Intravenous fluid administration is a staple of modern medical care. The advantage of crystalloid over colloid solutions is well established in the literature. However, few studies have examined differences in outcomes between “normal” saline solution (0.9% NaCl) and more “balanced” solutions such a Ringer’s Lactate or Plasma-Lyte A, which seek to more closely replicate “physiologic” conditions. The little literature that does exist is largely focused on the ICU setting. As such, there is little consensus on which fluid is superior in the resuscitation of the undifferentiated non-critically ill patient. The long history and tradition of the use of “normal” saline solutions have led to failure on the part of many providers to consider alternative options, particularly in light of the concern that saline solutions may precipitate a hyperchloremic metabolic acidosis, leading to adverse effects on the kidney.

Clinical Question

Does the use of balanced crystalloid solutions in the emergency department result in earlier hospital discharge or a lower incidence of major adverse kidney events in comparison to 0.9% saline?


Adults at least 18 years old who presented to one tertiary-care, academic, hospital-based emergency department between January 1, 2016, and April 30, 2017, received at least 500cc of an intravenous crystalloid solution in the emergency department, and were subsequently admitted to the hospital outside of an ICU.


Ringer’s Lactate or Plasma-Lyte A


0.9% saline


Primary: “Hospital-free days to day 28” (a composite outcome of in-hospital death and hospital length-of-stay, i.e. “days alive and out of the hospital between the index emergency department visit and 28 days later”)


  • Major adverse kidney events (MAKE) within 30 days
    • A composite of death, new renal-replacement therapy, or final serum creatinine >200% of baseline
  • Acute kidney injury (stage 2 or higher)
  • In-hospital death


  • Prospective, pragmatic, single-center, unblinded, cluster, multiple crossover trial
  • Paired with a sister-study (“SMART”) looking at patients admitted to an ICU


  • Patient younger than 18 years
  • Patients receiving less than 500cc crystalloid

Primary Results

  • 13,347 patients enrolled.
    • 6708 (50.3%) assigned to balanced crystalloids
    • 6639 (49.7%) assigned to saline
  • 88.3% of patients received exclusively the assigned crystalloid
  • Comparable demographic characteristics between populations
  • Per-protocol and intention-to-treat analyses had comparable findings
  • Robust, objective data collection, based on exclusively on data available in EMR
  • Critical Findings (Intention-to-Treat)
Solutions Median Hospital-Free Days to Day 28 Major Adverse Kidney Events (secondary) Acute Kidney Injury (secondary) In-Hospital Death (secondary)
0.9 % Saline 25 370 (5.6%) 560 (8.6%) 105 (1.6%)
Balanced Crystalloid 25 315 (4.7%) 528 (8.0%) 95 (1.4%)
Odds Ratio (95% CI) 0.98 (0.92-1.04) 0.82 (0.70-0.95) 0.91 (0.80-1.03) 0.88 (0.66-1.16)
Number Needed to Treat 111


  • Large, pragmatic study
  • Objective and robust follow-up
  • Very few exclusion criteria
  • Though providers could bypass study fluid choice, few did
  • Intention-to-treat analysis comparable to per-protocol analysis results
  • Pragmatic design suggests immediate clinical applicability
  • Large target population, so even a high NNT can have significant population level effects


  • Not blinded: Providers could order fluids outside of the protocol if they wanted to
  • Composite outcomes are, to a degree, inherently problematic
    • E.g. equal weight was given to in-hospital death and persistent AKI in calculating MAKE but these outcomes are not equivalent to the patient
  • Some outcome components (e.g. Creatinine levels) were indirect markers, less patient-centered
  • Only single-center
  • Almost all of the balanced solutions that were given was lactated ringers (95.3%). It’s unclear if these outcomes apply to Plasma-Lyte A
  • Study analyzed only a small amount of fluids given in ED
    • May have minimal impact
    • Subsequent inpatient providers could switch fluid choice

Author's Conclusions

“ Among noncritically ill adults treated with intravenous fluids in the emergency department, there was no difference in hospital-free days between treatment with balanced crystalloids and treatment with saline.”

Our Conclusions

In this study, the receipt of balanced crystalloids in the ED rather than saline had a modest effect on the composite secondary outcome of MAKE, but no effect on the primary composite outcome hospital-free days.

Potential Impact To Current Practice

Given the relative paucity of robust data on outcomes of crystalloid choice outside of the ICU setting, and given the comparable costs and minimal contraindications for either fluid choice, it would be reasonable, based on this study, to consider balanced crystalloids as the resuscitation fluid of choice in the undifferentiated emergency department patient.

Bottom Line

Saline and balanced crystalloid solutions are both reasonable choices for resuscitation fluids in the undifferentiated non-critically ill ED patient. Even a small reduction in the chance of adverse effects on the kidney may lead to significant population-wide benefits. As always, the need for more research remains, but paired with the results of the sister-trial “SMART,” the argument is growing for balanced crystalloid solutions as the “go-to” fluid in the ED.

Read More

PulmCrit: Get SMART – Nine Reasons to Quit Using Normal Saline for Resuscitation

REBEL EM: Is the Great Debate Between Balanced vs Unbalanced Crystalloids Finally Over?

REBEL EM: The SPLIT Trial – Saline vs Plasma-Lyte Fluid Therapy

Boring EM: Normal Saline: The Coke of Crystalloid Fluids