From Wiki Journal Club
Jump to: navigation, search
Semler MW, et al. "Balanced Crystalloids versus Saline in Critically Ill Adults". The New England Journal of Medicine. 2018. 378(9):829-839.
PubMedFull textPDF

Clinical Question

Among medical and surgical ICU patients, does the choice of fluid (normal saline versus a balanced crystalloid such as LR or Plasma-Lyte) impact rates of death, need for renal-replacement therapy, or persistent renal dysfunction?

Bottom Line

Among medical and surgical ICU patients, balanced crystalloids such as LR or Plasma-Lyte reduce the rate of death, need for renal-replacement therapy, or persistent renal dysfunction, when compared to normal saline.

Major Points

Normal saline has been associated with hyperchloremic metabolic acidosis and renal injury, prompting the hypothesis that isotonic crystalloids may offer inferior outcomes compared with balanced crystalloids such as lactated Ringer's solution (LR) and Plasma-Lyte. While preclinical models have suggested this hypothesis may be true, small studies in humans have failed to detect clinically significant differences.[1][2][3]

The Isotonic Solutions and Major Adverse Renal Events Trial in Medicine (SMART-MED) and Surgical patients (SMART-SURG) studies sought to address this question in a pragmatic, unblinded, multiple-crossover trial of approximately 15,000 critically ill adults admitted to medical and surgical ICUs. Most patients were admitted through the ED (50%), a subset had sepsis or septic shock (15%) or traumatic brain injury (9%), one-third required mechanical ventilation, and one-quarter required vasopressors. Patients were randomized to either normal saline or balanced crystalloids, with the primary outcome of major adverse kidney event within 30 days, defined as death, new renal-replacement therapy, or persistent renal dysfunction. The study demonstrated a small difference in rates of major adverse kidney events, which occurred in 15.4% patients of the normal saline group and 14.3% in the balanced crystalloids group (P=0.04). A subgroup analysis demonstrated the benefit of balanced crystalloids were primarily seen in medical ICU patients, those with sepsis, those without traumatic brain injury, and those with prior renal-replacement therapy. There were no statistically or clinically meaningful differences in secondary outcomes which included in-hospital death; ICU-, ventilator-, or vasopressor-free days; or rates of acute kidney injury.

The authors enumerate limitations of the study including its single-center design and non-blinded nature of treatment assignment. They conclude that although in percentage terms the between-group difference was relatively small, when scaled to the 5 million patients admitted to ICUs per year, the benefit of balanced crystalloids becomes substantial and meaningful. While questions regarding the generalizability of this study exist, it is a landmark paper supporting the utilization of balanced crystalloids. The concurrently published SALT-ED demonstrated similar results among non-ICU patients.


No guidelines have been published that reflect the results of this trial.


  • Pragmatic, unblinded, cluster-randomized on per-month basis, multiple-crossover trial
  • N=15,802 patients admitted to ICU
    • Balanced crystalloids (LR or Plasma-Lyte A, n=7,942)
    • Normal saline (0.9% sodium chloride, n=7,860)
  • Setting: Vanderbilt University Medical Center ICUs
  • Enrollment: 2015-2017
  • Follow-up: 30 days
  • Analysis: Intention-to-treat
  • Primary outcome: Composite of death, new renal-replacement therapy, or persistent renal dysfunction


Inclusion Criteria

  • Admission to a participating adult ICU at Vanderbilt University Medical Center
  • Eligible again if discharged and later readmitted to ICU

Exclusion Criteria

  • Age <18 years

Baseline Characteristics

Baseline characteristics are from the balanced crystalloid group.

  • Demographics: median age 58 years, female sex 42.8%, white race 80.4%, weight 80 kg
  • Coexisting renal conditions: CKD stage ≥3 (17.5%), prior RRT (4.8%)
  • Source of admission to ICU: ED (50.1%), OR (21.8%), ward (9.9%), outpatient (4.6%)
  • Diagnosis at admission: sepsis (14.7%), TBI (8.8%)
  • Mechanical ventilation: 34.3%
  • Vasopressors: 26.4%
  • Predicted in-hospital mortality risk (mean): 9.4%
  • Baseline creatinine (median): 0.89 mg/dL
  • Acute kidney injury: 8.6%


  • ICUs were randomized to use either normal saline on even-numbered months and balanced crystalloids (LR or Plasma-Lyte A) on odd-numbered months, or vice versa. Group assignments were non-blinded.
  • The choice of LR or Plasma-Lyte A in the balanced crystalloids group was at the discretion of the treating physician.
    • For patients with a relative contraindication to balanced crystalloid (eg, hyperkalemia or brain injury) or in whom clinicians felt saline was most appropriate, the treating physician was permitted to use saline.


Comparisons are balanced crystalloids vs. normal saline. OR adjusted for random effects (which ICU the patient is assigned to) and prespecified covariates (source of admission, primary diagnosis, mechanical ventilation, vasopressors, and baseline renal function).

