Low Morbidity and Mortality in Children with Diabetic Ketoacidosis Treated

with Isotonic Fluids

Perrin C. White, MD, and Bryan A. Dickson, MD
Objective To assess current rates of complications of diabetic ketoacidosis (DKA), particularly cerebral edema, in
a large tertiary-care pediatric hospital with a consistent management protocol.

Study design We report our single-center retrospective experience with 3712 admissions with DKA in
1999-2011. Our DKA protocol features a 3-bag system using 2 bags of rehydration fluids, identical except for
the presence in 1 bag of 10% dextrose, to allow rapid adjustment of glucose infusion rate. The third bag contains
insulin. Fluids are administered at a total rate of 2-2.5 times maintenance fluid requirements. Total electrolyte
concentration is kept approximately isotonic. Billing and medical records databases at Childrens Medical Center
Dallas were examined for cases of DKA, cerebral edema, other morbidities, and death.
Results We ascertained 20 cases of cerebral edema (0.5%). Most presented early (median duration of treatment 2
hours). Only 10 of 20 computed tomography scans were graded as moderate edema or worse. Only 10 patients
received treatment other than routine DKA management. There was 1 death in a patient with sickle cell trait who developed intravascular sickling. Two patients had neurologic sequelae at hospital discharge but both recovered fully.
Conclusions Compared with data in recent consensus statements, the Dallas protocol is associated with
extremely low rates of death and disability (0.08% vs 0.3%) from DKA. (J Pediatr 2013;163:761-6).

iabetic ketoacidosis (DKA) is the most common reason for hospitalization of children with type 1 diabetes after initial
diagnosis, occurring
1-10 times per 100 patient-years.1 It is the result of insulin deficiency, as occurs in patients with
newly diagnosed diabetes or in situations of poor compliance with the diabetes regimen, often in the context of intercurrent illness. Treatment of DKA requires intravenous administration of insulin and correction of fluid and electrolyte losses with
appropriate intravenous fluids. Cerebral edema is one of the most feared complications of DKA. If severe, it can result in brain
herniation, with a significant risk of death or permanent neurologic damage. The most recent large clinical series from the US
(which includes 1 Australian center)2 and from the United Kingdom,3 as well as recent reviews4 and consensus statements,1,5,6
all cite a prevalence of 0.5%-0.9% of cases with DKA and state that this complication accounts for 60%-90% of DKA mortality
and that the risks of death and permanent disability are 20%-25% and 10%-25% of cerebral edema cases, respectively.
The causes of cerebral edema are incompletely understood. Retrospective studies7 suggested that rates of fluid administration
exceeding 4 L/m2/24 h were associated with an increased risk of cerebral edema, but more recent studies,2,3 including a small
randomized controlled trial,8 have failed to show correlations between the development of cerebral edema and fluid administration rates. Current protocols1,5,6 call for rates of
3 L/m2/24 h. In 2000, we described our experience with a DKA protocol we
instituted in 1997,9 distinguished by a 3-bag system with 2 bags of rehydration fluidsidentical except for the presence in 1
bag of 10% dextrose (the third bag contained insulin)to facilitate rapid and continuous adjustment of net glucose concentration in the administered fluids to control the rate of decrease of blood glucose concentration.

Methods
This study was approved by the Institutional Review Board of the University of Texas Southwestern Medical Center. Hospital
billing databases were examined as far back as possible (January 1999) for admissions of $1 days duration with International
Classification of Diseases, Ninth Revision (ICD-9) diagnoses of 250.1x (where x equals an integer).
These admissions were searched for ICD-9 codes of 348.x (cerebral edema, compression of brain) or 96.x (intubation,
ventilation). The medical records database was searched for deaths with an ICD-9 code of 250.xx (diabetes of all types). All
records on patients with admitting diagnoses of 250.1x were examined individually if the billing database indicated an intensive
care unit admission and computed tomography (CT) scanning and/or magnetic

ADA
CT
DKA
ICD-9
NaCl

American Diabetes Association

Computed tomography
Diabetic ketoacidosis
International Classification of Diseases, Ninth Revision
Sodium chloride

From the Division of Pediatric Endocrinology, University

of Texas Southwestern Medical Center and Childrens
Medical Center, Dallas, TX
P.W. is the Audry Newman Rapoport Distinguished Chair
in Pediatric Endocrinology. The authors declare no conflicts of interest.
0022-3476/$ - see front matter. Copyright 2013 Mosby Inc.
All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2013.02.005

