyzed as categorical mL/kg strata, the conventional ≥30 mL/kg threshold, and a continuous variable using restricted cubic splines. The primary outcome was in-hospital mortality. Multivariable logistic regression, propensity score weighting, and marginal structural models were used to address the baseline severity, treatment intensity, and time-varying confounding.


Results: Among 18,742 septic ICU care unit, early fluid volume showed a nonlinear dose–response association with mortality. Patients receiving 20–30 mL/kg had the lowest adjusted mortality, whereas both lower-volume resuscitation (< 10 mL/kg) and liberal resuscitation (≥ 40 mL/kg) were associated with increased mortality. The ≥30 mL/kg threshold was not consistently associated with improved survival after the adjustment. The findings remained robust across sensitivity analyses, alternative exposure windows, and causal inference models.


Conclusion: Early fluid resuscitation volume in patients with sepsis is associated with mortality in a non-linear, dose-dependent pattern. These findings challenge the universal applicability of fixed 30 mL/kg resuscitation and support individualized physiology-guided fluid strategies in critically ill patients with sepsis.