1139982

research-article2022
CPJXXX10.1177/00099228221139982Clinical PediatricsBonadio

Review Article
Clinical Pediatrics

The Evaluation and Management 2023, Vol. 62(6) 551­–564

© The Author(s) 2022
Article reuse guidelines:
of Pediatric Diabetic Ketoacidosis: sagepub.com/journals-permissions
DOI: 10.1177/00099228221139982
https://doi.org/10.1177/00099228221139982

A Comprehensive Review journals.sagepub.com/home/cpj

William Bonadio, MD1

Abstract
Diabetic ketoacidosis (DKA) is a common, serious acute complication in children with diabetes mellitus (DM).
DKA can accompany new-onset type 1 insulin-dependent DM, or it can occur with established type 1 DM, during
the increased demands of an acute illness or with decreased insulin delivery due to omitted doses or insulin pump
failure. In addition, DKA episodes in children with type 2 DM are being reported with greater frequency. Although
the diagnosis is usually straightforward in a known diabetes patient with expected findings, a sizable proportion of
patients with new-onset DM present with DKA. The purpose of this comprehensive review is to acquaint clinicians
with details regarding the pathophysiology, treatment caveats, and potential complications of DKA.

Keywords
diabetes, ketoacidosis, hyperglycemia, cerebral edema, insulin

Introduction Definition of DKA

Diabetic ketoacidosis (DKA) is the most common acute The presence of hyperglycemia (serum glucose concen-
complication requiring hospitalization of children with tration >200 mg/dL) with metabolic acidosis (venous
type 1 diabetes mellitus (DM).1-3 It is potentially fatal, pH <7.3 or serum bicarbonate [HCO3] concentration
accounting for 70% of DM-related deaths in children <15 mEq/L) and ketonemia or moderate/high ketonuria
younger than <10 years.3,4 Most DKA fatalities are is defined as DKA.3 The severity of DKA is determined
caused by cerebral edema (CE).3-5 by the degree of acidosis (Table 1).4
Successful emergency department (ED) management
of DKA requires timely intervention, meticulous moni-
Critical Appraisal of the Literature
toring, and protocol-based hour-by-hour therapy adjust-
ment aimed at ensuring tissue perfusion, correcting fluid A review of PubMed entries from 2010 to 2022 using
and electrolyte depletion, arresting ketogenesis, and the keyword pediatric diabetic ketoacidosis revealed
restoring normal cellular utilization and metabolism of limited new medical literature comparing the efficacy of
glucose. To successfully manage the crucial initial phase various protocols for managing DKA. Most protocols
of DKA, ED clinicians must thoroughly understand the are traditional and time-tested and are intuitively based
pathophysiology of DKA and the appropriate therapeu- on well-defined pathophysiologic considerations. They
tics and monitoring requirements of this complex disor- share the aims of carefully restoring metabolic homeo-
der, as well as maintain a keen awareness of potential stasis through rehydration, replacing electrolyte and
complications as the therapy proceeds. mineral deficits, insulin replacement to reverse ketogen-
The basic tenets of DKA management have changed esis, and diligent clinical monitoring.4
a little over the past 40 years. The cornerstone still con- Because the most serious complication of DKA—
sists of accurate and timely diagnosis, provision of sup- namely, CE—is uncommon, studies describing risk
plementary insulin and fluid rehydration, repletion of
electrolyte and mineral deficiencies, and meticulous 1
Mount Sinai Medical Center, New York, NY, USA
monitoring with frequent patient reevaluation.
Corresponding Author:
The purpose of this review is to discuss the pathogen-
William Bonadio, 20 East 84th Street, A2, New York, NY 10019,
esis, evaluation, and management of pediatric DKA in USA.
the outpatient setting. Email: william.bonadio@mountsinai.org
552 Clinical Pediatrics 62(6)

