Abstract Rationale IV fluids are critical for treating sepsis. Yet, optimal fluid strategies remain debated. On September 26, 2024, Hurricane Helene disrupted IV fluid production at a manufacturing facility that produced 60% of IV solutions nationwide. This created an opportunity for a natural time-series ecological study on fluid management strategies. Here, we characterize fluid utilization during this restrictive period and its impact on patients admitted with sepsis. Methods We conducted a retrospective cohort study using the Common Longitudinal ICU data Format (CLIF). We identified adult patients with sepsis admitted from the emergency department during four time periods: Temporal control (October 2023-January 2024), Pre-shortage (June 2024-September 2024), During shortage (October 2024-January 2025), Post-shortage (February 2025-May 2025) from a 10-hospital health system in Minnesota. Encounters were identified using ICD-10 codes and demographics, illness severity, volume of fluid administration, and hospital stay characteristics were collected. Primary outcome was 30-day mortality. Secondary outcomes included vasopressor and mechanical ventilator use. Multivariable regression models estimated the association between admission period and patient outcomes. Sensitivity analyses adjusted for the first 24-hour fluid administration and patients with qSOFA scores equal or greater than 2. Results Among 13,098 patients admitted with sepsis across all time periods, cohorts were similar in age and race but differed in qSOFA score. During the shortage, a median volume of fluids of 1.96L were given in the first 24 hours and was significantly less compared to the other time periods (2.09-2.39L). Similarly, hospitalization fluid administration was decreased during the shortage (2.66L vs. 3.11-3.81L). Pairwise comparisons using Wilcoxon rank sum tests of fluid administration in the first 24 hours were statistically significant for all comparisons except for between the temporal control and pre-shortage periods. In multivariable logistic regression models comparing the shortage period with other cohorts, there were no differences of in-hospital, 30-day, or 90-day mortality. There was a significant decrease in the odds of mechanical ventilation when compared to the temporal control (OR 1.25, 95% CI 1.05-1.49, p = 0.013), but no difference in vasopressor initiation. Sensitivity analyses for 24-hour fluid administration and severity of illness showed similar findings. Conclusions During a nationwide IV fluid shortage caused by Hurricane Helene, fluid administration decreased by 20-30%, yet mortality and vasopressor use remained unchanged while mechanical ventilation rates declined. These findings suggest that more conservative fluid management in sepsis may be safe and potentially beneficial and highlight the resilience of health systems in mitigating supply chain disruptions. This abstract is funded by: None
Shyu et al. (Fri,) studied this question.