Balanced crystalloids versus saline in critically ill adults

Dr Bram Rochwerg, associate professor in the Division of Critical Care Medicine at McMaster University, and Dr James Douketis comment on the use of balanced multielectrolyte solution versus saline in critically ill adults.

For a Publications of the Week article discussing the use of balanced multielectrolyte solution versus saline in critically ill adults, click here.

James Douketis, MD: Hello everyone and welcome to another edition of the McMaster Textbook of [Internal] Medicine Paper of the Week. My name is Jim Douketis, I am an internist, and I am very pleased to be joined today by Dr Bram Rochwerg, who is a critical care specialist and associate professor of medicine and critical care medicine here, at McMaster University. Welcome, Bram, and I am delighted that you are able to join us today.

Bram Rochwerg, MD, MSc: Thanks, Jim, for the invitation.

James Douketis: It is our pleasure. We are going to be talking about the use of fluids in the critical care setting and focusing on the trial that is comparing the balanced multielectrolyte solution—let’s maybe call it BMES, if that is okay—with the standard, if you will, of saline. We will be talking a little bit about the results of the trial and what it means for our day-to-day practice. Perhaps I can start by asking you, Bram, what—kind of—came about that led to the development of these more sophisticated solutions? Can you put them in context versus the standard that myself and most physicians are aware of, which is the normal saline?

Bram Rochwerg: Thanks, Jim. This is a question that has gotten a lot of attention just in the last 5 or 10 years. I would say, before then people just felt like the big question was “colloids versus crystalloids.” But when it came to crystalloids, all the crystalloids are the same, whether you give saline or the Ringer [solution], or these newer options [that are now] available, it does not really matter. I think, as some of the colloid options, like starches, fell off of our menu of choices for patients who require fluid resuscitation, there has been a renewed interest in whether it matters which crystalloid we give.

I think, despite a lack of robust evidence, we have seen this transition in practice over the last 5 or 10 years—away from administering normal saline and towards using these so-called balanced fluids. “Balanced” really just refers to the fact that they use a lower chloride content, one that more closely matches serum, and they put it in an organic anion like lactate or acetate, or something else to replace that chloride. When we talk about a balanced [solution], there is the Ringer [solution], and then there is Plasma-Lyte, which is in fact what these investigators used. I know the title of the New England Journal of Medicine (NEJM) article is a mouthful, but it is essentially Plasma-Lyte, which is one of these designer balanced crystalloids. [It] looks very similar to the Ringer [solution]. Maybe some of it is that it is a little bit pricier and the companies can charge a little bit more for the products as compared with the Ringer [solution].

Despite the fact that we have not had robust evidence, most people have transitioned their practice with the thought that hyperchloremia is bad, it leads to respiratory acidosis... There are all these data in in vitro models and in vivo models [showing] that too high chloride contents can be harmful. And I think it has almost taken on this era of “I can’t believe those who are still using normal saline are still doing it.” And then it is in this context that we have had... I know today we are talking about the PLUS trial, but, Jim, there have actually been 2 large randomized [trials]—PLUS and BASICS (which is [carried out] by the Brazilian group)—published within the last 6 months, addressing this question, looking at thousands of critically ill patients, randomizing them to 2 different fluid strategies, trying to say “All right, despite the fact that folks think that balanced crystalloid is better, does it actually bear out when we look at it in large-scale randomized controlled trials (RCTs)?”

James Douketis: Right. That is a terrific background. Let’s now go to the trial itself. I mean, the results were essentially neutral in terms of the mortality outcome, the outcome of kidney replacement need… I think one of the justifications they used was that maybe the balanced solution would mitigate against kidney injury. Yet, the trial was neutral. How do you interpret these results? Were there any major weaknesses of the trial? Could it have been a false-negative result? What is your overview of the trial?

Bram Rochwerg: Yeah, you are exactly right. I think both trials, in fact, Jim, PLUS and BASICS (the Brazilian trial, which was a little bit larger than this one and published, again, ~6 months ago), showed no difference in the primary outcome of mortality between the 2 fluids, and similarly in secondary outcomes—acute kidney injury, duration of life support, this sort of stuff—no difference. I think the critical care community was a bit shocked by this. I think we all… many have transitioned their practice even before these studies and now [it is] shocking to say that maybe it does not make a difference. I think it is exceedingly hard to show slightly different electrolyte contents in these different fluids to show an effect on mortality. Maybe that was too high a bar to try to show a difference in the first place.