761

THE JOURNAL OF PEDIATRICS

www.jpeds.com

resonance imaging. We excluded cerebral edema cases and/or

deaths where the patients were not in DKA or had preexisting
complicating conditions.
In comparing the Dallas protocol9 with the 2006 American
Diabetes Association (ADA) consensus,5 we assumed for the
former a single intravenous fluid bolus of 20 mL/kg 0.9%
NaCl over 1 hour and subsequent fluid replacement with
0.675% NaCl and 49 mEq/L potassium at 3.75 L/m2/24 h.
For the ADA protocol, we assumed a single fluid bolus of
10 mL/kg 0.9% NaCl over 1 hour, then 0.9% NaCl and 49
mEq/L potassium for the next 3 hours at 3 L/m2/24 h, and
then 0.45% saline with 49 mEq/L added potassium at the
same rate for the remainder of 24 hours (note that both protocols specify adding 20 mmol/L each of potassium chloride
and potassium phosphate; the latter reagent is a mixture of
monobasic and dibasic potassium phosphate, and contains
1.47 mEq of potassium per mmol).
Demographics were compared using c2 tests, age by t tests,
and admission frequencies by Mann-Whitney tests. Outcomes were compared with use of c2tests.

Results
There were 3712 admissions to Childrens Medical Center Dallas in 1999-2011 with ICD-9 codes of 250.1x, denoting DKA
(Table I). These admissions involved 2346 unique patients;
538 patients accounted for 1902 admissions. Of these 2346
patients, 150 (6% of the total) were admitted $4 times and
accounted for 1014 admissions (27% of all admissions), and
22 patients accounted for 305 admissions. Sex did not
represent a risk factor for multiple admissions, but ethnicity
did, with African Americans being twice as likely to have >1
episode of DKA than either whites or Hispanics (generally
Mexican Americans in our population).
Of 100 random admissions with DKA, 66 were managed
on our DKA protocol; the remaining patients had milder
acidosis and were managed with intravenous fluids and subcutaneous insulin. In our hospital, patients with a routine
case of DKA are admitted to a specialized endocrinology
floor, but there were 358 intensive care unit admissions
and 229 patients underwent CT scanning.
We considered a patient to have cerebral edema if the patient was in DKA and either cerebral edema was diagnosed
on a CT scan or the patient received any treatment for altered
mental status other than routine DKA management

Vol. 163, No. 3

(Table II). Using these criteria, we ascertained 20 cerebral
edema cases in 19 patients among the 3712 admissions
(0.5%). Age, race, sex, and newly diagnosed diabetes were
not risk factors for cerebral edema (Table I). However, the
patients with cerebral edema were admitted more often over
the study period, a median (IQR) of 2 (1-4) times, versus all
other patients (median and IQR of 1 admission, P < .0001).
All 20 patients underwent CT scanning for altered mental
status (16 patients) and headache, seizure, suspected cerebral
edema, and hyperglycemia (1 patient each). Fourteen CT scans
were performed within 3 hours of arrival in the referring hospital or our hospitals emergency department (Figure). Of the
20 scans, 3 were read as normal, 7 as having mild cerebral
edema (3 of which were originally read as normal and
reclassified after subsequent review by the attending
radiologist), and 7 as moderate or ungraded. Three scans had
additional findings (herniation, hemorrhage, thrombosis).
Only 10 of the 20 patients received any treatment other
than routine DKA management. Seven received such
treatment at the referring hospital or in our emergency
department. Eight patients were intubated, and 7 patients
received intravenous boluses of mannitol. Six patients had
both interventions (including the 3 patients who subsequently had normal CT scans). Three patients required placement of ventricular drains.
There was 1 death of a patient with DKA in 13 years (0.03%;
5% of cerebral edema cases): a 13-year-old obese African
American girl who presented in a markedly hyperosmolar state
with a blood glucose level of 1483 mg/dl and pH 7.16. She had
sickle cell trait. Her autopsy showed no cerebral edema, but instead disseminated intravascular sickling that was judged to be
the cause of death. Two patients had neurologic sequelae at
hospital discharge. One was a 2-year-old patient with cerebral
edema and herniation. He made a gradual recovery and is
stated as having good school performance in the ninth grade;
he has attention-deficit disorder and is treated with atomoxetine. The other patient was an 8-year-old boy who had a posterior cerebral artery infarct. He also recovered gradually, has no
apparent permanent sequelae, and is in the 11th grade.
We examined subsequent outpatient records for the remaining patients with cerebral edema to identify possible
cognitive sequelae. Comments on school performance were
available for 12 of 17 patients; all were reported as having satisfactory school performance, and thus far 5 have graduated
from high school.

