Saline Is the Solution for Crystalloid Resuscitation
Paul Young, MBChB
Key Words: 0.9% saline; acute kidney injury; buffered crystalloids; As well as its effects on acid-base physiology, 0.9% saline
critical care medicine; fluid resuscitation; intravenous fluid therapy; may affect renal physiology because chloride seems to have an
perioperative care; Plasma-Lyte; Ringer lactate important role in tubuloglomerular feedback mechanisms (8).
Specifically, the chloride concentration in the fluid delivered to
the distal tubule seems to be one mediator of tubuloglomeru-
lar feedback. As the chloride concentration in the distal tubule
F
luid resuscitation is a fundamental component of fluid rises, feedback occurs via the macula densa, the afferent
the management of acutely ill patients. The choice arteriole constricts, and the glomerular filtration rate drops
of fluid has been an issue of longstanding debate (1). (9, 10). In healthy volunteers, renal artery blood flow velocity
With respect to the choice of crystalloid for fluid resuscita- and renal cortical tissue perfusion fall after administration of
tion, such debate has often been framed around the notion 2L of 0.9% saline but not after administration of 2L of a buff-
that saline is less physiologic than the other commercially ered crystalloid (11).
available crystalloids (2). However, although 0.9% saline has These effects on biochemistry and acid-base physiology are
approximately 1.5 times the chloride concentration of human not in dispute, but there is little evidence that they translate
plasma, the lactate concentration of Ringer lactate is 28 times into clinically important effects on patient-centered outcomes
that of plasma, and the acetate concentration in Plasma-Lyte (8). When it comes to large-scale randomized controlled tri-
148 (Baxter, Deerfield, IL) and Normsol-R (Hospira Inc, Lake als, 0.9% saline has been administered to more than 8,000
Forest, IL) exceeds that found in plasma by as much as 1,000 ICU patients without any evidence, suggesting that it results
times (3). In other words, none of the commercially available in worse outcomes for patients than its comparator (5, 12, 13).
crystalloids are inherently more physiologic than the others. Indeed, 0.9% saline was associated with a reduced risk of death
There is little doubt that 0.9% saline administration has and disability compared with albumin in patients with trau-
effects on acid-base physiology (4). Because 0.9% saline has matic brain injury (14) and with a reduced risk of requiring
a strong ion difference of zero, its administration would be renal dialysis compared with hydroxyethyl starch (Voluven) in
expected to cause metabolic acidosis (4), and indeed, the devel- an all-comers population of critically ill adults requiring fluid
opment of mild or even moderate hyperchloremic metabolic resuscitation (12).
acidosis is often observed at the bedside in critically ill patients In the Saline versus Plasma-Lyte 148 for Intensive care
who have received moderate volumes of IV saline. That said, unit fluid Therapy (SPLIT) trial, there were no differences in
in the Saline versus Albumin Fluid Evaluation study (5), even patient-centered outcomes between the treatment groups in
when 0.9% saline was used for fluid resuscitation in ICU, there 2,278 critically ill patients (13). There was, however, one patient
was a significant increase in serum bicarbonate and base excess in the Plasma-Lyte 148 group who died after a serious adverse
over time after randomization, often leading to significant event judged by the treating clinician to be potentially related
metabolic alkalosis (6). As balanced crystalloids have a positive to study treatment. This patient developed severe lactic acidosis
strong ion difference compared with saline (7), it seems likely and progressive multiple organ failure culminated in circula-
that large volumes of balanced crystalloids will exacerbate this tory collapse and death with no specific cause of death iden-
common acid-base disturbance among critically ill patients tified at autopsy. Hyperlactemia occurring in association with
compared with saline. sodium acetate infusion has been reported previously (15).