I think that there is a couple of other key features with these trials that limit our ability to show a difference if one exists. It is really hard to capture these patients early on in RCTs. As we both know, the majority of fluid resuscitation happens early on. A lot of these patients actually got a lot of an open-label fluid prior to randomization, which might have diluted further down the difference between the fluids. Also, these patients, although critically ill in both trials, were relatively healthy critically ill patients, if that makes sense. The severity of illness scores were low, [there were] a lot of postoperative admissions, and the volume of the administered study fluid was only in the range of a couple of liters. So, if you take critically ill patients who are not that sick to start with, do not give them a ton of study fluid, mix in a lot of open-label crystalloid administration, perhaps it is not a big surprise that we do not see a difference in mortality in the long run. These are some of the ideas that folks have brought up.

That being said, Jim, if you look at a lot of the outcomes in both trials, there certainly appears to be a trend towards benefit with the balanced crystalloids—not statistically significant. When you look at both trials, you look at all the secondary outcomes, it seems (...) that maybe there could be a benefit with using a balanced crystalloid. Folks have now proceeded with doing some of these fancy Bayesian analyses, looking at the probability of benefit, and found that—based on posterior and prior probabilities—maybe balanced crystalloids are most optimal. The criticism balanced crystalloids have always got was that they are more expensive. It might still be the case for Plasma-Lyte, which was studied in these RCTs. The Ringer [solution], which looks like Plasma-Lyte, is much cheaper.

I still think it leaves us in a difficult spot insomuch that [it] probably does not matter, which fluid you give to patients for resuscitation when it comes to crystalloids. But that being said, I think it would be hard pressed to sort of... If you had equal access to both [fluids] and the cost [of both fluids] was [similar], [you would probably be] still inclined to use a balanced crystalloid.

I should say—and maybe just to end on this—I mentioned to you before [that] I am doing an RCT myself, looking at the Ringer [solution] versus saline. And the key in our RCT is the Ringer [solution], which is the more widely available balanced crystalloid that is out there. And in our trial we are trying to get around some of these limitations insomuch that we are only looking at septic shock patients. So, a very high severity of illness. And [in] patients in the FISSH trial, which is the name we call our trial, they have on average gotten 6 or 7 liters of the study fluid, with the mindset here being that if you want to show a difference, take a sick group, give them lots of the study fluid, try to avoid that upfront contamination, and if you are going to see a difference, I think it is in this population. We are still recruiting and maybe the subject of one of these talks in a year or two will be to discuss the FISSH trial results.

James Douketis: The message I am getting is that the jury is still out with this, but it seems that it is very sensible if you want to use a balanced solution, to do so. And the author said [that] maybe [it should be used] in certain patient groups, like the ones that you described—who are sicker, maybe some who have hyperchloremic metabolic acidosis for whatever underlying cause. The other issue is [the one] of the upfront fluids... Because in this trial they only got ~4 liters of the study fluid, and I am sure they got a whole bunch more before, so, [we get] back to our point, [whether] we could have had a false-negative result.

At the end of the day, are there patients in whom you would definitely go to balanced [crystalloids] and [the ones in whom] you would definitely go to just saline? Maybe we can end with that?

Bram Rochwerg: Definitely the one that I feel most strongly about would be your traumatic brain injury (TBI) population. And both trials did enroll patients with TBI and showed harm with a balanced crystalloid and the fact that they are slightly hypo-osmolar compared with saline. So, if you have a population with TBI, the answer is still saline. And I feel comfortable telling you that.

The rest of the critically ill population—again, I think it is a toss-up. I think you should feel comfortable using either. It is not like we are in a position that if somebody uses saline, we can call them up and be like “I can’t believe you gave this patient saline, what were you thinking?” And probably the same with balanced [crystalloids]. As I said, I think that probably, on the balance of odds, maybe giving balanced [crystalloids] makes the most sense, but I think equipoise does persist. It is interesting if you look... There was a meta-analysis that has been published now including BASICS and PLUS. It was published in that new NEJM Evidence [monthly]. It is one of these new publications to come from the New England Journal Group. They provided a subgroup analysis in the meta-analysis and if there is one group that might have a signal towards benefit—the upper end of the confidence interval just crossed 1 in terms of benefit for a balanced crystalloid—it was in the sepsis subgroup. My inclination, maybe even more so in the sepsis group, would be to give a balanced crystalloid. But again, I think we need further data before we can say that strongly.

James Douketis: Bram, thank you very much for these insights and the clarity in which you presented them. I think they have been very helpful. I will end with... I wish you the best of luck and we look forward to maybe having another Paper of the Week with Dr Rochwerg’s trial. Best of luck to you and your research team.

Bram Rochwerg: Thanks again for the invitation. My pleasure.

James Douketis: Thank you again.