Table I. Demographic and biochemical characteristics of cases of DKA and cerebral edema
Age, mean  SD, y
Sex, No. (boys/girls)
Race, % (white/African American/Hispanic/other)
New diagnoses, No.
Admissions, median No. (IQR)
Initial pH, median (IQR)*
Initial glucose, median (IQR), mg/dl*

All DKA cases (N = 3712)

Cerebral edema cases (n = 20)

11.3  4.2
1769 (48%)/1923 (52%)
1830 (49%)/1019 (27%)/649 (17%)/194 (5%)
1888 (51%)
1 (1-1)
7.25 (7.15-7.33)
389 (313-552)

11.1  4.2
7 (35%)/13 (65%)
7 (35%)/7 (35%)/4 (20%)/2 (10%)
6 (30%)
2 (1-4)
7.07 (6.99-7.13)
530 (414-856)

NS
NS
NS
NS
<.0001
<.0001
<.0001

NS, Nonsignificant.
*Biochemical statistics determined on a random sample of 100 patients with DKA, and all cerebral edema cases.

762

White and Dickson

ORIGINAL ARTICLES

September 2013

Table II. Outcomes

Definition of DKA
Definition of cerebral edema

Total No. of DKA cases

Total No. of cerebral edema
cases, %

Dallas protocol
(present study)

US study2

UK study3

ICD-9 codes

ICD-9 codeszz

DKA AND altered mental

status AND [radiographic
evidence of cerebral
edema OR specific
treatment (hyperosmolar
therapy or intubation)]

Blood glucose >300 AND [pH <7.25

OR bicarbonate <15 mmol/L]
AND altered mental status AND
[radiographic or pathologic evidence
of cerebral edema OR specific
treatment (hyperosmolar therapy or
hyperventilation) followed by clinical
improvement]
6977
61 (0.9)*

Physician reports;
pH <7.3 or bicarbonate <18 mmol/L or heavy
ketonuria
Deterioration in level of consciousness AND $1 of:
Hypertension, bradycardia, pupillary changes,
ophthalmoplegia, papilledema, respiratory
arrest, decerebrate or decorticate posturing,
OR radiologic evidence of moderate or severe
cerebral edema

3712
20 (0.5)*,

20 (0.7)z

13 (0.3),z
Patients presenting >7 h (%)
Mild-moderate neurologic
disability, No.
Severe neurologic disability, No.
Death, No.
Total death and disability,
No. (%)

2940

2 (0.06){,**
2

30 (0.4){
8

10 (0.3)**, 34x
9x

0
1
3 (0.08)

4
13
25 (0.3)

8x

*P = .06 (for the comparison between the two quantities bearing this symbol).
The 2 figures for the Dallas data use definitions of cerebral edema as similar as possible to those of the US and UK studies, respectively.
zP = .06.
xCerebral edema cases were oversampled, and death and disability data are available only for the oversampled total.
{P = .0007.
**P = .006.
P = .008.
zzThe US study required pH <7.25 for subjects included in a nested case-control study.

Two patients had significant complications outside the central

nervous system; one had a radial artery thrombosis after line
placement, and the other developed acute renal failure requiring
hemodialysis. Follow-up information on the latter patient is unavailable. Median (IQR) length of hospitalization for the 19 surviving patients with cerebral edema was 4 (3-7) days, 2 days
longer (P < .0001) than for other patients admitted with DKA.

