Griesdale et al. Critical Care 2013, 17:401
http://ccforum.com/content/17/1/401
LETTER
Hypernatremia and intracranial pressure:
more questions than answers
Donald E G Griesdale*1-3, Mypinder S Sekhon1 and William R Henderson1
See related research by Wells et al., http://ccforum.com/content/16/5/R193
A critical limitation of observational methodology is
Abstract confounding by indication, which exists when variables
The observational literature suggests that associated with exposure are also associated with out-
hypernatremia is associated with worse outcomes comes in the study base [3]. Clinicians administer hyper-
in patients with traumatic brain injury. In a previous osmolar therapy and induce hypernatremia on the basis
issue of Critical Care, Wells and colleagues add to this of measured and unmeasured characteristics of patients.
literature by failing to show an association between Lack of a formalized TBI protocol, including indications
hypernatremia and reduced intracranial pressure. for hypertonic saline (HTS), can exacerbate this bias. For
However, we must bear in mind many limitations example, clinicians in the study by Wells and colleagues
of observational methods before eliminating used both boluses and infusions of HTS with a range of
hyperosmolar therapy from our armamentarium. administered concentrations (3% versus 7.5%). No
attempt was made to adjust for differences in baseline
characteristics of patients. Furthermore, although regres-
In a previous issue of Critical Care, Wells and colleagues sion is an important method to help reduce bias, several
[1] report on their retrospective cohort study examining assumptions must be met for the analysis to be valid.
the relationship between serum sodium and intracranial Linear regression is predicated upon independence of
pressure (ICP) with severe traumatic brain injury (TBI). data. With 1,230 paired sodium and ICP measurements
This is an ongoing topic of interest to neurointensivists as in 81 patients, there is likely to be marked within-subject
hyperosmolar therapy remains a treatment option for correlation of data that is not taken into account by this
elevated ICP [2]. We commend the authors for their analysis. Consequently, a linear mixed model or other
contribution to a fundamentally important issue in analysis is warranted to specifically model and account
neurocritical care; however, several issues deserve further for this correlation [4]. Failing to do so limits the
discussion. interpretability of their results.
Authors’ response
Diana L Wells, Joseph M Swanson, G Christopher Wood, Louis J Magnotti, Bradley A Boucher, Martin A Croce, Charles G Harrison,
Michael S Muhlbauer and Timothy C Fabian
We thank Griesdale and colleagues for their thoughtful evaluated this practice in one of the largest studies to
letter describing concerns with our study, specifically the date [1,5,6].
retrospective design and statistical tests. Owing to the We agree with the limitations mentioned by Griesdale
lack of literature supporting the benefits of HTS to induce and colleagues in regard to the lack of a formalized TBI
hypernatremia in patients with TBI, we pragmatically protocol and the inclusion of various doses of HTS, and
we addressed this in our discussion. Importantly, patients
*Correspondence: in our study were treated with the same general approach
1
Department of Medicine, Division of Critical Care Medicine, Vancouver General
Hospital, University of British Columbia, 855 West 12th Avenue, Vancouver, BC,
to hyperosmolar therapy (84% of patients received 3%
V5Z 1M9, Canada NaCl boluses) [1].
Full list of author information is available at the end of the article In regard to adjustment for baseline characteristics, we
performed linear regression in important subgroups
© 2010 BioMed Central Ltd © 2013 BioMed Central Ltd divided by baseline ICP and Glasgow Coma Scale score.
http://ccforum.com/content/17/1/401
LETTER
Hypernatremia and intracranial pressure:
more questions than answers
Donald E G Griesdale*1-3, Mypinder S Sekhon1 and William R Henderson1
See related research by Wells et al., http://ccforum.com/content/16/5/R193
A critical limitation of observational methodology is
Abstract confounding by indication, which exists when variables
The observational literature suggests that associated with exposure are also associated with out-
hypernatremia is associated with worse outcomes comes in the study base [3]. Clinicians administer hyper-
in patients with traumatic brain injury. In a previous osmolar therapy and induce hypernatremia on the basis
issue of Critical Care, Wells and colleagues add to this of measured and unmeasured characteristics of patients.
literature by failing to show an association between Lack of a formalized TBI protocol, including indications
hypernatremia and reduced intracranial pressure. for hypertonic saline (HTS), can exacerbate this bias. For
However, we must bear in mind many limitations example, clinicians in the study by Wells and colleagues
of observational methods before eliminating used both boluses and infusions of HTS with a range of
hyperosmolar therapy from our armamentarium. administered concentrations (3% versus 7.5%). No
attempt was made to adjust for differences in baseline
characteristics of patients. Furthermore, although regres-
In a previous issue of Critical Care, Wells and colleagues sion is an important method to help reduce bias, several
[1] report on their retrospective cohort study examining assumptions must be met for the analysis to be valid.
the relationship between serum sodium and intracranial Linear regression is predicated upon independence of
pressure (ICP) with severe traumatic brain injury (TBI). data. With 1,230 paired sodium and ICP measurements
This is an ongoing topic of interest to neurointensivists as in 81 patients, there is likely to be marked within-subject
hyperosmolar therapy remains a treatment option for correlation of data that is not taken into account by this
elevated ICP [2]. We commend the authors for their analysis. Consequently, a linear mixed model or other
contribution to a fundamentally important issue in analysis is warranted to specifically model and account
neurocritical care; however, several issues deserve further for this correlation [4]. Failing to do so limits the
discussion. interpretability of their results.
Authors’ response
Diana L Wells, Joseph M Swanson, G Christopher Wood, Louis J Magnotti, Bradley A Boucher, Martin A Croce, Charles G Harrison,
Michael S Muhlbauer and Timothy C Fabian
We thank Griesdale and colleagues for their thoughtful evaluated this practice in one of the largest studies to
letter describing concerns with our study, specifically the date [1,5,6].
retrospective design and statistical tests. Owing to the We agree with the limitations mentioned by Griesdale
lack of literature supporting the benefits of HTS to induce and colleagues in regard to the lack of a formalized TBI
hypernatremia in patients with TBI, we pragmatically protocol and the inclusion of various doses of HTS, and
we addressed this in our discussion. Importantly, patients
*Correspondence: in our study were treated with the same general approach
1
Department of Medicine, Division of Critical Care Medicine, Vancouver General
Hospital, University of British Columbia, 855 West 12th Avenue, Vancouver, BC,
to hyperosmolar therapy (84% of patients received 3%
V5Z 1M9, Canada NaCl boluses) [1].
Full list of author information is available at the end of the article In regard to adjustment for baseline characteristics, we
performed linear regression in important subgroups
© 2010 BioMed Central Ltd © 2013 BioMed Central Ltd divided by baseline ICP and Glasgow Coma Scale score.
