The SPLIT Trial: Saline vs Plasma-Lyte Fluid Therapy

There has been a lot of debate over the recent years about the safety of crystalloid fluid therapy in acutely ill patients. Several observational studies have shown an increased risk of acute kidney injury (AKI)  with the use of normal saline (NS). Other observational studies have shown a decreased risk of AKI when using a buffered solution (Hartmann’s solution, Plasma-Lyte (PL)). What is the best fluid to give to our patients who need fluid resuscitation? The answer to this question is not known, but another step in finding the answer was taken with the release of the SPLIT trial online by JAMA on October 7, 2015.

0.9% Saline vs Plasma-Lyte 148 (PL-148) for ICU fluid Therapy (SPLIT)


NS is the most common intravenous fluid used in fluid resuscitation, but it is associated with increased risks of AKI and mortality. The stated objective of this study was to determine the effect of a buffered crystalloid solution vs NS on renal complications in the intensive care unit (ICU).

What They Did:

  • Double blind, cluster randomized, double crossover trial in 4 ICUs in New Zealand.
  • Crossover done in 7-week blocks to allow each ICU to use each fluid twice over the 28-week study period.
  • Included all patients admitted to the ICU requiring crystalloid fluid therapy
  • Patients were excluded if: established AKI, plans for renal replacement therapy (RRT) in the following days, or patients already receiving RRT


  • Primary: Proportion of patients with AKI
  • Secondary:
    • Incidence of RRT
    • In-hospital mortality

What They Found:

  • 2278 patients eligible
    • 1152/1162 (99.1%) received buffered crystalloid
    • 1110/1116 (99.5%) received saline
  • Proportion of Patients with AKI within 90 Days
    • Buffered Crystalloid Group 102/1067 (9.6%)
    • Saline Group 94/1025 (9.2%)
    • Absolute Difference 0.4% (95% CI -2.1% – 2.9%)
    • RR 1.04 (95% CI 0.80 – 1.36; p = 0.77)
  • Incidence of RRT
    • Buffered Crystalloid Group 38/1152 (3.3%)
    • Saline Group 38/1110 (3.4%)
    • Absolute Difference -0.1% (95% CI -1.6% – 1.4%)
    • RR 0.96 (95% CI 0.62 – 1.50; p = 0.91)
  • In-Hospital Morality
    • Buffered Crystalloid Group 87/1152 (7.6%)
    • Saline group 95/1110 (8.6%)
    • Absolute Difference -1.0% (95% CI -3.3% – 1.2%)
    • RR 0.88 (95% CI 0.67 – 1.17; p = 0.40)


  • Physicians were blinded to type of fluid used (NS vs PL)
  • Pragmatic design in that treating clinicians determined the rate and frequency of fluid administration. The two groups had similar amounts of fluid administered (median=2 liters).
  • Statistical analysis plan was published before completing recruitment which helped eliminate analysis bias
  • Primary end point was based on a serum creatinine measurement which helped removed observer bias
  • 3% of patients eligible for study were analyzed which removed selection bias


  • Even though clinicians were blinded 2/3 were able to guess the assigned treatment. Maybe the development of hypercholoremia and metabolic acidosis in patients receiving saline may have unblinded some physicians
  • No sample size calculations performed. There are no study specifications for cluster-randomized trials; so all patients were enrolled during the specified time period for the study.
  • >90% of patients had received fluids, mostly PL, prior to study enrollment. This fluid was not considered in the analysis.
  • This was a heterogenous population with a low incidence of AKI, which makes it difficult to extrapolate findings to higher-risk groups
  • There was a low overall mortality in this patient population, therefore a clinically important increase or decrease in risk of in-hospital mortality could be missed.  The study included mostly elective surgery patients with relatively low APACHE scores.


  • The majority patients included in this study were patients admitted to the ICU following an elective surgery.  Of note many of these patients were not septic and did not have many comorbid conditions
  • Unanswered Questions:
    • What is the effect on all patients, namely those admitted from the emergency department?
    • Is there a difference in patients requiring more than 2 liters of fluid (e.g. sepsis)?

Author Conclusion: “Among patients receiving crystalloid fluid therapy in the ICU, use of a buffered crystalloid compared with saline did not reduce the risk of AKI. Further large randomized clinical trials are needed to assess efficacy in higher-risk populations and to measure clinical outcomes such as mortality.”

Clinical Take Home: Although no definitive answer is given to the question of which fluid to give for fluid resuscitation in the critically ill, at the moment it seems you are safe if you still want to use 0.9% saline or low-chloride crystalloid solutions at a volume of 2L or less. Remember the Goldilocks principle: not too much, not too little, but just the right amount. Fluids are a drug and we must monitor the effect they have on patients.


  1. Young P et al. Effect of a Buffered Crystallid Solution vs Saline on Acute Kidney Injury Among Patients in the Intensive Care Unit: The SPLIT Randomized Clinical Trial. JAMA 2015; E1 – E10. [epub ahead of print]
  2. Kellum JA et al. Assessing Toxicity of Intravenous Crystalloids in Critically Ill Patients. JAMA 2015; E1 – E3 [epub ahead of print]

For More Thoughts on This Topic Checkout:

Post Peer Reviewed By: Salim Rezaie (Twitter: @srrezaie)

Cite this article as: Matt Astin, "The SPLIT Trial: Saline vs Plasma-Lyte Fluid Therapy", REBEL EM blog, October 15, 2015. Available at:

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?