Discussion
Whereas it is commonly thought that cerebral edema from
DKA is the leading cause of death in children with type 1

Figure. Comparison between time of onset (in hours) of cerebral edema in the present study.2

diabetes,1 this is not consistent with our experience in a large

tertiary-care childrens hospital. Besides the 1 death already
discussed, 2 patients with poorly controlled type 1 diabetes
had sudden cardiorespiratory arrests at home and were
brought to the hospital but never regained consciousness; sudden death has been previously reported in patients with poorly
controlled type 1 diabetes.10 Thus, DKA may actually be an unusual cause of death in children with type 1 diabetes.
Our study is unique among single-center case series in its
size, enabled by the size of our hospital and its position as the
only hospital in a large metropolitan area admitting pediatric
patients with diabetes, and in the consistent use of a single
protocol in all venues within the hospital during the entire
study period. The total number of patients we included
exceeds that of a multicenter UK study3 and is more than
one-half the number of a multicenter US study2 that included
10 participating centers. This allowed many of the observed
differences in outcome between our center and these studies
to reach statistical significance (Table II).
At our hospital, DKA and cerebral edema are associated
with lower morbidity and mortality than reported in the
multicenter studies from the US and United Kingdom
(Table II).2,3 A comparison of our rate of cerebral edema
with the rates of those studies is complicated by different
definitions of both DKA and cerebral edema and by local
differences in the use and interpretation of diagnostic tests.
To be as conservative as possible in comparing our
outcomes with the outcomes of these studies,2,3 we used

Low Morbidity and Mortality in Children with Diabetic Ketoacidosis Treated with Isotonic Fluids

763

THE JOURNAL OF PEDIATRICS

www.jpeds.com

less stringent inclusion criteria and included patients with

cerebral edema who probably would not have been
reported elsewhere. For example, most patients with
significantly altered mental status underwent a CT scan in
our emergency department regardless of the presence or
absence of other signs of cerebral edema (eg,
ophthalmoplegia, bradycardia, or hypertension); indeed,
6% of all patients admitted with DKA had CT scans
performed. By contrast, only 21 of 34 patients with cerebral
edema in the UK study had a CT scan performed.3 An
additional 26 patients in that study had altered level of
consciousness without other signs of increased intracranial
pressure, did not have a CT scan performed, and were not
included as having cerebral edema. Moreover, the UK
study only included patients on the basis of radiographic
evidence if they had moderate or worse cerebral edema,
whereas we included all patients with any reported grade of
cerebral edema. If we eliminate our patients who received
no additional treatment and had mild radiographic
findings, 13 patients remain, for a prevalence of 0.3% (P =
.06 for the comparison with the UK prevalence). In the US
multicenter study,2 intubation per se was not considered
specific treatment for cerebral edema, and patients
receiving specific treatment for cerebral edema were
apparently included only if they improved after such
treatment. We believed that it would be difficult to
distinguish clinical improvement owing to ongoing DKA
therapy from that resulting from specific cerebral edema
interventions. Keeping these differences in mind (and
including all 20 of our patients with cerebral edema), the P
value for the difference in the prevalence of cerebral edema
between our center and that of the US study2 is, again, .06.
Future comparisons between studies would be aided by
a generally accepted clinical definition of cerebral edema.
Formal diagnostic criteria have been proposed6,11; 15 of
our 20 patients met these criteria (0.4% of cases with DKA).
To address possible biases in diagnosis of cerebral cases
within the first few hours of treatment, we examined the duration of treatment at diagnosis of cerebral edema. Most of
our patients with cerebral edema presented to medical care
with altered mental status, rather than deteriorating later
(Figure). Using the time of CT scan to define its onset, the
median time to develop cerebral edema in our series was
<2 hours versus 7 hours in each of the other studies. Only
2 of our patients presented at >7 hours. Thus, frequencies
of such patients are 2 of 3712 (0.06%) in our study versus
30 of 6977 (0.4%, P = .0007) and 10 of 2940 (0.3%, P =
.006) in the US2 and UK3 studies, respectively. The Dallas
and US multicenter cases also differ in hour of onset
(Figure; P = .00009). The incidence of late-presenting
patients seems more likely to be affected by differences in
treatment than were cases that were apparent on, or shortly
after, the patients arrival in the emergency department.
The altered mental status at presentation in many of our
patients may have been due to dehydration and acidosis12
rather than cerebral edema, which may have been an
incidental radiographic finding.13
764

