In Sepsis, Fluid Choice Matters


During a large-volume sepsis resuscitation, your choice of fluids – specially which crystalloid solution – could mean the difference between life, death and dialysis

Included in the emergency physician’s skill set is their ability to resuscitate critically ill patients; an example of this is the emergency department care of the septic patient. Given the general delay in translating medical knowledge to the bedside, it’s remarkable to see the vast change in the management of these patients since Dr. Rivers published his ground-breaking paper [1]. In a relatively short period of time, we’ve made aggressive fluid resuscitation and early antibiotics the standard of care and now focus our attention on improving other aspects of the resuscitation. Recent literature has studied goal MAP requirements [2], endpoints such as lactate clearance vs ScvO2 [3], and how best to evaluate volume responsiveness (IVC measurement [4], passive leg raise [5], carotid velocity time integral [6]). One element that has received far less attention is the type of fluid that is administered during the resuscitation. As it turns out, the type of fluid you choose does matter; it may be the difference between your patient requiring dialysis or even dying. Specifically, which crystalloid solution should be your fluid of choice in patients requiring large-volume resuscitations, such as those with sepsis or diabetic ketoacidosis?

There are different types of crystalloid fluids. Crystalloids such as lactated ringers (LR) or PlasmaLyte are considered “balanced fluids,” while chloride-rich fluids such as normal saline (NS) are not. Colloids including albumin and starches are not considered in this discussion. What makes some fluids “balanced” and others not is discussed later, but more importantly, does your choice of fluid really change outcomes? I was always taught that critically ill patients needed aggressive fluid resuscitation, so pick whichever one was most convenient, and make sure to start your resuscitation as soon as possible. And this practice seemed to be supported by my experiences in the ED. A patient would present in shock, and I’d resuscitate them with 4-6 liters of NS, trying to make sure I didn’t give them too much or too little fluid. By the time an ICU bed became available, the patient generally seemed to have improved, and I thought I had done a great job. As it turns out, I probably could have done better, as the choice of fluid does matter. Recent literature demonstrates important differences in outcomes such as the need for renal replacement therapy and mortality in patients that are resuscitated with balanced fluids, as opposed to chloride-rich solutions such as NS.

A recent review cited several studies associating NS infusions with the development of a hyperchloremic metabolic acidosis, in addition to other findings of questionable clinical importance [7]. Recently, other studies demonstrated that NS leads to more adverse events, and worse patient outcomes, compared to resuscitation with a balanced fluid. A large retrospective cohort study compared patients undergoing either elective or emergent open general surgical operations that received either NS or a balanced fluid the day of the procedure [8]. Unadjusted in-hospital mortality (5.6% CI 5.3-5.8 vs. 2.9% CI 2.0-4.2; p<0.001) and the number of patients developing major complications (33.7 vs. 23%) were significantly greater in the group that received NS compared to the group that received balanced crystalloids. After using propensity scoring to correct for multiple variables, the difference in mortality was no longer significantly different; however, patients that received NS were 4.8 times more likely to require dialysis (p<0.001). In addition, an analysis of patients requiring emergent general surgery showed an adjusted odds of death nearly 50% less in the cohort that received a balanced resuscitation compared to NS (OR 0.51 CI 0.28-0.95). In a prospective, open-label study of consecutive patients admitted to an intensive care unit (ICU), those that received balanced crystalloids had a decreased incidence of acute kidney injury (OR 0.52 CI 0.37-0.75; p<0.001) and less need for renal replacement therapy (OR 0.52 CI0.33-0.81; p=0.004) [9]. Over a six-month control period, patients were admitted to an ICU and given intravenous fluids including chloride-rich solutions, such as NS. This was followed by a six-month intervention period where chloride-rich fluids could only be administered by the attending physician. Seven hundred and sixty patients were admitted during the control and 773 during the interventional period. Patients had similar baseline creatinine concentrations, SAPS II, and APACHE scores. Mortality was similar during the control and study periods.

In a meta-analysis more relevant to emergency medicine, septic patients that received balanced fluids had a trend towards a lower mortality than those that received NS (OR 0.78 95% credibility intervals 0.58-1.05) [10]. All included studies were randomized, controlled trials of adults with severe sepsis or septic shock. The authors evaluated multiple fluid resuscitation strategies (among them balanced vs. unbalanced crystalloids) with a primary endpoint of 90-day mortality. While the complete analysis included 14 studies and 18,916 patients, no trial that was included directly compared balanced to unbalanced crystalloid solutions. The authors do not report how many patients were included in the subanalysis (balanced vs. unbalanced crystalloids).

So what causes a fluid to be “balanced?” Crystalloids that are “balanced” have the presence of an organic anion (such as lactate) and a lower chloride content that more closely resembles the composition of plasma. The difference between the strong cations and the strong anions (the positives and the negatives) in “balanced fluids” is 24-28. In plasma, the actual difference between the sodium (Na) (142 mEq/L) and chloride (Cl) (103 mEq/L) is approximately 39. However by using 24-28 instead of 38-42, you also account for the dilutional effect of the fluid that you are infusing, which dilutes the patient’s albumin and alters the acid-base status [11].

