EDITORIAL| VOLUME 164, ISSUE 4, P812-813, OCTOBER 2023

Fluid Therapy in Sepsis: Does It Matter How Much?

Jonathan A. Silversides, Bram Rochwerg

Chest 2023; 164: 812-813

DOI:https://doi.org/10.1016/j.chest.2023.05.027

Although IV fluid resuscitation is a fundamental element of sepsis management, debate regarding the ideal volume, type, timing, and physiologic targets is ongoing. A number of large randomized trials have been published addressing this important topic, and in this issue of CHEST, Sivapalan et al¹ have updated a previous systematic review and meta-analysis comparing lower with higher volumes of IV fluid for adult patients with sepsis.

IV fluid resuscitation aims to increase venous return and cardiac output, thereby restoring blood flow to underperfused organs. Since the 2010s, a multitude of observational studies have described a consistent association between administration of greater volumes of IV fluid, leading to positive fluid balance, and adverse patient outcomes in sepsis and other critical illness. This has stimulated an interest in alternative, more conservative, fluid strategies.² The landmark Fluid Expansion as Supportive Therapy (FEAST) trial demonstrated higher mortality with IV fluid bolus than with slower IV fluid infusions in children with infection-associated shock in sub-Saharan Africa.³ This led to a paradigm shift: rapid administration of IV fluid boluses came to be seen as potentially harmful, and paved the way for large randomized trials evaluating restrictive fluid strategies in adults with sepsis, several of which are included in this systematic review and meta-analysis.^4,5

This review included 13 randomized controlled studies, most of which were small and single center. Most of the evidence was from two large trials, which together contributed 3,112 of the 3,978 patients.^4,5 There was no important difference in mortality between fluid strategies, and trial sequential analysis indicated that the predefined boundary for futility had been crossed, meaning the likelihood of a relative risk reduction of greater than 15% was very unlikely. For other outcomes, including incidence of serious adverse events, duration of mechanical ventilation, use of vasopressors, and renal replacement therapy, there was similarly no difference between lower and higher volumes of IV fluid, with varying degrees of certainty across outcomes. None of the included trials reported on health-related quality of life. Subgroup analyses evaluating early (up to 6 h) vs late (greater than 6 h) stages of resuscitation, and comparing successful vs unsuccessful separation in fluid balance between arms, did not reveal evidence of effect modification.

On the basis of the results of this review, it seems that the volume of fluid administered is unlikely to have a moderate to large effect on mortality and other clinical outcomes in adult patients with sepsis. Nevertheless, despite the methodologic rigor with which the review was conducted, some important limitations are worth highlighting. First, a small effect on mortality and other clinical outcomes cannot be excluded, and given the ubiquitous nature of IV fluid resuscitation for patients with sepsis, even a small effect may be important on a global scale. Second, the findings of this review are limited by the design of the included studies. Fluid resuscitation is a complex intervention, and the triggers, timing, and end points for administration matter and were relatively heterogeneous across the included studies. None of the included studies sought to replicate or even approximate a restrictive strategy similar to that used in FEAST, in that rapid administration of significant volumes of IV fluids had already occurred before randomization. Importantly, the differences amongst included studies between those randomized to lower-volume as compared with higher-volume arms were small, on the order of 1 to 2 L. Even for those in the higher-volume group, amounts of IV fluid administered were relatively low based on historical terms given evolving practice toward a more judicious use of IV fluids.^6,7 It is almost implausible that such small differences in volume of IV fluid administered would lead to a 15% relative reduction in mortality, the threshold for futility that was used in the trial sequential analysis.

Although this review is informative, there remain important unanswered questions in relation to fluid therapy. It is remarkable that no randomized controlled trial has attempted to replicate FEAST in higher-income settings given the clear harm demonstrated with bolus fluid administration at the point of presentation. Most trials focus on bolus fluids; however, much of the fluid administered to critically ill patients is obligate, for example, in nutrition or drug diluents, and accumulation of a positive fluid balance over time is not dependent solely on IV fluid classically referred to as “resuscitation.”⁸ Future trials should investigate comprehensive strategies to avoid accumulation of a positive fluid balance over time, and not focus solely on “resuscitation.”

For the practicing clinician, faced with a patient with sepsis and hypotension and/or hypoperfusion, the results of this review should be reassuring. Within the parameters tested in the included trials, the decision to administer carefully titrated fluid boluses with the aim of restoring end-organ and tissue perfusion is unlikely to cause harm. At the same time, using a more restrictive fluid administration strategy seems equally safe, and concerns that this might lead to harm through excessive vasopressor use are likely unwarranted. For the vast majority of patients, we can be reassured there is no imaginary tightrope to walk when deciding how much fluid to administer in the early stages of sepsis. Clinicians should use their judgment to target fluid on the basis of rational clinical end points,⁹ and to stop administering unnecessary fluid when a clinical response is not seen.