A multicenter, randomized feasibility analysis comparing restrictive fluids to standard therapy for people with sepsis in the emergency department (REFACED)

Abstract

Aim: Sepsis presents a problem for fluid management, and clinical consensus exists about intravenous (IV) volume requirements. When adult patients with sepsis who were not in shock presented to the emergency room, our goal was to assess if a 24-hour strategy restricting IV fluid was possible (ED). Methods: The REFACED Sepsis trial is an investigator-initiated, multicenter, randomised, open-label, feasibility study in which sepsis patients who are not in shock are randomly assigned to receive either conventional therapy or 24 hours of restrictive, crystal IV fluid delivery. Fluid boluses were only allowed in the IV fluid restriction group when certain conditions for hypoperfusion were met. The treating team determined the standard of care. Total IV crystalloid fluid volumes were the main result 24 hours after randomization. Total fluid volumes, feasibility tests, and patient-centered outcomes were also considered secondary outcomes. Results: In the initial analysis, we included 123 patients (61 patients under restrictive care and 62 patients under normal treatment). The proportion of eligible patients who met all inclusion criteria and no exclusion criteria was 32% (95% confidence interval [CI] 28%-37%). At 24hrs, the restriction versus standard care groups had mean (± SD) IV crystalloid fluid amounts of 562 (1076) ml and 1370 (1438) ml, respectively (mean difference -801 ml, 95% CI - 1257 to - 345 ml, p = 0.001). In the fluid-restrictive group, 21 (34%) of the patients experienced protocol violations. Adverse events, the need for mechanical breathing or vasopressors, acute renal failure, length of stay, or mortality did not differ across groups. Conclusions: In comparison to normal care, a regimen restricting IV crystalloid fluids in ED patients with sepsis resulted in lower 24-hour fluid amounts. It is possible that a future experiment may focus on patient-centered outcomes.