The New England Journal of Medicine published an important trial today that aimed to answer the question: how much fluids should we give to patients with septic shock? Before we get into that result, let’s revisit the reasons we give fluids in the first place: to normalize low blood pressure, to normalize perfusion (blood flow) and oxygen delivery to body tissues, to restore organ function, and to clear abnormally elevated lactate. All these things sound good right? You would be forgiven for thinking that more of this would be better. Meyhoff et al. tested whether the opposite strategy – giving less fluids – would be better for patients. They conducted a multi-center randomized controlled trial of a restrictive fluid strategy for patients with septic shock. The trial involved a total of 1554 patients already admitted to the ICU, 770 of whom received the restrictive-fluid strategy and 784 who received the standard, larger, quantity of fluids. The restricted group in this study received less than half the quantity of fluids, 1798 ml on average compared to those getting usual care, at 3811 ml. The bottom line result: there was no difference in survival. The strategy of giving much less fluid after they arrived in the ICU did not improve or worsen outcomes in these patients.
In an accompanying NEJM editorial by McIntyre: “The amount of intravenous fluid that was received by the patients in the standard-care group in this trial was substantially less than that observed in previous national and international trials” of septic shock. McIntyre’s key take away point: “These findings show that a highly restrictive fluid management strategy is safe and raise important new questions that challenge conventional wisdom regarding the management of shock.”
My first reaction was that this result might validate the current practice of giving lots of IV fluids (typically these fluids are normal saline or lactated ringers, among others). After reading the accompanying article, another take away came into focus – there is no evidence to support the current practice of giving large volumes of fluids to these patients. In this report, less was the same as more. ICU and emergency physicians who prefer a restrictive strategy of giving less intravenous fluids to patients can find support from this study that this practice is safe.
Earlier work has shown worse outcomes in pediatric patients with septic shock receiving IV fluids. I wrote previously about the eye-popping results of the FEAST trial, performed in three sub-Saharan countries by a team led by Kathryn Maitland. That study measured survival in children receiving either local standard of care (no fluids) compared to children receiving intravenous fluids just like sick kids in North America and Europe do. The physicians involved in the FEAST trial were surprised to find higher mortality in children with shock who received intravenous fluids. A similar randomized controlled trial found higher mortality in adults with sepsis in Zambia who received intravenous fluids (along with vasopressors) compared to getting no fluids and no vasopressors (the usual therapy in that location). These are the only randomized controlled trials that I am aware of that compared giving fluids compared to not giving fluids. (It has not been possible to do a similar study in the United States, for example, because not giving fluids would be an unacceptable departure from the standard of care – physicians, institutional review boards, and probably many patients would reject such a study.)
However, a previous study suggested that giving less intravenous fluids may benefit sepsis patients by reducing the progression of kidney failure. The current study did not show such a benefit. However, it did not show any increase in kidney injury with fluid restriction (the trend was in the opposite direction) which might surprise some. Today’s NEJM paper suggests that we are not harming patients by giving them less fluids. It also does not answer the question of whether we should give less fluids in the emergency department. That is the scenario that interests me most as an emergency physician. Luckily, a research study of initial fluid management of sepsis patients is gathering data on that question.
Meanwhile, today’s NEJM study highlights questions about current Surviving Sepsis Campaign advice on fluids in sepsis. The current Surviving Sepsis bundle recommends 30 ml per kilogram of body weight (more than two liters of fluids for the typical adult) within 3 hours of arrival. We don’t know if giving 20 ml per kilo, or 10 ml per kilo or even no fluids is better during this initial phase of sepsis treatment. This current guideline is evidence-free, but we must report whether we meet the guideline as one of the performance goals promulgated by the United States Center for Medicare and Medicaid Services (CMS). In 2015, CMS instituted an all-or-none sepsis performance measure bundle (termed SEP-1) as a core quality of care measurement. SEP-1 is closely modeled on the Surviving Sepsis Campaign guidelines. If we don’t give a 30 ml per kilo fluid bolus in the first three hours for patients with sepsis criteria, our treatment is out of compliance.
Hospitals are required to report CMS quality measures including SEP-1. Because administrators are terrified of receiving negative feedback from CMS, and especially bad press for being out of compliance, attending emergency physicians and trainees receive the strong message that fluids are not just good for sepsis, they are necessary. I agree with critics of SEP-1 and critics of the Surviving Sepsis campaign that many of these performance measures are simply not evidence-based. Today’s NEJM report highlights the fact that we simply don’t know how much fluid, and when, we should give our patients with sepsis.
Moreover, the recent history of the Surviving Sepsis Campaign (SSC) is one of setbacks and medical reversals. In 2004, the SSC endorsed early goal directed protocol (EGDT) that was pioneered by Rivers et al. 2001. EGDT involved invasive monitoring of oxygen delivery, along with liberal use of the vasopressor dobutamine and red blood cell transfusions aimed at increasing oxygen delivery to shocked tissues. Dobutamine was a first line sepsis therapy in the 2004 SSC and 2012 SSC guidelines, but no longer, after being associated with worse sepsis outcomes. Enthusiasm for blood transfusions deflated after the TRISS trial showed higher mortality when we give more red blood cells to patients. Prior to 2016, the SSC recommended inserting a central venous catheter to measure and achieve a central venous pressure CVP of >8 mmHg and a central venous oxygen saturation (ScvO2) >70%. The ProMISe, ARISE, and ProCESS trials revealed that EGDT is ineffective, so we no longer measure central venous catheter pressure and central oxygen. Early SSC recommendations also included infusing recombinant human activated protein C (rhAPC) and giving corticosteroids per protocol, and normalizing blood glucose. Tight glucose control, routine corticosteroid use, and activated protein C have each been contradicted by subsequent trials, with activated protein C perhaps most disappointingly and most definitively. Read more about the story of activated protein C, trade name Xigris, and its evolutionary medicine implications in this previous post and podcast.
Many of the 2004 SCC Initial Resuscitation guidelines focused on Early Goal Directed Therapy and are no longer included:
• Obtaining central venous pressure (CVP) 8–12 mmHg
• A Central venous (superior vena cava) or mixed venous oxygen saturation of > 70%.
• Blood transfusion to achieve hematocrit > 30%
• Inotropic therapy – Dobutamine for CVP
• Activated Protein C (rhAPC, Xigis)
The 2021 Surviving Sepsis Campaign Initial Resuscitation treatment bundle is greatly diminished (in the number and intensity of interventions) compared to its predecessors. Current guidelines appropriately note that the quality of evidence for many of the remaining initial resuscitation guidelines is low, or very low (with early antibiotics as a notable exception). If history is our guide, we might expect that in future SSC guidelines, the 30 ml per kilogram IV fluid bolus mandate might go the way of activated protein C, blood transfusions, and central venous oxygen catheters.
What evolutionary lessons can we take from the incredible shrinking Surviving Sepsis treatment bundle? One lesson is that evolutionary medicine tends to favor a less is more approach, as I have written recently. Another lesson is that features that appear dysregulated, like low activated protein C or low levels of vitamin C in sepsis, may be maintained by natural selection. If so, a complex interplay of costs and benefits will accompany these traits. Further testing is needed, of course, but a key benefit of an evolutionary medicine perspective is that it generates novel and testable hypotheses.
Update: click here for Part 2 on the Incredible Shrinking Sepsis Guideline.
And for further reading, peruse my critique of how sepsis is defined: The Emperor has no Clothes? Searching for Dysregulation in Sepsis.
Copyright © Joe Alcock MD
Emergency Physician, Educator, Researcher, interested in the microbiome, evolution, and medicine