2023 Top Story in Respiratory Medicine: Early Restrictive or Liberal Fluid Management for Sepsis-Induced Hypotension
According to the Centers for Disease Control and Prevention, 1.7 million adults in the US develop sepsis annually, and 270,000 die as a result. Sepsis accounts for a substantial proportion of deaths that occur in the hospital, and the overall incidence has been rising. Globally, the toll is substantial, affecting nearly 50 million people each year.1 Intravenous fluid resuscitation and vasopressor therapy are two treatments used in combination to combat sepsis-induced hypoperfusion.
Despite the ubiquitous use of fluids and vasopressors, to date, there has been a lack of definitive data to help guide frontline physicians in terms of specifics, such as how much and when to use these therapies, particularly early in the course of sepsis. Both fluids and vasopressors have potential drawbacks. Large-volume administration of fluids, for example, may lead to fluid overload, pulmonary edema, and dilutional coagulopathy. In one recent randomized controlled study in the ICU, patients with early septic shock (defined as development of shock within 12 hours of screening) were allocated to either a restrictive fluid strategy intervention or standard care. The difference between the average volumes received was approximately 1.5 L in the restrictive group compared with 3 L net positive in the standard care group over 5 days. There was no difference in 90-day mortality or other outcomes.2
The Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis (CLOVERS) trial builds upon this study by incorporating the use of vasopressors to more closely resemble the sepsis management that is already happening more holistically.3 Patients were enrolled earlier, within 4 hours of sepsis-induced hypotension once they had received at least 1 L of fluid. In the restrictive fluid group, the study investigators prioritized the use of vasopressors with rescue fluids administered in prespecified aliquots for indications outlined in the protocol. In contrast, the liberal fluid group received an initial 2 L of fluid, which was followed by additional boluses based on preset criteria, with rescue vasopressors employed for indications as outlined in the protocol. A total of 1563 patients were enrolled into CLOVERS, with about 20% of the patients in each intervention arm already on vasopressors at the time of enrollment. Protocol adherence was excellent, and the separation of groups in terms of fluid volumes received was meaningful. In the first 6 hours, the median fluids received for those in the restrictive group was 500 mL, as opposed to 2300 mL in the liberal group. By 24 hours, the difference was 1267 mL in the restrictive group versus 3400 mL in the liberal group. In the first 24 hours, 59% in the restrictive ("vasopressor first") group received vasopressors, as opposed to 37% in the liberal ("fluids first") group. Bottom line: there was neither a difference in the primary outcome of 90 day all-cause mortality nor in any of the secondary outcomes, such as need for renal replacement therapy, ICU length of stay, or ventilator days. For all subgroups examined, there was no difference in primary outcome, with the exception of a possible trend favoring the restrictive strategy among those with underlying end-stage renal disease.
The results of the CLOVERS trial have important implications for early sepsis care. This study supports that a more restrictive fluid resuscitation approach is noninferior to a more liberal strategy. Of note, it is possible that the structured approach to reassessment in both arms may have been ultimately beneficial and may have led to smaller differences in outcomes. There are important limitations and future considerations. The 90-day mortality rate was relatively low at around 14%, with only 59% of the patients even in the restrictive fluid group needing vasopressors. The differential impact for sicker patients is therefore unknown. This study does not answer how best to approach patients with late sepsis recognition or those who develop sepsis later in their hospital course (as CLOVERS excluded patients if more than 24 hours had lapsed since hospital presentation). Many patients hospitalized with chronic illness with varying degrees of pre-existing organ dysfunction and the presence of fluid overload may indeed have different and nuanced needs that are not answered by this trial. As a study generalizable to various practice settings, neither were dynamic assessments of fluid responsiveness included in the protocol, nor was assessment of sepsis-induced (or other) cardiac dysfunction required. Interestingly, 500 patients with early receipt of vasopressors had the medication administered peripherally. Adverse events related to extravasation were rare. While the study is not intended to examine the safety of peripheral vasopressors, the data support this increasingly common practice.
Click on any of these tags to subscribe to Topic Alerts. Once subscribed, you can get a single, daily email any time PracticeUpdate publishes content on the topics that interest you.
Visit your Preferences and Settings section to Manage All Topic Alerts
Additional Info
- Global sepsis burden is double previous estimates. Fogarty International Center. 2020;19(1). Accessed November 13, 2023.
- Meyhoff TS, Hjortrup PB, Wetterslev J, et al. Restriction of intravenous fluid in ICU patients with septic shock. N Engl J Med. 2022;386(26):2459-2470.
- Shapiro NI, Douglas IS, Brower RG, et al. Early restrictive or liberal fluid Management for sepsis-induced hypotension. N Engl J Med. 2023;388(6):499-510.
Disclosure statements are available on the authors' profiles: