Are balanced crystalloids better than saline? SMART Talk with Dr. Michael Pinsky
Jon-Emile S. Kenny MD [@heart_lung] with illustrations by Carla M Canepa MD
“Each time new experiments are observed to agree with the predictions, the theory survives and our confidence in it is increased; but if ever new observation is found to disagree, we have to abandon or modify the theory.”
— Stephen Hawking
A 47 year old man with fever, right upper quadrant pain, jaundice, hypotension and evolving confusion is admitted from the emergency department for management of sepsis. He received roughly 50 mL/kg of 0.9% sodium chloride as initial fluid resuscitation followed by norephinephrine infusion via a large peripheral IV. Overnight he receives an emergent cholecystostomy from the on-call interventional radiology team yet his renal function progressively worsens. Estimation of his right atrial pressure via IVC ultrasound triggers many more litres of saline IV and in the morning his labs demonstrate a chloride of 118 mEq/L a bicarbonate of 12 mEq/L and an ascendant lactate of 5.5 mEq/L. He is completely anuric, hypoxemic, acidemic and on high flow nasal cannula to relieve his work of breathing. On morning rounds the senior ICU fellow collects her thoughts and readies the patient for invasive mechanical ventilation.
Background
With the 2016 Surviving Sepsis Guidelines mandating early and aggressive intravenous fluid resuscitation, the illustrative case above is a common clinical variety. Yet despite the more than 200 million litres of 0.9% saline administered per year in the United States, relatively little hard data underlies its unquestioned dominance atop the heap of clinically-available crystalloids.
While 0.9% saline is called ‘normal’ saline, it’s actually quite abnormal. 0.9% saline [henceforth called saline] is hypertonic and rather acidic. Additionally, there is data illustrating that the non-gap acidosis engendered by saline may lead to hyperkalemia, diminished renal perfusion and increased mortality.
There have been two relatively recent studies in the critically-ill population exhibiting no clear benefit to balanced solutions [e.g. lactated ringers] over saline – though there were trends favouring balanced solutions. With the above context, the SMART trial was recently published in the New England Journal of Medicine. In the same issue, the SALT-ED trial was also circulated – though this investigation studied non-critically-ill patients admitted from the ED to the non-ICU setting.
The discussion below will focus mostly on the SMART Trial and will be followed by a brief question-and-answer session from the renowned Dr. Michael Pinsky.
What They Did
SMART was a pragmatic, non-blinded, cluster-randomized, multiple-crossover trial, performed in 5 ICUs at Vanderbilt University Medical Center. The authors enrolled all patients 18 years of age or older who were admitted to the ICU. Participating ICUs were randomly assigned to use saline during even-numbered months and balanced crystalloids during odd-numbered months, or vice versa. The trial occurred over a 16-month period in an attempt to balance both seasonal and academic variation.
Patients could have their randomized IV fluids switched at the discretion of the treating clinician for concerns over hyperkalemia, brain injury or other reasons.
The primary outcome was a major adverse kidney event within 30 days — a composite of death from any cause, new renal-replacement therapy, or persistent renal dysfunction [the latter was defined as an elevation of the creatinine level to ≥200% of baseline] — all censored at hospital discharge or 30 days, whichever occurred first.
What They Found
Over the first 7 days in the ICU, the group randomized to balanced crystalloid received about 2 L of balanced solution [mostly lactated ringers] and about 500 mL of saline. By contrast, the group randomized to saline received a little more than 2 L of saline and roughly 250 mL of balanced solution.
There was a statistically and clinically-significant reduction in the primary end-point [absolute risk reduction of 1.1%] favouring balanced crystalloid; interestingly this reduction was driven mostly by improved mortality in the balanced group [absolute risk reduction of 0.8% with an odds ratio of 0.82 to 1.01].
During months assigned to balanced crystalloid, 12,227 of 13,085 isotonic crystalloid orders [93.4%] were for either lactated ringer’s or Plasma-Lyte. Of the 6.6% of those who were assigned to balanced fluids and who did not receive them, 2.8% [348 of 13,085] were for ‘hyperkalemia.’
Among those who required new renal replacement therapy, the difference between the groups was largely driven by oliguria [76% versus 82% for balanced versus saline]. There was no difference in refractory acidemia or hyperkalemia between the groups as indications for new renal replacement.
The average potassium for both groups was 4.1 mEq/L on admission to ICU and no difference between the groups in terms of hypo or hyperkalemia. Notably, the absence of potassium difference was present despite chloride levels being significantly higher and bicarbonate levels significantly lower in the saline group.
Thoughts
It is tempting to wield SMART like a scientific axe – using it to carve a deeper hollow into which one’s prejudiced view of crystalloid may be placed. Personally, I have always felt that balanced solutions made more sense, save for patients with severe hypochloremia in need of volume resuscitation. My opinions were based on previous data suggesting renal protection with balanced crystalloids. I was taken aback, however, by the clinically-small benefits noted in SMART – at least for renal-specific indices. Is it time to modify or abandon this theory?
Lactated ringers superior for kidneys?
