PHARMACY COLUMN
Hyperchloremia in the Neurocritically Ill Patient
By Caroline Der Nigoghossan, PharmD, BCCCP, and Shivani Ghoshal, MD
Intravenous crystalloid administration is
one of the most common interventions
performed in critically ill patients.
However, during the last decade, a growing
body of literature suggests potential harm
with 0.9 percent sodium chloride (also
known as normal saline) and its associated
hyperchloremia, most notably in the
critically ill.
In actuality, the adjective “normal” is a misnomer; 0.9 percent
sodium chloride solution contains a supraphysiologic chloride
concentration of 154 mEq/L – 40-60 percent higher than normal
human plasma levels. Hyperchloremia itself may lead to nonanion
gap metabolic acidosis and has been associated with renal
dysfunction, among other complications. On the other hand,
balanced crystalloids contain organic anions such as acetate and
lactate, which are then metabolized to bicarbonate and electrolyte
concentrations similar to human plasma. Clinicians may be
reluctant to use some of the balanced crystalloids in patients with
hyperkalemia, hypercalcemia, hyponatremia, hyperlactatemia and
cerebral edema.
The majority of studies comparing 0.9 percent sodium chloride
to balanced crystalloids (Table 1) thus far have been conducted
in general critical ill populations and large studies including
neurocritically ill patients are lacking. In 2012, Yunos et al. studied
the association of chloride-restrictive versus chloride-liberal
intravenous fluid administration with acute kidney injury (AKI)
in a mixed-bed intensive care unit (ICU). The group found that a
chloride-restrictive strategy among their critically ill patients was
associated with a significant decrease in both the incidence of AKI
as well as the use of renal replacement therapy (RRT).1 Subsequent
large retrospective analyses have also found association between
higher chloride loads and in-hospital mortality in patients with
systemic inflammatory response syndrome and septic shock, even
after controlling for received volume and other confounding
variables.2,3 The much-awaited double-blind, cluster-randomized,
double-crossover 2015 SPLIT feasibility trial did not show reduced
incidence of kidney injury in critically ill patients receiving
buffered crystalloid versus 0.9 percent sodium chloride. Both
groups showed a low utilization of RRT and low mortality. The
study was limited by unreported serum chloride levels, relatively
low volumes of administered fluids and low patient acuity, making
it poorly generalizable. The majority of patients were admitted
following elective surgery, with a mean APACHE II score of 14 and
a median two liters of fluid administered in both groups.4
The most recent Isotonic Solutions and Major Adverse Renal
Events Trial (SMART) published by Semler et al. at Vanderbilt, is
the largest trial to date comparing the use of balanced crystalloids
(Lactated Ringer’s or Plasma-Lyte A) versus 0.9 percent sodium
chloride in the critically ill. Five intensive care units were
randomized to a sequence of alternating crystalloid group on a
monthly basis. Relative contraindications to balanced crystalloids
included hyperkalemia and brain injury and decision to use
saline instead was left to the clinician’s preference. Fewer patients
in the balanced crystalloids group had a plasma chloride > 110
mEq/L (24.5 percent versus 35.6 percent, p<0.001). The primary
endpoint was a composite of major adverse kidney events
within 30 days: new renal replacement therapy, persistent renal
dysfunction (defined as an increase of the serum creatinine by
≥ 200 percent from baseline), and death from any cause. The
composite endpoint occurred less in the balanced-crystalloids
group, a difference that was mirrored in the subgroup analysis
of neurologic ICU patients. Although the effect size was low, the
widespread use of crystalloids renders this difference clinically
significant. These results were inconsistent among patients with
traumatic brain injury; however, clinicians had the option of
administering 0.9 percent sodium chloride for these patients
regardless of group assignment, making conclusions difficult to be
drawn for this patient population.5
Hyperchloremia is often seen as a necessary evil in management
of the neurocritically ill, where hypertonic saline is a mainstay
treatment for increased intracranial pressure (ICP). Two recent
studies, summarized in Table 1, have evaluated the impact of
hyperchloremia in neurocritically ill patients. The first, by Riha
et al., evaluated the impact of hyperchloremia (chloride ≥ 115
mEq/L) in a multicenter retrospective propensity-matched cohort
study of intracerebral hemorrhage (ICH) patients with continuous
infusion of 3 percent hypertonic saline. Patients exhibiting
hyperchloremia were associated with a higher incidence of inhospital
mortality, though patients in the hyperchloremia group
had lower baseline Glascow Coma Scale score, higher ICH score,
and required more mechanical ventilation and cerebrospinal fluid
diversion.
After adjusting for differences between both groups by propensity
score matching, in-hospital mortality and AKI still occurred more
frequently in the hyperchloremia group.6 Another retrospective
analysis from Sadan et al. evaluated AKI prevalence in patients
with subarachnoid hemorrhage. Patients with AKI were found to
have higher Hunt and Hess scores, longer duration of mechanical
ventilation and ICU length of stay compared to patients without
AKI. Additionally, these patients received more hypertonic saline
(3 percent or 23.4 percent NaCl) treatment, had higher mean
serum creatinine, chloride and sodium concentrations during
their ICU stay. The authors identified male gender, hypertension,
abnormal admission creatinine, mean serum chloride
concentration and hypertonic saline therapy as predictors of AKI
development.7 In both studies, patients in the AKI group were
sicker and older, and it is difficult to determine if the mortality
difference between both groups was a result of hyperchloremia,
a higher incidence of AKI, or a higher severity of illness in the
hyperchloremic patients.
Despite study limitations in the literature evaluating the impact
of hyperchloremia in neurocritically ill patients, the available data
in the critically ill suggests that hyperchloremia should not be
disregarded in patient care. Growing evidence prompts clinicians
to develop new strategies to reduce chloride accumulation in
patients with ICP crises. An alternative strategy includes the use of
mixed sodium acetate/sodium chloride solutions for both crisis
and continuous infusion therapies. Further research is similarly
needed to evaluate the role of isotonic balanced crystalloids
containing physiologic sodium concentration (e.g. Plasma-Lyte)
for fluid resuscitation in neurocritically ill patients.
33