Vol. 163, No. 3

In total, 3 of 3712 of our patients with DKA (0.08%) died or
were (mildly) disabled. In the US multicenter study,2 13 patients died, 8 had mild-moderate neurologic sequelae, and 4
had severe sequelae, for a total of 25 of 6977 (0.3%, P = .008).
In the UK study,3 50% of the patients with cerebral edema
died or were disabled; correcting for the oversampling of cerebral edema cases in their study design, we estimate that 10 of
2940 of the UK study patients with DKA (0.3%) died or became
disabled due to cerebral edema. An additional 6 deaths were
eliminated as not being due to in-hospital cerebral edema,
but the causes of death were not stated and the validity of these
exclusions for comparison purposes cannot be determined.
Although we and the US multicenter study2 (Nicole
Glaser, MD, University of California at Davis, personal communication via e-mail, December 11, 2012) ascertained DKA
cases similarly by using ICD-9 codes (the US study imposed
a pH criterion of <7.25 only for a nested case-control analysis), our study might have included a larger proportion of
mild cases. However, the differences between our center
and the US multicenter study in rates of death and disability
(P = .05), and of late-presenting cerebral edema cases (P =
.007) remained significant after reducing our denominator
by 30%, the proportion of a random sample of 100 patients
with DKA with pH $7.30 (Table I).
The risk of death or severe complications of DKA might be
further reduced by decreasing the incidence of DKA itself. A
relatively small number of our patients accounted for a disproportionate number of admissions with DKA, and these
appear to represent an important target population for interventions to improve compliance with diabetes management.
Our frequency of multiple admission rates are similar to
those reported in the UK study, although the UK study involved a much shorter (2-year) reporting period.3
Features of the Dallas protocol include consistent use of
isotonic fluids for the entire course of treatment, a constant
rate of fluid infusion (after an initial fluid bolus) of 2.0-2.5
times the maintenance rate with no calculation of degree
of dehydration, and a 3-bag system that facilitates controlled
resolution of hyperglycemia. The retrospective US and UK
cases series were not restricted to any particular protocol(s),
but, in comparison with the 2006 ADA consensus recommendations5 for a hypothetical 30-kg child with a 1-m2
body surface area (Table III), the Dallas protocol specifies
31% more fluid in the first 24 hours, 70% more sodium,
and 24% more potassium. Possible mechanisms by which
these differences might be beneficial could include quicker
correction of severely depleted intravascular volume and
thus faster restoration of normal central nervous system
perfusion. It has been suggested that cerebral edema in
DKA is a form of ischemiareperfusion injury, with edema
before or early in treatment being cytotoxic in nature (ie,
edema due to ischemic cellular dysfunction) and
vasogenic edema (caused by loss of vascular integrity)
occurring later.8,14 If this is so, the Dallas protocol may
decrease the duration or severity of the ischemic insult.
However, a recent small randomized trial of different rates
of fluid administration did not demonstrate any treatmentWhite and Dickson

ORIGINAL ARTICLES

September 2013

Table III. Comparison of fluid volumes and electrolytes specified by different protocols for a hypothetical 30-kg, 1-m2
patient
ADA 2006 consensus5

Childrens Medical Center Dallas

Time, h

% NaCl

Rate, mL/h

Volume, mL

Na, mEq

K, mEq

% NaCl

Rate, mL/h

Volume, mL

Na, mEq

K, mEq

0-1
1-4
4-24
Mean [cation], mEq/L
24-h totals

0.9
0.675
0.675
155

600
155
155

600
465
3100

92
54
358

0
18
123

0.9
0.9
0.45
130

300
125
125

300
375
2500

46
58
193

0
15
100

4165

504

175

3175

297

141

related differences using magnetic resonance diffusionweighted imaging to measure the apparent diffusion
coefficient of brain water (a measure of edema formation).8
Cerebral vascular thromboses may occur along with cerebral edema and are another cause of death and poor neurologic outcomes in DKA15,16; 1 of our 2 patients with
neurologic sequelae had such an event. Perhaps the higher
volumes of fluid provided by the Dallas protocol more
quickly reduce the hypercoagulable state associated with
DKA,17,18 thus decreasing the likelihood of thrombosis and
thereby lowering the complication rate. However, comparative data are lacking for the incidence of cerebral vascular
thrombosis in DKA.
The higher amounts of sodium correct sodium depletion
more quickly and presumably reduce the likelihood of
hyponatremia.19 Failure of serum sodium to increase with
treatment has been suggested as a risk factor for cerebral
edema,20,21 although the cause and effect relationship in this
association is uncertain; hyponatremia might be caused in
some cerebral edema cases by inappropriate antidiuretic hormone secretion triggered by increased intracranial pressure.22
Previous large studies failed to identify an effect of fluid
composition on cerebral edema risk2,3 (see review23), but,
based on recent consensus statements1,5,6 that recommend
either 0.45% or 0.9% NaCl, we suspect that relatively few patients were treated with nearly isotonic fluids. Whereas 0.45%
NaCl is hypotonic, even with added potassium, 0.9% NaCl is
hypertonic when 40 mEq/L potassium is added. A study using fluids with 100-125 mEq/L sodium also claimed a low
complication rate in 635 patients with DKA, although 5%
of those patients did receive mannitol, a rate much higher
than in the present study.24 We speculate that there may be
a relatively narrow optimum of tonicity for DKA resuscitation fluids and suggest that the tonicity of the fluid regimens
recommended in recent consensus statements may miss this
optimum on either side. Similarly, an ongoing randomized
trial (NCT01365793) of fluid regimens in DKA does not include an isotonic fluid arm and thus it will not be able to determine whether isotonic fluids are to be preferred.
There may be additional explanations that are not related
to the composition or rate of fluid administration. Consistent
use of a uniform protocol in a large childrens hospital with
hundreds of cases of DKA yearly may minimize management
errors; similar resources may not be available in many
venues.25 The 3-bag protocol minimizes delays in responding
to changes in blood glucose concentration; we previously re-

