The Effect of Early Fluid Resuscitation on Mortality in Sepsis: A Review of a Systematic Review (The Dude Abides)

To the readers who can conjure an original thought without drifting off into comedies from the 1990s, I hope this review retains some useful coherence. To everyone else, I am confident that a few well-placed quotes from the 1998 classic, The Big Lebowski, will help to “tie” the literature together on this topic...”there is no literal connection, Dude!” The who, why, and volume of fluids administered to septic patients remain topics of considerable debate. The current Surviving Sepsis campaign guidelines weakly recommend 30 mL/kg of crystalloids within 3 hours of sepsis-induced hypoperfusion, based only on low-quality evidence (think Creedence tapes).1 And, to the chagrin of some, these guidelines became a U.S. mandate. “This aggression will not stand, man!” An American College of Emergency Physicians (ACEP) policy statement agrees that many will benefit from 30 mL/kg—BUT for all septic patients?! “Dios mio, man!” In the true parlance of the college, “We do not support a prespecified volume of fluid for all patients...administration of an initial volume of 500-1,000 mL of crystalloid is a common and reasonable practice” and note that further volume or vasopressor support should then be guided by assessments of fluid status. ACEP backs this statement with a collection of observational, ICU-based studies highlighting the harms of positive fluid balances days out from sepsis onset, but lacking well-resourced emergency department (ED)-based literature examining the early stages of resuscitation.2 “Donny, you're out of your element!” CLOVERS was supposed to be our “sweet prince” and answer this vexing question by comparing a fluid-restricted, early vasopressor strategy to liberal fluid resuscitation.3 Well, “there is a problem, Dude.” The fluid-restricted group likely received >30 mL/kg and no outcome differences were found, except for vasopressor administration. Perhaps this supports the case of the medical nihilists. “We believe in nothing, Lebowski. Ja, nas-zing.” So, is a global prespecified fluid minimum, like 30 mL/kg (may not be reached in >50% cases), ensuring more patients are receiving vital volume replacement that may not be otherwise prioritized because bedside assessments are not realistically feasible or reliable early in current sepsis care?4 Several mechanisms may support this theory, including common prodromal symptoms contributing to fluid deficits (vomiting/diarrhea, reduced intake, and high insensible losses), vasodilatation leading to reduced stressed volume and cardiac output, and microvascular dysfunction, which may improve with early fluids. Or does 30 mL/kg too often exceed the needed fluid volume causing deleterious issues of fluid overload where more tailored resuscitative strategies, such as inclusion of earlier vasopressors, might be a safer alternative? Sepsis-induced glycocalyx perturbations may cause pathologic extravasation of infused fluids, which is one proposed mechanism for the deleterious impacts of increased fluids found in some low-resource cohorts.5,6 To help answer these questions, “the Royal we,” Dr Hani Kuttab and Dr Bob Badgett, went hunting for “leads.” They had us “working in shifts” except for the Shabbos. As such, we recently published The Effect of Early Fluid Resuscitation on Mortality Sepsis: A Systematic Review and Meta-Analysis that included 119,583 patients from 30 studies assessing fluid dosing and timing within 8 hours of sepsis diagnosis.7 From this we derived the following conclusions (summarized in Table 1): Table 1. - The quality of assessment and summary of findings with respect to the four conclusions derived from this review. The quality assessment was performed using the GRADE system. Further details are provided in the full publication outlining the factors used to adjudicate the five domains (limitations, inconsistency, indirectness, imprecision, and publication bias) that may decrease the certainty of evidence. Certainty assessment Summary of findings based on mortality # studies (design) Risk of bias Inconsistency Indirectness Imprecision Publication bias Number of patients or studies categorized by statistical significance∗ Effect∗ Certainty of evidence Comments Study question 1: Mortality difference between more liberal & more restrictive fluid strategies More liberal More restricted Relative risk (CI) 3 (RCT) Not serious Not serious Serious Serious Undetected Mortality: 200 of 1274 patients Mortality: 199 of 1286 patients RR 1.00 (0.81-1.24) ⊕⊕ Low ↓ Differing volume ranges assessed between studies ↓ few studies included with one low n-size Take away: Evidence suggests No mortality impact Study question 2: Mortality impact of receiving low-volume fluid dosing (45 mL/kg) High fluids Benefit High fluids Risk High fluids No impact Effect direction/sign test 2 (RCT), 9 (observational) Serious Not serious Not serious Serious Undetected 0 studies 6 obs. studies 2 RCT, 3 obs. studies High fluids with increased mortality effect; 6 of 11 studies, P = 0.55 ⊕ Very low Many obs. studies begin low certainty ↓ Poor study control in cases ↓ Mixed results Take away: no mortality effect but evidence very uncertain Study question 4: Earlier timing (