CC Nerd-The Case of the Unbalanced Solution

The use of 0.9% saline has undergone a degree of scrutiny in recent years. Detractors claim high chloride content, leads to acidosis and kidney injury. But clinical evidence supporting the clinically deleterious effects is lacking. The SPLIT Trial published in JAMA in 2015 by Young et (1) al failed to find any difference in outcomes for patients who received crystalloid infusions of either normal saline or Plasma-Lyte. The paper has been criticized due to its small sample size (only 2,278 patients), and the limited amount of fluids administered (median fluid in the saline group was only 1L). These factors limit the conclusions we can draw from the results. Our calls for a large RCT, possessing the statistical power to appropriately differentiate clinical signal from diagnostic noise were heard. Be careful what you wish for…

Published in the NEJM, two massive single center trials examining the utility of balanced solutions compared to normal saline. The first, by Self et al, the Saline against Lactated Ringer’s or Plasma-Lyte in the Emergency Department (SALT-ED) Trial (2), was a pragmatic, non-blinded, multiple-crossover trial that enrolled adults (?18 years old) who received at least 500 ml of intravenous isotonic crystalloids in the Emergency Department and were subsequently hospitalized outside an ICU. Treatment assessment was determined according to calendar month, with all patients in the trial Emergency Department during the same month assigned to the same fluid, either balanced crystalloids or saline. The initial month was selected at random and then fluid type alternated during each subsequent month. Clinicians had the option of ordering off-protocol crystalloids if they believed it was clinical necessary.

During the 16-month trial period the authors enrolled a total of 13,347 patients, including 6,708 (50.3%) assigned to receive balanced crystalloids and 6,639 (49.7%) assigned to the saline group. In general patients received between 1L and 2L during their stay in the Emergency Department. Lactated Ringer’s made up the vast majority (95.3%) of the fluid administered in the balanced solution group, and 88.3% of the patients received only the assigned crystalloid in the Emergency Department and did not receive any off protocol crystalloids. Finally, the volume of crystalloid that was administered and the adherence to crystalloid assignment were similar in the balanced-crystalloids and saline groups.

Not surprisingly, patients in the balanced crystalloids group had lower chloride and higher bicarbonate concentrations, and experienced less episodes of hyperchloremia or acidemia. From a clinical standpoint, the authors reported no difference in the primary outcome, hospital free days (median, 25 days in each group). They did note that patients in the balanced crystalloids group experienced a lower incidence of major adverse kidney events within 30 days (MAKE-30, defined as the composite of death, new receipt of renal-replacement therapy, or persistent renal dysfunction, with inpatient creatinine value ?200% of the baseline value), 4.7% vs. 5.6%. This 0.9% absolute difference was statistically significant with a p value of 0.01

The second trial, by Semler et al, the Isotonic Solutions and Major Adverse Renal Events Trial (SMART) (3), examined the use of balanced crystalloid solutions, only this time in patients admitted to the ICU. This was a pragmatic, non-blinded, cluster randomized, multiple-crossover trial, that was conducted in 5 ICUs at a single medical center over a 16-month period. The authors enrolled all adults (18 years of age or older) who were admitted to a participating ICU during the trial period. Participating ICUs were randomly assigned to use saline during even-numbered months and balanced crystalloids during odd-numbered months, or vice versa.

Similar to the previous trial, the authors noted a statistically significant difference in MAKE-30, the trial’s primary outcome,14.3% vs 15.4% in the balance solution and saline groups respectively. And like the SALT-ED Trial this 1.1% absolute difference reached statistical significance with a p-value of 0.04.

We now have two methodologically robust, large studies finding a statistically significant benefit for the use of balanced solutions in both critically ill patients admitted to the ICU and non-critically ill patients presenting to the Emergency Department. On first glance these trials appear to be two neat bookends to the story of normal saline’s slow demise, but before saline makes its final departure there are a number of caveats to discuss.

Both the SALT-ED and SMART Trials found their success in a composite outcome. Composite outcomes offer a number of benefits to trialists. First they allow for the combination of multiple event types into a single outcome, increasing the effect size and potentially augmenting the efficiency and statistical power of any given sample size. It provides an additional benefit to the indecisive trialist by combining multiple endpoints into one primary outcome without violating the singular nature of the primary endpoint . In this case the composite outcome was made up of death, new receipt of renal-replacement therapy, or persistent renal dysfunction (inpatient creatinine value ?200% of the baseline value). While composite endpoints augment statistical precision, they come at the cost of added clinical uncertainty. Each individual component of a composite outcome is weighted equally, placing equal value on events of varying degrees of clinical importance. In this case there is a dramatic difference in the clinical importance to a rise in creatinine of greater than 200% from baseline, a purely disease-oriented outcome, to death, an outcome of true patient-centered relevance. And so it is important to examine the individual components of a trial’s composite outcome, in order to assess the true clinical meaning of their results.