Perhaps an easier way to think about this is to compare the difference between the positives and negatives of the fluid (the SID or strong ion difference) and compare it to the patient’s bicarbonate concentration. If the SID is less than the patient’s bicarbonate, the fluid will be acidotic; if the SID is greater than the patient’s bicarbonate, the fluid will be alkalotic. As an example, let’s compare 2 commonly used fluids, 0.9% NS and LR. For NS, the difference between the Na and Cl is 0 (154-154=0). In a patient with a normal bicarbonate concentration (24 mEq/L), the bicarbonate is greater than the SID and so the fluid will essentially be acidotic [12]. This explains why patients that receive large amounts of NS develop a non-anion gap metabolic acidosis. The difference between the sodium and chloride in LR is 21 (130-109=21), which is nearly equal to a patient’s normal bicarbonate of 24 mEq/L and so is considered a “balanced fluid;” it does not cause the acidosis associated with NS. For a more in depth discussion of SID and acid-base status, I’d highly recommend the EMCrit Podcast #50.

While some emergency physicians have the good fortune to practice in an ED that stocks multiple types of fluid, many of us aren’t as lucky. We’ll have to work with hospital administration and the pharmacy to change the culture. This may mean educating your administration and adjusting your order sets so balanced fluids are an option under frequent orders. Balanced solutions such as LR cost the same as NS and are just as available in most places, so there is not a great reason for nursing or pharmacy not to make them readily accessible in the ED. While it’s easy to just give in to the status quo, it’s important to remember that we only have 24/7 cath lab activation, use of ketamine and propofol for sedations, and emergency ultrasonography because emergency physicians showed these improved patient care and pushed for change. Next to these, improving access to balanced fluids seems like it should be an easy thing to change.

So how does this change my practice? For the patient that is a little dehydrated that only needs 1-2 liters of fluid, not much; it probably doesn’t matter which crystalloid you use. For those patients that are already receiving NS and are going to require a large-volume resuscitation, there is likely a benefit in switching to a more balanced approach.

1. Rivers et al. Early Goal-Directed Therapy in the Treatment of Severe Sepsis and Septic Shock. N Engl J Med 2001;345:1368-1377.
2. Asfar et al. High Versus Low Blood-Pressure Target in Patients with Septic Shock. N Engl J Med 2014;370:1583-1593.
3.Jones et al. Lactate Clearance vs Central Venous Oxygen Saturation as Goals of Early Sepsis Therapy: A Randomized Clinical Trial. JAMA 2010;303(8):739-46.
4. Nagdev et al. Emergency Department Bedside Ultrasonographic Measurement of the Caval Index for Noninvasive Determination of Low Central Venous Pressure. Ann Emerg Med 2010;55(3):290-5.
5. Duus et al. The Reliability and Validity of Passive Leg Raise and Fluid Bolus to Assess Fluid Responsiveness in Spontaneously Breathing Emergency Department Patients. J Crit Care 2015;30(1):217.e1-e5.
6. Marik et al. The Use of Bioreactance and Carotid Doppler to Determine Volume Responsiveness and Blood Flow Redistribution following Passive Leg Raising in Hemodynamically Unstable Patients. Chest 2013;13(2):364-70.
7. Lobo et al. Should Chloride-Rich Crystalloids Remain the Mainstay of Fluid Resuscitation to Prevent ‘Pre-Renal’ Acute Kidney Injury:con. Kidney International 2014;86:1096-1105.
8. Shaw et al. Major Complications, Mortality, and Resource Utilization after Open Abdominal Surgery: 0.9% saline compared to Plasma-Lyte. Ann Surg 2012;255:821-829.
9. Yunos et al. Association between a Chloride-Liberal vs Chloride-Restrictive Intravenous Fluid Administration Strategy and Kidney Injury in Critically Ill Adults. JAMA 2012;308:1566-1572.
10. Rochwerg et al. Fluid Resuscitation in Sepsis. A Systematic Review and Network Meta-Analysis. Ann Intern Med 2014;161:347-355.
11.Morgan TJ. Clinical review: The meaning of Acid-Base Abnormalities in the Intensive Care Unit-Effects of Fluid Administration. Critical Care 2005;9(2):204-211.
12. Carlesso et al. The rule Regulating pH Changes during Crystalloid Infusion. Intensive Care Med 2011;37(3):461-68.


Evan Schwarz, MD is a faculty member in Emergency Medicine at Washington University in St. Louis.


  1. Dr. Schwarz

    Thank you for your article.  Your piece has helped me to start thinking more about the IVF I use and I can share it with my colleagues in the emergency department I work in.

  2. I’m wondering if half normal saline plus 3 amps of bicarb would be better and more available in departments that just stock saline

  3. Very interesting article! How does using a fluid with lactate in it affect the use of lactate levels as an index of hypoxia?

  4. If I may chime in from a different perspective; there is an argument that beginning the fluid resuscitation pre-hospital is desired, when possible. However, not everyone carries LR, they only carry NS. Now what?

    • Seth Hawkins on

      We operate in that exact situation. EMS starts a NS bolus. Their volume never reaches the levels described as needed for the risks associated with NS. LR is started upon hospital arrival.

Leave A Reply