The overall exposure to the study fluids in the SPLIT trial
Intensive Care Unit, Wellington Regional Hospital, Wellington, New Zealand. was small (a median of 2L), and the study population had low
Dr. Young received support for this article research from the Health illness acuity. As a result, further large randomized trials are
Research Council of New Zealand and disclosed other support (Baxter needed to assess the effects of 0.9% saline compared with buff-
Healthcare Corporation provided intravenous fluids for the Saline Versus
Plasma-Lyte Intensive care unit fluid Therapy study). ered crystalloids in high-acuity ICU patients receiving larger
For information regarding this article, E-mail: [email protected] fluid volumes. Until such trials are conducted, the compara-
Copyright 2016 by the Society of Critical Care Medicine and Wolters tive effectiveness of 0.9% saline and buffered crystalloids in
Kluwer Health, Inc. All Rights Reserved. critically ill patients can only be inferred from small-scale ran-
DOI: 10.1097/CCM.0000000000001844 domized controlled trials and observational trials. Apart from
Critical Care Medicine www.ccmjournal.org 1
Copyright 2016 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
Young
the SPLIT trial, the existing evidence base from randomized Saline (0.9%) is the dominant IV crystalloid fluid in North
controlled trials in critically ill patients comparing saline with America (23). Fundamentally, this is a debate about where the
other crystalloids is extremely limited with only 293 patients threshold for practice change lies. I submit that the current
enrolled in such randomized controlled trials in total and level of evidence falls far below that threshold. The substan-
no significant difference reported in any clinically important tial and important work described in the studies outlined in
outcome (16). this Viewpoint has paved the way for a definitive large-scale
In the most widely cited observational cohort study remov- randomized controlled trial. The Australian and New Zealand
ing chloride-rich fluids, including 0.9% saline, from a single Intensive Care Society Clinical Trials Group will soon begin
ICU was associated with a reduction in the cumulative inci- enrolling patients into the Plasma-Lyte 148 versUs saline (PLUS)
dence of acute kidney injury (AKI) and reduced requirements trial, an 8,800 participant double-blind randomized controlled
for renal replacement therapy (17). However, in this study, there trial with a primary end point of day 90 mortality. PLUS will
were differences in albumin use in the pretreatment and post- definitively establish the relative efficacy and safety of buffered
treatment phases, and one of the fluids whose use was discon- crystalloid compared with 0.9% saline in critically ill patients
tinued was a synthetic gelatin-based colloid. The use of gelatins with high mortality risk. Until the results of PLUS are known,
has previously been associated with an increased risk of AKI 0.9% saline should remain the first choice for crystalloid fluid
in patients with sepsis (18), and thus, it is plausible that the resuscitation. It is a choice supported by level I evidence (5, 12).
observed difference was related to removing gelatin from the The alternatives are fluids that have either not been shown to be
ICU not removing 0.9% saline. That said, as multiple changes superior to saline in randomized controlled trials in critically ill
in fluid therapy occurred simultaneously and the study was an patients (i.e., Plasma-Lyte 148) (13) or have not been tested in
open-label before-and-after trial (17), it is impossible to say large-scale randomized controlled trials at all (e.g., Ringer lac-
with any certainty what component of the fluid change strategy tate). Saline is the first choice crystalloid fluid and is supported
was responsible for the observed changes or even if the fluid by 150 years of clinical experience (24). Our options are to stick
therapy was responsible for the observed changes at all. with what is tried and tested or to change to more expensive flu-
A recent meta-analysis of randomized controlled trials and ids on the basis of inductive physiologic reasoning and observa-
observational studies comparing chloride-liberal fluids with tional data that are subject to bias and confounding.
chloride-restrictive fluids in perioperative and critical care
reported that the use of chloride-liberal fluids was associated
with a significant increase in the risk of AKI (16). However,
REFERENCES
1. Myburgh JA, Mythen MG: Resuscitation fluids. N Engl J Med 2013;
77% of the patients included in this meta-analysis were from 369:12431251
the single centre before and after study described above (17). 2. Lira A, Pinsky MR: Choices in fluid type and volume during resuscita-
When this study was excluded, there was no significant effect on tion: Impact on patient outcomes. Ann Intensive Care 2014; 4:38
AKI (16). Furthermore, two additional studies published since 3. Tollinger CD, Vreman HJ, Weiner MW: Measurement of acetate in
human blood by gas chromatography: Effects of sample preparation,
this meta-analysis reported no significant association between feeding, and various diseases. Clin Chem 1979; 25:17871790
choice of IV crystalloid and AKI risk (19, 20). The first study 4. Guidet B, Soni N, Della Rocca G, et al: A balanced view of balanced
compared 3,365 patients with sepsis treated with balanced flu- solutions. Crit Care 2010; 14:325
ids with a propensity-matched cohort of 3,365 patients treated 5. Finfer S, Bellomo R, Boyce N, et al; SAFE Study Investigators: A com-
parison of albumin and saline for fluid resuscitation in the intensive
with 0.9% saline. The second study evaluated the impact of fluid care unit. N Engl J Med 2004; 350:22472256
composition on outcomes in patients with systemic inflamma- 6. Bellomo R, Morimatsu H, French C, et al; SAFE Study Investigators:
tory response syndrome and included 1,158 0.9% saline-treated The effects of saline or albumin resuscitation on acid-base status and
patients and 1,158 propensity-matched patients treated with a serum electrolytes. Crit Care Med 2006; 34:28912897
7. Morgan TJ, Venkatesh B, Hall J: Crystalloid strong ion difference
calcium-free balanced solution as the primary fluid (20). determines metabolic acid-base change during in vitro hemodilution.