In the case of the SALT-ED Trial the authors found a 0.9% decrease in the rate of MAKE-30 in patients who received a balance crystalloid solution during their ED stay. This was a statistically significant outcome boasting a p-value of 0.01. But an examination of the individual components of this outcome certainly calls into question the clinical relevance of these results. There was no difference in 30-day mortality between the patients in the balanced solution and saline groups (1.4% and 1.5%, respectively). Nor was there any difference in the number of patients who required new renal replacement therapy (0.3% vs 0.5%). In fact, the entire difference was powered by a 0.7% increase in the rate of patients experiencing a 200% increase in the creatinine above baseline at 30-days. Essentially after enrolling 13,347 patients the only difference in outcomes the authors were able to identify, was a minute difference in the rate of elevations in creatinine levels above baseline.

An additional concern with the use of composite outcomes, is mistaking statistical noise for clinical signal. The argument for the use of a composite outcome is that the events being measured are rare. Sometimes prohibitively so, requiring an impossibly large sample to identify the true effect size. In these instances assembling an assortment of outcomes all representing a shared underlying physiological benefit, provide one the ability to identify small benefits of the treatment in question, that would otherwise go unnoticed. But if the composite outcome does not represent a unified physiological process, it is in danger of piecing together small random variations in effect size, augmenting statistical chance rather than identifying a clinical reality. In these cases, the composite outcomes serve only to provide statistical credibility to random noise.

In the case of the SMART Trial the authors again utilized the MAKE-30 composite outcome. And again they found a statistically significant decrease in the incidence of 30-day major kidney events in the patients who received a balanced crystalloid solution. Similar to the SALT-ED trial, none of the individual components that made up the MAKE-30 outcome were independently statistically different. Only when added together with an additional adjusted analysis, a p value less than 0.05 was achieved. 30-day mortality was 10.3% vs 11.1%, with 0.8% absolute difference. There was a 0.4% difference in new renal replacement therapy (2.5% vs 2.9% ). And only a 0.2% difference in the rate of patients experiencing a 200% increase in creatinine level at 30-days (6.4% vs 6.6). When the individual components of this outcome are examined one cannot help but notice that there does not appear to be any difference in the rates of kidney injury between the two groups. In fact, there was no difference in renal replacement free days, the incidence of stage 2 or higher AKI, highest creatinine level, or change in creatinine level above baseline between the two groups. And so if the MAKE-30 criteria were intended to identify the clinically significant deleterious effects of saline on renal function, then one has to ask the obvious question,

Where is the signal of renal dysfunction?

If there is no difference in the rates of AKI between the two groups, can we confidently state that combining these 3 metrics into one composite is an appropriate clinical measure?

The final answer is of course unclear. The results of the SALT-ED Trial confidently demonstrate the lack of benefit of balanced solutions in a non-critically ill cohort of Emergency Department patients. The findings of the SMART Trial may in fact represent a true clinical benefit offered by a balanced crystalloid strategy. There are some signals that this may be the case. Patients enrolled in the the trial with the diagnosis of sepsis appeared to do significantly better when randomized to received a balanced-solution. The rate of MAKE-30 was 33.8% vs 38.3% in the balanced and saline groups, respectively. Hopefully the publication of the PLUS trial, intended to enroll a sicker subset of ICU patients, will shed some light on these uncertainties. And while Semler et al have offered us statistically significant results, it is important we remember how they reached this answer, enrolling 15,802 patients in a non-blinded study, using a composite outcome, in which the individual components demonstrated minute differences of questionable clinical relevance. And even then, an adjusted analysis was required to demonstrate statistical significance, as the unadjusted comparison resulted in a non-significant p-value. In the end the real question to ask,

Do these results represent a true clinical effect, elucidated by a composite outcome that embodies the physiological benefits of balanced solutions, or is this just statistical noise stacked on top of random chance, piled on top of happenstance, that when assembled one on top of the other offers the false sense of clinical import?

Additional FOAM resources:

Josh Farkas in a great review tells you why my analysis is all wrong at PulmCrit
Well done detailed synopsis of both trials over at Rebel EM by Salim Rezaei

Sources:

Young P, Bailey M, Beasley R, Henderson S, Mackle D, McArthur C, McGuinness S, Mehrtens J, Myburgh J, Psirides A, Reddy S, Bellomo R, . Effect of a Buffered Crystalloid Solution vs Saline on Acute Kidney Injury Among Patients in the Intensive Care UnitThe SPLIT Randomized Clinical Trial. JAMA. 2015;314(16):1701–1710. doi:10.1001/jama.2015.12334
Self WH, Semler MW, Wanderer JP, et al. Balanced crystalloids versus saline in noncritically ill adults. N Engl J Med. DOI: 10.1056/NEJMoa1711586.
Semler MW, Self WH, Wanderer JP, et al. Balanced crystalloids versus saline in critically ill adults. N Engl J Med. DOI: 10.1056/NEJMoa1711584.