Fluid Volumes in Adults with Sepsis

Less versus more intravenous fluid did not affect mortality in adults with sepsis

Benefits in NNT

No one was helped (no death prevented)
No one was helped

Harms in NNT

No one was harmed (no difference in adverse events)
No one was harmed
View As:


Long B, Gottlieb M. Fluid volumes in adults with sepsis. Academic Emergency Medicine. Published online April 3, 2024:acem.14912.

Study Population: 4,006 participants in 13 trials of adult patients with sepsis

Efficacy Endpoints

All-cause mortality

Harm Endpoints

Serious adverse events


Sepsis, defined by organ dysfunction and a dysregulated response to infection, accounts for up to one in six hospital admissions in the United States and nearly 270,000 deaths annually.1, 2 Treatment includes source control, antibiotics, and ensuring oxygenated blood flow supported by intravenous (IV) fluid administration. However, limited evidence on “more versus less” IV fluids has led to conflicting, uncertain recommendations.3, 4, 5, 6 The 2021 international guidelines for management of sepsis and septic shock “Surviving Sepsis Campaign” suggests administering 30 mL/kg of fluids in the first 3 h.3 However, other studies have shown either no benefit or potentially increased mortality with such a strategy.4, 5, 6 A systematic review published in 2020 found no difference in mortality between lower and higher fluid volumes,7 but several large, high-quality randomized trials have been published since then.8, 9, 10, 11

The review summarized here included 13 trials of 4006 adult participants with sepsis.12 The review included trials comparing fluid strategies intended to “obtain a separation in IV fluid volumes.” The authors focused on clinical outcomes (e.g., mortality) and excluded trials comparing different fluids and those focused on burn patients or severe blood loss. The primary outcomes were assessed at roughly 90 days and included (1) mortality, (2) adverse events, and (3) health-related quality of life. Secondary outcomes included duration and need for mechanical ventilation, need for renal replacement therapy (e.g., dialysis), and acute kidney injury. Additional outcomes included blood products, intensive care unit (ICU) length of stay, and hospital length of stay.

All 13 included trials were published between 2015 and 2023. Seven were multicenter, 10 were conducted in the ICU, and three in the emergency department (ED). All enrolled septic patients with hypoperfusion, severe sepsis, or septic shock.

Meta-analysis of all 13 trials found that less versus more fluid had no effect on mortality (relative risk [RR] 0.98, 95% confidence interval [CI] 0.9–1.1, moderate-quality evidence) or serious adverse events (RR 0.95, 95% CI 0.8–1.1, low-quality evidence).12 No data were reported on health-related quality of life, and no difference was found in any other outcome. The average fluid volume for resuscitation was 1679 mL in the less fluid group and 2775 mL in the more fluid group. The mean total volume of fluid was 6950 mL in the less fluid group and 7828 mL in the more fluid group.


There are several important considerations when interpreting these results. First, five of 13 trials were unable to separate fluid volumes between the groups. However, analysis of only trials with separation also found no differences. Moreover, analysis of only high-quality trials found no difference in mortality. Second, definitions of sepsis and septic shock varied across studies. Third, the patient populations and interventions were significantly different. Seven trials included ICU patients only, three ED patients only, and three combined both. Importantly, several trials used complex protocols with a bundle of interventions, and in most, patients received IV fluids before inclusion. Crystalloids were the most common fluid; however, one trial used hydroxyethyl starch, which may be harmful.13, 14 This heterogeneity in populations and interventions could dilute any effects. Fourth, serious adverse events are often underreported. Accordingly, only six trials evaluated serious adverse events using varying definitions.

It is reassuring, however, that the two largest and highest quality trials, representing more than 85% of all subjects in high-quality trials, tested similar methods and populations and achieved significant separation in fluid volumes.8, 10 In both trials, mortality, the primary outcome, was unchanged by fluid strategy.

Based on current data, it seems highly likely there is no meaningful difference in clinical outcomes between less versus more IV fluid strategies for people experiencing severe sepsis. We have therefore selected a color recommendation of red (no benefits) for the choice of fluid volumes in adult patients with sepsis. Further randomized controlled trials using well-defined fluid protocols in well-defined populations, as well as those comparing timing of vasopressors, will hopefully shed further light on this topic.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.


Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD


April 26, 2024