ported that instituting this protocol reduced the number of

ordered changes in fluid composition by
80%, and reduced
the mean time to resolve DKA from 16 to 12 hours.9
Our cases were collected later (1999-2011) than those in
the US (1982-1997) and UK (1995-1998) studies2,3; perhaps
DKA management has improved over time with the implementation of uniform protocols,1,5,6 thereby reducing morbidity and mortality. However, the US death rate for
children with diabetes who are <10 years old has remained
flat during 1995-2011, and the death rate has steadily increased between 1984 and 2011 in children 10 to 19 years
old.26 To the extent that DKA mortality contributes to these
death rates, there is no evidence for improved outcomes during the time periods in question.
We conclude that, although their relative importance cannot be determined from the available data, the main features
of the Dallas protocol appear to be collectively associated
with very low rates of death and disability from DKA. n
We thank the many physicians and nurses who have cared for patients
with diabetes at Childrens Medical Center Dallas. We are grateful to
Dr Thomas Abramo (Vanderbilt) for suggesting this study.
Submitted for publication Oct 30, 2012; last revision received Jan 15, 2013;
accepted Feb 4, 2013.
Reprint requests: Perrin C. White, MD, Department of Pediatrics, UT
Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9063.
E-mail: Perrin.white@utsouthwestern.edu

References
1. Dunger DB, Sperling MA, Acerini CL, Bohn DJ, Daneman D, Danne TP,
et al. European Society for Paediatric Endocrinology/Lawson Wilkins
Pediatric Endocrine Society consensus statement on diabetic ketoacidosis in children and adolescents. Pediatrics 2004;113:e133-40.
2. Glaser N, Barnett P, McCaslin I, Nelson D, Trainor J, Louie J, et al. Risk
factors for cerebral edema in children with diabetic ketoacidosis. The Pediatric Emergency Medicine Collaborative Research Committee of the
American Academy of Pediatrics. N Engl J Med 2001;344:264-9.
3. Edge JA, Hawkins MM, Winter DL, Dunger DB. The risk and outcome
of cerebral oedema developing during diabetic ketoacidosis. Arch Dis
Child 2001;85:16-22.
4. Rosenbloom AL. The management of diabetic ketoacidosis in children.
Diabetes Ther 2010;1:103-20.
5. Wolfsdorf J, Glaser N, Sperling MA. Diabetic ketoacidosis in infants,
children, and adolescents: a consensus statement from the American Diabetes Association. Diabetes Care 2006;29:1150-9.
6. Wolfsdorf J, Craig ME, Daneman D, Dunger D, Edge J, Lee W, et al. Diabetic ketoacidosis in children and adolescents with diabetes. Pediatr Diabetes 2009;10(Suppl 12):118-33.

