Trick of the Trade: Lactated Ringers for Sepsis Complicated by Hyponatremia

Sodium Na canstockphoto12825701An 82-year-old female is brought into the Emergency Department by family for a several day history of progressive altered mental status. You initiate a broad workup. However, soon after initial evaluation, you are called back into the room. The patient’s vitals are as follows and concerning for septic shock and an alarming serum sodium level.

Vital signs:

  • Blood pressure 89/65 (MAP 55)
  • Heart rate 105 bpm
  • Respiratory rate 18/min
  • Temperature 38.3C

Initial labs:

  • Serum sodium 105 mmol/L
  • Lactate 4 mmol/L

What is the safest way to volume resuscitate a hyponatremic, hypovolemic patient?

Likely the first instinct is to provide an isotonic fluid bolus of 20-30 mL/kg for septic shock. However, in chronic, severe, symptomatic hyponatremia, sodium correction should not exceed 0.5-1 mEq/L/h with a goal of 8-12 mEq/L/d to reduce the risk of central pontine myelinolysis (also known as osmotic demyelination syndrome).1,2 Acutely symptomatic (<48 hours) patients can be corrected safely at a faster rate with goal of increasing sodium by approximately 1-2 mEq/L/hr for 3-4 hours.

This is a nuanced case of fluid management, but can have long-term implications for an acutely ill patient. The ideal fluid choice would allow for adequate fluid resuscitation without rapid overcorrection of the sodium concentration.

Options for fluid resuscitation

Fluid Na+
(mEq/L)
Cl-
(mEq/L)
K+
(mEq/L)
Ca++
(mEq/L)
Mg++
(mEq/L)
Buffer pH Osmolality
(mOsm/L)
0.9% NaCl 154 154 5.7 308
Lactated Ringers 130 109 4 3 Lactate
(28)
6.4 273
3% saline 513 513 5.0 1026

 

Trick of the Trade: Lactated Ringers may be the safest choice

Compared to normal saline, Lactated Ringers provides twice the volume expansion for the same degree of correction of sodium due to the lower sodium concentration per liter.

This was calculated using MDCalc, which was derived based on a great NEJM review paper.3 For example, the amount of fluid needed to increase the serum sodium by 1 mmol/L/hr for an elderly female with an estimated weight of 60 kg and an initial sodium concentration of 105 mmol/L:

Lactated Ringers: 1240 mL/hr
0.9% Normal Saline: 633 mL/hr
3% Saline: 76 mL/hr

 

MDcalc Na LR

 

The exact amount of volume that can be given for an expected 1 mmol/L/hr increase in serum sodium will change depending on your patient’s age, sex, weight, and starting sodium value and should be calculated for each individual patient.

Bottom Line

Be concerned about rapid sodium correction in a patient with symptomatic hyponatremia. Twice the amount of volume resuscitation with Lactated Ringers can be given compared to normal saline for the same degree of sodium correction.

 

(c) Can Stock PPhoto

1.
Sterns R, Cappuccio J, Silver S, Cohen E. Neurologic sequelae after treatment of severe hyponatremia: a multicenter perspective. J Am Soc Nephrol. 1994;4(8):1522-1530. [PubMed]
2.
Vachharajani T, Zaman F, Abreo K. Hyponatremia in critically ill patients. J Intensive Care Med. 2003;18(1):3-8. [PubMed]
3.
Adrogué H, Madias N. Hyponatremia. N Engl J Med. 2000;342(21):1581-1589. [PubMed]

ALiEM Copyedit

>

  • Nice discussion of fluid resuscitation in hyponatremia. Might want to mention this is mainly a concern in chronic, severe hyponatremia and that acute hyponatremia is thought to be safely corrected at a faster rate. Also, Na of 105 is extremely rare (1 in 1,000 patients with hyponatremia in community sample: Clin J Am Soc Nephrol. 2011 May; 6(5): 960–965). Might be helpful to remind people that the same lesson (can give way more LR than NS) is also true if her Na is 120. This broadens the applicability of the case.
  • Lactate is listed as a vital sign, but it\'s a laboratory value.
  • Most people will get a sodium level off a chemistry, not a VBG. Specifying VBG doesn\'t seem to add anything to the case. Consider deleting. Also may remove () around units.
  • Repeatedly specify LR is \"best\" choice but that\'s a subjective statement. Realistically it may be the \"safest\" choice in terms of reducing risk of demyelination.
  • Nice reference to online calculator, but MDCalc references JAMA Adrogué HJ, Madias NE. Hyponatremia. NEJM, 2000. for this. Would be more appropriate to cite the initial source and mention the MDCalc calculator as a helpful reference.
  • This concept of LR having only slightly less sodium but being able to give twice as much volume may be hard for some people to initially grasp. Would a picture help people understand this better?
  • You specify that numbers should be calculated for each patient but also seem to say that LR is a better resuscitation fluid for any hyponatremic patient at risk of rapid Na correction. Isn\'t it always better to give LR to hyponatremic patients? If so, isn\'t calculating for every patient unnecessary? If not, why not?
  • Matthew Zuckerman, MD
    Assistant Professor, University of Colorado School of Medicine, Department of Emergency Medicine, Medical Toxicology Section; Creator and host of ToxNow.org

    Expert Peer Review

    The clinical condition of an undifferentiated hyponatremic AND hypotensive patient poses a rare but significant clinical dilemma. While Early Goal Directed Therapy (EGDT) demands rapid correction of hypovolemia when sepsis is suspected, there exists a significant risk of cerebral demyelination from overly aggressive sodium correction.

    Remarkably, based on the MDCalc application (derived from Adrogué HJ, Madias NE. Hyponatremia. NEJM, 2000.), seemingly mild differences in the sodium and potassium content of Ringer’s lactate, allows for infusion of much larger crystalloid volumes compared to Normal Saline for the same degree of serum sodium increase. In this case, a “typical” 2 liter bolus of Normal Saline would be expected to increase the serum sodium concentration almost 2.5 mmol/L which is already about a quarter of the recommended daily sodium change in hyponatremia. Ringer’s Lactate (or other “balanced” crystalloid solutions such as Plasma-Lyte) would be expected to cause HALF of this effect on serum sodium.

    While some might argue that the sodium increase from a single 2 liter bolus of Normal Saline has minimal significance a hyponatremic patient’s immediate morbidity and mortality, I would argue that there are still compelling reasons to change practice. For one, as described, Ringer’s Lactate is a clearly superior crystalloid compared to Normal Saline in these circumstances. Two, we ought to be mindful of how our decisions in the ED have downstream effects. In such a complicated patient, we should attempt to optimize the conditions for success of our ICU colleagues and the patient over the ensuring hours and days of care. It is not hard to imagine the massive deliberation needed to balance the needs of multiple continuous IV infusions, volume resuscitation, and careful sodium titration in this patient. Why make it more difficult for our ICU colleagues? Finally, it is really in these critically-ill patients, where single, sometimes seemingly minor decisions, can have major impacts on their ultimate morbidity and mortality.

    As an aside, for me, the most striking aspect of this trick of the trade is how it exposes the danger of using automated behaviors (a form of anchoring) irrespective of individual patient conditions. Rapid volume resuscitation is an EGDT cornerstone, which has been ingrained as gospel in medical trainees, and as strict policy at many emergency departments/ intensive care units. Additionally, use of Normal Saline as the crystalloid of first choice is largely the result of convenience and routine, rather than superiority to other crystalloid formulations. For me, this trick of the trade teaches me that I shouldn’t always use a 2 Liter bolus of Normal Saline just because it’s on my medication order “quick list.”

    In summary, it is ideal to use Ringer’s lactate (or other “balanced” crystalloid solutions) as the initial choice for volume resuscitation in the undifferentiated hyponatremic, hypotensive patient, until more definitive diagnosis and disease-centric treatments can be initiated.

    Derrick Lung, MD MPH
    Assistant Clinical Professor, Department of Emergency Medicine, UCSF-San Francisco General Hospital; Assistant Medical Director, California Poison Control System – San Francisco Division

    Expert Peer Review

    Dr. Li,

    Thank you for writing up this pearl. Both of the reviewers bring up great points.

    1. I agree with Dr. Zuckerman\'s point #4 as I am unable to find an outcome specific study in fluid resus between the usual standard care (NS 20-30 ml/kg bolus vs LR) with patient-specific safety outcomes. Based on the math, it theoretically should be a safer choice for a carefully selected cohort of patients. Please consider stating \"may be the safest choice,\" rather than \"[it] is the safest choice.\"
    2. The second small change is to be consistent with your units. (cc/kg -->ml/kg).
    3. From an emergency department perspective, differentiating acute vs chronic hyponatremia can be very challenging. Perhaps majority of our patients will have chronic hyponatremia, and in cases where the onset period is unknown (and there are no life-threatening symptoms), it may be best to assume chronic hyponatremia, rather than treat aggressively with NS or higher osmolality fluids. Regardless of the strategy, it would be wise to recommend close and frequent monitoring of Na (every 2 to 4 hours), starting in the emergency department while these patients continue to wait for a bed. Repeat Na can be easily overlooked during a busy shift.

    Thank you again for this great pearl!

    Zlatan Coralic, PharmD
    Assistant Clinical Professor of Pharmacy & EM; Emergency Department Clinical Pharmacist; University of California, San Francisco
    Kai Li, MD

    Kai Li, MD

    Toxicology Fellow
    Department of Emergency Medicine
    University of California, San Francisco
    • R. Garrett Hanzel

      Interesting. How about hyponatremia plus rhabdo? I had a Na of 100 with CPK of 10k a few weeks ago, which presents the same dilemma of course.

      • Similar scenario of needing large volume fluid resuscitation while slowly correcting low sodium levels.

    • We should not be defaulting to saline for resuscitation in sepsis anyway. Far too much strong ion (chloride) without buffer, and associations with additional kidney injury (http://www.ncbi.nlm.nih.gov/pubmed/23073953) and mortality (http://www.ncbi.nlm.nih.gov/pubmed/24674927).

    • andrew

      Metabolic Theory of Septic Shock
      Please do a search for the above

    • andrew

      Metabolic Theory of Septic Shock
      Please check it out by doing a search

    • Mark Culver

      When using LR for sepsis fluid resuscitation – be aware that certain antibiotics are not compatible with LR (ceftriaxone, imipenem, meropenem, ampicillin, etc.)… Big fan of LR resuscitations for the most part!

      • Mohammed Abbasi

        That’s interesting – what’s the evidence or reasoning behind this?

        • Adam M. Spaulding

          The ceftriaxone moiety and the calcium contained in LR can form a precipitate in the blood. Similar with the other agents. I would also add amphotericin.

    • Joel Topf

      Any estimate of the risk of osmotic demyelination syndrome if you exceed the guidelines on correction?

      • Great question Joel. Obviously the calculator tool provides only a rough estimate of the change in sodium, since you have to factor in intravascular volume and tonicity. Also depends on rate of Na decline for that patient. ODS risk is greater for those with a slow gradual decline over days. Would definitely check serial serum sodium levels more frequently (q2-4 hrs for example) at the beginning of the resuscitation to trend Na levels.

    • Joel Topf

      Another issue that is not addressed is assuming the same degree of volume resuscitation with LR and NS, while that is normally true, I suspect that with the degree of hyponatremia you describe, it would not be the case. Since NS is highly hypertonic it would draw water from the intracellular compartment fortifying it’s volume resuscitation potential.

    • Joel Topf

      Also you should be aware that when the Androgue formula was tested it dramatically under-estimted the change in sodium: http://cjasn.asnjournals.org/content/2/6/1110.full.pdf

      • Thanks for sharing this article. Hadn’t heard of this. Makes me want to check serial Na’s more often in the acute setting! Hmm, I learn something new every day!

    • robert donnell

      As alluded to in one of the other comments the calculation only gives a rough estimate. It must be kept in mind that such a patient probably has increased vasopressin levels driven by non osmotic (volume) stimuli. If you replete the volume and turn that signal off a water diuresis may ensue and drive the sodium up way faster than the formula would predict.

      • Jason Schwaber

        What role should DDAVP play in the resuscitation of the hyponatremic hypovolemic septic patient?

    • Mohammed Abbasi

      You have to be careful though if pt has component of SIADH, and if the urine is very concentrated giving a fluid that’s hypotonic in comparison to what the kidneys are putting out will worsen the hyponatremia as kidneys will continue to resorb more water and hence concentrate the urine.