IVbags2 copyA 6-month-old male presents to the emergency department with diarrhea and vomiting. Despite antiemetic therapy, the the child is unable to tolerate oral intake in the ED and so you opt to admit him to the hospital for IV fluids.  The pediatric hospitalist requests that you write maintenance fluids prior to admission to the floor. Utilizing the 4-2-1 rule you calculate maintenance needs and choose D5 ½NS as your fluid. This is what you had been taught to utilize in children. It seems appropriate… but is it?

Background

Holliday and Segar published their seminal work on the maintenance caloric and fluid needs of children in Pediatrics in 1957 [1]. As the paper utilized hypotonic solution to match presumed solute needs, subsequent generations of emergency physicians and pediatricians have relied upon hypotonic solutions to serve as the primary vehicle for which to deliver caloric and electrolyte needs. The original calculations recommended 0.2% saline however this has largely been supplanted by 0.45% saline with dextrose as a primary intravenous maintenance fluid. Though D5 0.45% saline is chemically hypertonic, in vivo it is an effective hypotonic solution due to the rapid uptake and metabolism of dextrose. 

Recently there have been a few trials (reviewed in the systematic review we are discussing) that question the wisdom of using hypotonic solutions as maintenance fluid. It is theorized that hospitalized and critically ill children may have a non-osmotic stimulus for anti-diuretic hormone secretion potentially leading to hyponatremia and/or cerebral edema.    

Article Citation

Foster BA, Tom D, Hill V. Hypotonic versus Isotonic Fluids in Hospitalized Children: A Systematic Review and Meta-Analysis. J Pediatr. 2014 Feb 27. PMID:  24582105

Objective 

  • Systematic review of all studies comparing isotonic to hypotonic maintenance fluids in chilren assessing for hyponatremia

Study Methods 

  • Cochrane style systematic review in which a total of 10 studies met inclusion criteria and were included in the final analysis
  • 5 ICU studies, 4 ward studies, 1 mixed study
  • Patients had variety of illness (many were very sick) 
    • Large representation of PICU and post-operative patients 
  • Multiple different hypotonic fluids included across studies including 0.18%, 0.3%, and 0.45% saline
  • Primary outcome: hyponatremia (Na <135 mmol/L)
  • Secondary outcomes: 
    • Change in serum sodium from baseline
      • Moderate (<130 mml/L)
      • Severe (< 125 mmol/L)
    • Adverse events of hypernatremia (> 145 mmol/L)
    • Mortality

Results 

  • 11 RCTs included
  • Primary outcome
    • Relative risk for hyponatremia = 2.37 (1.72-3.26)
    • Assuming an estimated control event rate (CER) for hyponatremia of 5%, the Number Needed to Harm (NNH) = 15 (9-28)
    • Assuming an estimated CER for hyponatremia of 20%, the NNH = 4 (3-7)
      • The calculations of these NNHs are based upon the varying CER found in the various studies.
        • The control event rate describes how often an event in study occurs within the control group
        • To determine the NNH (as the NNT) we utilize the control event rate and the experimental event rate (EER—how often the event in study occurs in the treatment group).
          • NNH= 1/(EER-CER)
      • The authors utilized both the high and low end of the CER to give a range of NNH (4-15) with corresponding confidence intervals (3-28) depending upon the CER 
  • Secondary outcome
    • Change in serum sodium (5/11 studies described this statistic) = -2.46 (-3.11 to -1.81)
    • Mortality: none identified
    • Relative risk for hypernatremia (8/11 studies described this statistic) = 0.81 (0.32-2.04)
      • Reported about 0-6% incidence of hypernatremia using isotonic fluids
      • NNH not calculated due to nonsignificant findings 

Analysis

The studied population, that which the systematic review included, was heterogeneous and included disparate disease states lumping together floor patients admitted for various reasons with post-operative patients admitted to the PICU setting.  Though the underlying question of hyponatremia in the entire cohort may be equivalent (the I2 statistic did not demonstrate significant statistical heterogeneity) it may also be the case that sicker and post-operative patients have altered physiology from increased disease burden and represent the primary population in which ADH excess is triggered by non-osmotic stimuli (the actual at risk cohort).   

Due to the few studies included with routine pediatric EM admissions (e.g. dehydrated gastroenteritis) it is difficult to secondarily generalize these findings into the ED setting. It is also worth noting that there were no disease oriented outcomes delineated in either group from shifts in serum sodium concentrations. Though hyponatremia may predict subsequent neurological deterioration and cerebral edema, this systematic review did not find deleterious patient responses either because they do not occur or they are rare enough to not be found in the final analysis.  

Future Directions

This article forces us to reassess conventional wisdom in the light of new experimental evidence. Hypotonic maintenance fluids were originally established using a now 60 year old study on the basis of presumed rather than clinically confirmed patient physiology. While this particular systematic review failed to find patient oriented harm associated with hypotonic maintenance fluids it did show an absolute alterations in serum sodium potentially predictive of poor patient outcomes.

The next step will be to verify the study results and make it more applicable to our ED patient population. A prospective study of pediatric ED patients admitted for disease entities requiring maintenance fluids could be undertaken comparing the two intravenous fluid tonicities, using laboratory and clinically relevant outcome measures.  

References

  1. Holliday MA, Segar WE. The maintenance need for water in parenteral fluid therapy. Pediatrics. 1957 May;19(5):823-32. PMID: 13431307.
William Paolo, MD

William Paolo, MD

Residency Program Director
Assistant Professor of Emergency Medicine
SUNY-Upstate Medical Center