The Best Crystalloid for the Critically Ill

Aaron B. Holley, MD


February 08, 2022

Hemodynamic instability is rewarded with a sojourn in the intensive care unit (ICU). When the intensivist sees it, they're going to throw fluids at it. Most likely a crystalloid of some type. This has been true for decades, centuries even. When I was a medical student, which was decades but not centuries ago, I used crystalloids every day on the surgical wards, in the operating room, in the emergency department, or on the medicine wards. Medicine docs preferred normal saline (NS) and surgeons used lactated Ringer's (LR). I never gave this a second thought.

During medical school, I was drawn to internal medicine by the heavy emphasis on evidence-based medicine in the field. Prior to 2015 though, there wasn't much data to support using one crystalloid formulation over another. Pre-2010, we had an American Thoracic Society (ATS) consensus statement on using crystalloid vs colloid, making recommendations largely drawn from the SAFE trial. The ATS statement also suggested starches may be harmful, a view that was confirmed in a series of articles published in 2012 and 2013. There was less discussion about what type of crystalloid was best.

In 2014, I finally read a paper that compared crystalloid formulations. It was a network meta-analysis, which is "statistician speak" for combining disparate trials to make indirect comparisons. In the absence of large, randomized trials, this approach was a welcome addition to the data we had at the time. The authors concluded that "balanced" (typically LR or Plasma-Lyte) are superior to "unbalanced" (another term for NS) crystalloids. Balanced fluids typically have acetate or lactate and have a higher pH and lower chloride than NS. I found the signal for balanced fluids interesting at the time but promptly forgot about it.

Since 2015, the critical care community has rallied to produce a bevy of large trials comparing balanced vs unbalanced crystalloids. The first was the SPLIT trial, which showed equivalence. Then came the SMART trial in 2018, which showed balanced fluids were better. Of note, another trial with an identical design (SALT-ED) was published in the same issue of The New England Journal of Medicine as SMART. SALT-ED enrolled patients in the emergency department, not the ICU, but also found benefit to using balanced fluids, albeit not for their primary outcome. I admit, after SMART and SALT-ED were published, I made the switch to LR. A secondary analysis of patients with sepsis pushed me further toward LR, while others withheld judgement.

Then we saw publication of the BaSICs trial; another large, randomized study evaluating crystalloid composition. I was hoping this one might put the issue to rest. That nephrologist who perseverated on every patient's chloride during morning report would be vindicated. NS would prove to be too unbalanced and would finally be retired. No such luck. This is critical care medicine, where the initial signal is rarely confirmed in the follow-up trials. BaSICs found no difference between crystalloids for most important outcomes. They did find balanced fluids may worsen outcomes for patients with head injuries.

Finally, there's the PLUS trial, a large, multicenter randomized controlled trial comparing Plasma-Lyte vs NS in the ICU. I could make the argument that this trial was the best of the bunch, and it was negative. They did an excellent job of showing that serum pH and chloride levels did vary by fluid composition, but despite this, mortality and renal outcomes did not differ. Case closed? Crystalloid composition doesn't matter, right?

An editorial that accompanies the BaSICs trial does an outstanding job of reviewing SPLIT, SMART, and BaSICs. The authors discuss design and population differences that may have led to differing results, and there are many. They conclude for most patients in the ICU, there's no compelling reason to choose one crystalloid over another. Perhaps they're right.

An updated meta-analysis that included all the studies I've mentioned concluded there was an 89% probability that balanced fluid reduces mortality for ICU patients. How could the meta-analysis authors reach this conclusion given all the negative trials? It has to do with their statistical methods — they performed both standard, frequentist (if statistical significance isn't reached the study, is considered negative) and Bayesian analyses (posterior probability of benefit is calculated, regardless of P value). The frequentist approach was negative, but the posterior probability for benefit remained high.

Personally, I see no reason not to favor LR when resuscitating ICU patients without head injuries. In particular, it seems that medical patients (who made up almost 80% of those in the SMART trial) and those with sepsis may benefit. The critical care community has again outdone itself by performing large, well-designed trials to address important questions. Despite not having a definitive answer on crystalloid resuscitation, we know a lot more than we did when I was a medical student.

Aaron B. Holley, MD, is an associate professor of medicine at Uniformed Services University and program director of pulmonary and critical care medicine at Walter Reed National Military Medical Center. He covers a wide range of topics in pulmonary, critical care, and sleep medicine.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.