Primary Outcome

Major adverse kidney event within 30 days
Death, new renal-replacement therapy, or persistent renal dysfunction
14.3% vs. 15.4% (OR 0.90, 95% CI 0.82-0.99, P=0.04)

Components of Primary Outcome

In-hospital death before 30 days
10.3% vs. 11.1% (OR 0.90, 95% CI 0.80-1.01, P=0.06)
Receipt of new renal replacement
2.5% vs. 2.9% (OR 0.84, 95% 0.68-1.02, P=0.08)
Final creatinine level ≥200% of baseline
6.4% vs. 6.6% (OR 0.96, 95% CI 0.84-1.11, P=0.60)

Secondary Outcomes

In-hospital death before ICU discharge
6.6% vs. 7.3% (OR 0.89, 95% CI 0.78-1.02, P=0.08)
In-hospital death before 60 days
11.7% vs. 12.4% (OR 0.92, 95% CI 0.83-1.02, P=0.13)
ICU-free days (median)
25.3 vs. 25.3 (OR 1.00; 95% CI 0.89-1.13, P=0.94)
Ventilator-free days (median)
28.0 vs. 28.0 (OR 1.06; 95% CI 0.97-1.16, P=0.22)
Vasopressor-free days (median)
28.0 vs. 28.0 (OR 1.05; 95% CI 0.97-1.14, P=0.26)
RRT-free days (median)
28.0 vs. 28.0 (OR 1.11; 95% CI 1.02-1.20, P=0.01)

Secondary Renal Outcomes

Stage 2 or higher AKI developing after enrollment
10.7% vs. 11.5% (OR 0.91; 95% CI 0.82-1.01, P=0.09)
Highest creatinine before discharge or day 30 (median, mg/dL)
0.99 vs. 0.99 (OR 1.01; 95% CI 0.97-1.05, P=0.58)
Change from baseline to highest creatinine value (median, mg/dL)
0.04 vs. 0.04 (OR 0.98; 95% CI 0.94-1.02, P=0.35)
Discharge creatinine or value at 30 days (median, mg/dL)
0.83 vs. 0.83 (OR 1.02; 95% CI 0.97-1.06, P=0.51)

Subgroup Analysis

Values are percentage of patients with the primary event. Comparisons are balanced crystalloids vs. normal saline.

Type of ICU (P-interaction=0.27)
Medical: 22.5% vs. 24.9% (OR 0.87, 95% CI 0.77-0.99, P=0.04)
Cardiac: 13.7% vs. 12.7% (OR 1.10, 95% CI 0.89-1.36, P=0.38)
Neurologic: 8.1% vs. 10.2% (OR 0.77, 95% CI 0.59-0.99, P=0.04)
Trauma: 8.0% % vs. 8.4% (OR 0.95, 95% CI 0.74-1.21, P=0.66)
Surgical: 11.4% vs. 12.2% (OR 0.93, 95% CI 0.66-1.30, P=0.66)
Sepsis (P-interaction=0.06)
No: 11.0% vs. 11.3% (OR 0.96, 95% CI 0.86-1.07, P=0.47)
Yes: 33.8% vs. 38.9% (OR 0.80, 95% CI 0.67-0.94 P=0.01)
Traumatic Brain Injury (P-interaction=0.24)
No: 14.3% vs. 15.5% (OR 0.89, 95% CI 0.81-0.98, P=0.01)
Yes: 15.0% vs. 14.0% (OR 1.09, 95% CI 0.81-1.47 P=0.58)
Categories of kidney function (P-interaction=0.19)
Normal: 8.5% vs. 9.2% (OR 0.91, 95% CI 0.80-1.04, P=0.16)
Acute kidney injury: 54.9% vs. 58.8% (OR 0.85, 95% CI 0.67-1.08 P=0.18)
Chronic kidney disease: 21.7% vs. 22.6% (OR 0.95, 95% CI 0.79-1.13 P=0.55)
Prior RRT: 12.2% vs. 18.4% (OR 0.61, 95% CI 0.41-0.91 P=0.01)

Adverse Events

No adverse events related explicitly to administration of fluids were reported.

Sensitivity Analyses

  • The authors conducted six prespecified sensitivity analyses. The OR varied between 0.87-0.93, although significance at the P<0.05 level was only obtained in the primary analysis (P=0.09-0.25 for all sensitivity models, page S10).
    1. Restriction of patients who received 500mL or more of isotonic crystalloid within 72 hours of enrollment
    2. Excluding patients admitted in the week preceding a crossover (attempting to exclude patients getting both fluid types)
    3. Excluding patients transferred to another ICU, or remaining in an ICU during a crossover
    4. Excluding readmitted patients
    5. Excluding those missing baseline creatinine values
    6. Alternative modeling


  • Concern about application to traumatic brain injury, where balanced crystalloid hypotonicity could adversely increase intracranial pressure
  • Single-center study limits generalizability
  • Non-blinded study
  • No separate analysis of LR vs. Plasma-Lyte A
  • Patients populations determined by hospital location, composite outcomes inherently not patient-centered and susceptible to confounding[4]
    • However, the authors state that the NIDDK recommends the adverse renal event outcome used this trial as a patient-centered outcome.[5]


  • Funded by the Vanderbilt Institute for Clinical and Translational Research
  • Primary author supported by grants from National Heart, Lung, Blood Institute

Further Reading

  1. Yunos NM et al. The biochemical effects of restricting chloride-rich fluids in intensive care. Crit. Care Med. 2011. 39:2419-24.
  2. Raghunathan K et al. Association between the choice of IV crystalloid and in-hospital mortality among critically ill adults with sepsis*. Crit. Care Med. 2014. 42:1585-91.
  3. Young P et al. Effect of a Buffered Crystalloid Solution vs Saline on Acute Kidney Injury Among Patients in the Intensive Care Unit: The SPLIT Randomized Clinical Trial. JAMA 2015. 314:1701-10.
  4. Myburgh J & Patient-Centered Outcomes and Resuscitation Fluids. N. Engl. J. Med. 2018. 378:862-863.
  5. Palevsky PM et al. Design of clinical trials in acute kidney injury: report from an NIDDK workshop on trial methodology. Clin J Am Soc Nephrol 2012. 7:844-50.