Certainly, in favour of balanced solutions, the ~ 1% reduction in mortality seen in SMART follows the trend observed in both SPLIT and SALT which, like SMART, studied the critically-ill. In SPLIT, 87 of 1152 patients [7.6%] in the buffered crystalloid group and 95 of 1110 patients [8.6%] in the saline group died in the hospital; while not statistically-significant, it is certainly of clinical note. Further, as a run-in for SMART, the SALT trial demonstrated a 30 day mortality of 15% in those randomized to saline [n= 454] and 13.8% in those randomized to balanced solutions [n = 520]. While SPLIT looked at 90 day mortality, these three trials reveal, in totality, 9614 critically-ill patients randomized to balanced solutions and 9424 patients randomized to saline with 30 [or 90] day mortality rates of 10.2% and 11.0%, respectively.
Surprisingly, the differences in kidney injury and new renal replacement were underwhelming in all trials – including the non-critically-ill patients in SALT-ED. There were trends favouring balanced solutions, but the absolute differences were small. Nevertheless, given the large amount of fluid prescribed such small differences could translate into large population-level differences.
While clearly only hypothesis-generating, both SALT and SMART found large clinically and statistically-significant mortality reduction in septic patients who received balanced solutions. The mechanisms by which this occurred are obscure, but potentially mediated by protection from septic renal pathophysiology. Another possibility includes pro-inflammatory chloride effects which may alter the glycocalyx and lead to excessive interstitial fluid. Nevertheless, I recall the excitement for albumin in sepsis following sub-group analysis of the SAFE trial – then tempered by the ALBIOS trial. It seems that these crystalloid sub-group findings must be explored in more dedicated trials.
Lastly, concerns over hyperkalemia with balanced solutions are overblown. In fact, data exists that saline raises serum potassium levels as compared to balanced solutions – potentially secondary to the non-anion gap acidosis that follows excessive chloride administration. Indeed in SALT, potassium levels in the ICU trended much higher beyond day 4 in those randomized to saline. Nevertheless, in the much larger SMART there was essentially no difference in potassium levels between the two groups. In line with this data, neither the SALT-ED trial nor SPLIT trial saw significant dyskalemias between those randomized to balanced solutions versus saline.
Questions for Dr. Michael Pinsky
I remember you mentioning many years ago at a conference that you were/are not a fan of ‘normal’ saline. With this I am interested to know your general thoughts on the recent SMART and SALT-ED Trials?
Dr. Pinsky: The two trials were block step randomized designs. The SALT-ED gave very little fluid but actually saw the same threshold 1% increased mortality from normal saline over a balanced salt solution, even though the amount of fluid was often less than a liter with mean fluid given of 1.6L. These are very important studies, requiring many thousand patients in each group to pull out the effects. A 1% increase in mortality when extrapolated to all patients given fluids during hospitalizations becomes a very big number of lives we could save by simply using the right fluids. All fluids are drugs and thus will have both beneficial and detrimental effects. There is no one perfect solution for all conditions and all people, but as a default using a balanced salt solution with a lower chloride level as a first choice seems prudent.
There are data demonstrating that ‘normal’ saline worsens hyperkalemia compared to balanced crystalloids because of the acidosis that accompanies hyperchloremia. Nevertheless, hyperkalemia was considered a contraindication to balanced solutions in these trials? Do you agree? If so, what is your potassium threshold for withholding balanced solutions?
Dr. Pinsky: All fluids are drugs and if high potassium were an issue then giving a potassium free fluid would be reasonable. However, if one were to choose, hypokalemia is both more common and more detrimental to cardiac function. Indeed, we tend to run potassium levels above 4.5 mEq/L in heart failure patients to support contractility. If the potassium was above 5.5 mEq/L I would hold supra-physiologic potassium solutions and if >6, I would give no more potassium along with the usual treatments for hyperkalemia.
SMART found a significant reduction in sepsis mortality for those who received balanced solutions (~ 24% versus ~29%). Do you think that there is a difference between ‘normal’ saline and balanced solutions in their effects on the micro-circulation and/or gylcocaylx – especially in the septic milieu?
Dr. Pinsky: We do not know if balanced salt solutions have as much of a deleterious effect on the micro-circulation’s glycocalyx as does normal saline. However, we also do not know if that effect is actually due to the crystalloid solutions or to sepsis itself. Also we do not know what detrimental things such changes cause. So speculation here is premature.
Before you administer crystalloids to a patient who is deemed fluid responsive by functional monitoring, what do you use to assess whether or not the patient may, nevertheless, still be harmed by additional fluid administration? For example, should clinicians extrapolate the mean systemic filling pressure? Should a high CVP be used to warn the clinician about further fluids? Or the change in mean systemic filling pressure relative to the CVP?
Dr. Pinsky: The only reason to give fluids for resuscitation is to increase mean systemic pressure in a fluid responsive patient in circulatory shock. Since all fluids will increase mean systemic pressure to a certain degree, one really needs only measure CVP and not if it also increases the mean systemic filling pressure. If the CVP does rise, then the fluid may not be causing cardiac output to increase. High CVP values mean only RV failure, not that the patient does not need fluids. Still, one would also consider vasopressors to maintain right ventricular coronary blood flow if pulmonary hypertension were the cause of RV failure; and inotropes if it was a generalized heart failure plus or minus afterload reduction. If one has access to CVP, mean arterial pressure and cardiac output values then they can easily estimate mean systemic pressure from standard equations, though the clinical value in doing so is unclear.
Thanks for reading,
JE
Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he is also the creator and author of a free hemodynamic curriculum at heart-lung.org