Low Morbidity and Mortality in Children with Diabetic Ketoacidosis Treated with Isotonic Fluids

765

THE JOURNAL OF PEDIATRICS

www.jpeds.com

7. Duck SC, Wyatt DT. Factors associated with brain herniation in the
treatment of diabetic ketoacidosis. J Pediatr 1988;113:10-4.
8. Glaser NS, Wootton-Gorges SL, Buonocore MH, Tancredi DJ,
Marcin JP, Caltagirone R, et al. Subclinical cerebral edema in children
with diabetic ketoacidosis randomized to 2 different rehydration protocols. Pediatrics 2013;131:e73-80.
9. Felner EI, White PC. Improving management of diabetic ketoacidosis in
children. Pediatrics 2001;108:735-40.
10. Secrest AM, Becker DJ, Kelsey SF, LaPorte RE, Orchard TJ. Characterizing sudden death and dead-in-bed syndrome in type 1 diabetes: analysis
from two childhood-onset type 1 diabetes registries. Diabet Med 2011;
28:293-300.
11. Muir AB, Quisling RG, Yang MC, Rosenbloom AL. Cerebral edema in
childhood diabetic ketoacidosis: natural history, radiographic findings,
and early identification. Diabetes Care 2004;27:1541-6.
12. Edge JA, Roy Y, Bergomi A, Murphy NP, Ford-Adams ME, Ong KK,
et al. Conscious level in children with diabetic ketoacidosis is related
to severity of acidosis and not to blood glucose concentration. Pediatr
Diabetes 2006;7:11-5.
13. Glaser NS, Wootton-Gorges SL, Buonocore MH, Marcin JP, Rewers A,
Strain J, et al. Frequency of sub-clinical cerebral edema in children
with diabetic ketoacidosis. Pediatr Diabetes 2006;7:75-80.
14. Glaser NS, Marcin JP, Wootton-Gorges SL, Buonocore MH,
Rewers A, Strain J, et al. Correlation of clinical and biochemical
findings with diabetic ketoacidosis-related cerebral edema in children
using magnetic resonance diffusion-weighted imaging. J Pediatr
2008;153:541-6.
15. Ho J, Mah JK, Hill MD, Pacaud D. Pediatric stroke associated with new
onset type 1 diabetes mellitus: case reports and review of the literature.
Pediatr Diabetes 2006;7:116-21.
16. Rosenbloom AL. Hyperglycemic crises and their complications in children. J Pediatr Endocrinol Metab 2007;20:5-18.

766

Vol. 163, No. 3

17. Carl GF, Hoffman WH, Passmore GG, Truemper EJ, Lightsey AL,
Cornwell PE, et al. Diabetic ketoacidosis promotes a prothrombotic
state. Endocr Res 2003;29:73-82.
18. Bilici M, Tavil B, Dogru O, Davutoglu M, Bosnak M. Diabetic ketoasidosis is associated with prothrombotic tendency in children. Pediatr
Hematol Oncol 2011;28:418-24.
19. Neville KA, Sandeman DJ, Rubinstein A, Henry GM, McGlynn M,
Walker JL. Prevention of hyponatremia during maintenance intravenous
fluid administration: a prospective randomized study of fluid type versus
fluid rate. J Pediatr 2010;156:313-9.
20. Harris GD, Fiordalisi I, Harris WL, Mosovich LL, Finberg L. Minimizing
the risk of brain herniation during treatment of diabetic ketoacidemia:
a retrospective and prospective study. J Pediatr 1990;117:22-31.
21. Hoorn EJ, Carlotti AP, Costa LA, MacMahon B, Bohn G, Zietse R, et al.
Preventing a drop in effective plasma osmolality to minimize the likelihood of cerebral edema during treatment of children with diabetic ketoacidosis. J Pediatr 2007;150:467-73.
22. Rivkees SA. Differentiating appropriate antidiuretic hormone secretion,
inappropriate antidiuretic hormone secretion and cerebral salt wasting:
the common, uncommon, and misnamed. Curr Opin Pediatr 2008;20:
448-52.
23. Brown TB. Cerebral oedema in childhood diabetic ketoacidosis: is treatment a factor? Emerg Med J 2004;21:141-4.
24. Fiordalisi I, Novotny WE, Holbert D, Finberg L, Harris GD. An 18-yr
prospective study of pediatric diabetic ketoacidosis: an approach to minimizing the risk of brain herniation during treatment. Pediatr Diabetes
2007;8:142-9.
25. Edge J, Ackland FM, Payne S, McAuley A, Hind E, Burren C, et al.
Inpatient care for children with diabetes: are standards being met?
Arch Dis Child 2012;97:599-603.
26. Diabetes death rates among youths aged </=19 yearsUnited States,
1968-2009. MMWR Morb Mortal Wkly Rep 2012;61:869-72.

White and Dickson