Although some observational studies have reported an Crit Care Med 2002; 30:157160
increased mortality risk associated with the use of 0.9% saline 8. Yunos NM, Bellomo R, Story D, et al: Bench-to-bedside review: Chlo-
(19, 21), the recent meta-analysis reports no overall signifi- ride in critical illness. Crit Care 2010; 14:226
cant increase or decrease in the risk of in-hospital mortality 9. Morsing P, Velazquez H, Ellison D, et al: Resetting of tubuloglomerular
feedback by interrupting early distal flow. Acta Physiol Scand 1993;
with the use of high chloride fluid (0.9% saline) compared 148:6368
with low chloride fluid (buffered crystalloid) (risk ratio, 1.13; 10. Schnermann J, Ploth DW, Hermle M: Activation of tubulo-glomerular
95% CI, 0.921.39) (16). Overall, there are no consistent sig- feedback by chloride transport. Pflugers Arch 1976; 362:229240
nals of harm from 0.9% saline in observational studies, and 11. Chowdhury AH, Cox EF, Francis ST, et al: A randomized, controlled,
double-blind crossover study on the effects of 2-L infusions of 0.9%
any potential harms identified need to be considered in light of saline and plasma-lyte 148 on renal blood flow velocity and renal corti-
the inevitable limitations of bias and confounding inherent in cal tissue perfusion in healthy volunteers. Ann Surg 2012; 256:1824
all observational studies (22). The patients who received 0.9% 12. Myburgh JA, Finfer S, Bellomo R, et al; CHEST Investigators; Austra-
saline may be systematically different from the patients who lian and New Zealand Intensive Care Society Clinical Trials Group:
Hydroxyethyl starch or saline for fluid resuscitation in intensive care.
received buffered crystalloids, and the treatments provided by N Engl J Med 2012; 367:19011911
doctors who use 0.9% saline may be different to the treatments 13. Young P, Bailey M, Beasley R, et al; SPLIT Investigators; ANZICS
provided by doctors who use buffered crystalloids. CTG: Effect of a Buffered Crystalloid Solution vs Saline on Acute
2 www.ccmjournal.org XXX 2016 Volume XX Number XXX
Copyright 2016 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
Pro-Con Article
Kidney Injury Among Patients in the Intensive Care Unit: The SPLIT 19. Raghunathan K, Shaw A, Nathanson B, et al: Association between
randomized clinical trial. JAMA 2015; 314:17011710 the choice of IV crystalloid and in-hospital mortality among critically ill
14. Cooper DJ, Finfer S, Bellomo R, et al; SAFE Study Investigators; Aus- adults with sepsis*. Crit Care Med 2014; 42:15851591
tralian and New Zealand Intensive Care Society Clinical Trials Group; 20. Shaw AD, Schermer CR, Lobo DN, et al: Impact of intravenous fluid
Australian Red Cross Blood Service; George Institute for Interna- composition on outcomes in patients with systemic inflammatory
tional Health: Saline or albumin for fluid resuscitation in patients with response syndrome. Crit Care 2015; 19:334
traumatic brain injury. New Engl J Med. 2007;357:874884 21. Raghunathan K, Bonavia A, Nathanson BH, et al: Association
15. McCague A, Bowman N, Wong DT: Lactic acidosis after resuscita- between initial fluid choice and subsequent in-hospital mortality dur-
tion with sodium acetate. J Surg Res 2012; 173:362364 ing the resuscitation of adults with septic shock. Anesthesiology.
16. Krajewski ML, Raghunathan K, Paluszkiewicz SM, et al: Meta-analysis 2015; 123:13851393
of high- versus low-chloride content in perioperative and critical care 22. MacMahon S, Collins R: Reliable assessment of the effects of treat-
fluid resuscitation. Br J Surg 2015; 102:2436 ment on mortality and major morbidity, II: Observational studies. Lan-
17. Yunos NM, Bellomo R, Hegarty C, et al: Association between a chloride- cet 2001; 357:455462
liberal vs chloride-restrictive intravenous fluid administration strategy 23. Finfer S, Liu B, Taylor C, et al; SAFE TRIPS Investigators: Resusci-
and kidney injury in critically ill adults. JAMA 2012; 308:15661572 tation fluid use in critically ill adults: An international cross-sectional
18. Bayer O, Reinhart K, Sakr Y, et al: Renal effects of synthetic colloids study in 391 intensive care units. Crit Care 2010; 14:R185
and crystalloids in patients with severe sepsis: A prospective sequen- 24. Awad S, Allison SP, Lobo DN: The history of 0.9% saline. Clin Nutr
tial comparison. Crit Care Med 2011; 39:13351342 2008; 27:179188
Critical Care Medicine www.ccmjournal.org 3
Copyright 2016 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