Table 1. DKA Classification of Degree of Severity. and the hyperglycemic hyperosmolar state (HHS).
Severe dehydration and hyperosmolarity are the charac-
Serum pH Serum HCO3 (mEq/L)
teristics of HHS that has a high risk of serious complica-
Mild 7.20–7.3 15–18 tions including coma and death. In contrast to DKA,
Moderate 7.0–7.20 10–15 with HHS, there is often sufficient insulin effect to pre-
Severe <7.0 <10 vent marked lipolysis and inhibit ketogenesis.
Abbreviations: DKA, diabetic ketoacidosis; HCO3, serum
bicarbonate. Newly Diagnosed IDDM
In the United States, it is not uncommon for children
factors have been largely elucidated by retrospective with newly diagnosed IDDM to present with DKA.10,11
multicenter case series6 and case reports.7,8 This is more common in younger children and is likely
the result of a delay in diagnosis.12-14 In a study of chil-
Epidemiology, Etiology, and dren with new-onset DM, nearly one-fourth presented
with DKA; specifically, 36% of children younger than 5
Pathogenesis years presented with DKA at the initial diagnosis com-
The most common metabolic/endocrine disorder of pared with only 16% of adolescents.12 Rates of DKA at
childhood is DM.4 The mechanism of acquired insulino- the time of DM presentation are even higher in develop-
penia is incompletely understood, although evidence ing countries.
points to an autoimmune destruction of pancreatic islet The symptoms of preverbal infants/toddlers with
beta cells in genetically predisposed individuals.9 The undiagnosed DKA are commonly misinterpreted as
prevalence of pediatric type 1 insulin-dependent DM respiratory disorders associated with tachypnea and are
(IDDM) is estimated at 2/1000 school-aged children.5 sometimes treated with beta-agonist bronchodilators
New cases most commonly occur in autumn and winter.4 and steroids, which can exacerbate DM metabolic
It is thought that symptoms of newly recognized type 1 derangements. As a result, symptom duration may be
DM manifest when the insulin-secreting reserve dimin- prolonged, leading to more severe dehydration and met-
ishes to <20% of normal. Symptoms include the “3 abolic acidosis and increased risk of neurologic compro-
P's”: polyuria, polydipsia, and polyphagia; patients can mise. A Canadian study documented that DKA at the
also experience weight loss and fatigue.4 The duration of time of DM presentation was significantly more com-
prediagnostic symptoms varies but is usually 1 to 2 mon in children younger than 3 years, and 39% had had
weeks.4 at least 1 prior health care visit (ie, a missed diagnosis)
Typically, DKA can manifest with type 1 DM when during the week prior to the DM diagnosis.14
>90% of pancreatic islet beta cells have been destroyed.
The overall risk of DKA in children with established
Established IDDM
type 1 DM is estimated to range between 1 and 10 per
100 person-years.2,5 In children with established type 1 DM, the risk of recur-
By contrast, the incidence of pediatric insulin-resis- rent DKA is 1% to 10% per patient per year.4 The
tant (type 2) DM has steadily risen, as it is associated increased risk of DKA is associated with the following
with epidemic rates of obesity. Some centers report type conditions15-17:
2 DM accounts for 50% of all newly diagnosed DM
cases in patients of the age range 10 to 21 years.10 This •• Poor metabolic control (reflected in higher hemo-
disorder typically presents in obese pubertal children of globin A1c [HgbA1c] levels)
minority ethnic background. At the time of type 2 DM •• Prior DKA
diagnosis, DKA is reportedly present in as many as 10% •• Female gender of adolescent age
to 25% of children.10 Type 2 disease is primarily due to •• Psychiatric disorders, including eating disorders
insulin resistance. Normally, insulin suppresses glyco- •• Unstable family circumstances
genolysis and gluconeogenesis in the liver; however, in •• Poor compliance with the insulin regimen
the setting of insulin resistance, the liver inappropriately •• Limited access to medical services and/or lack of
releases glucose into the blood. It can also be associated health care insurance
with relative insulinopenia due to insulin secretory •• Insulin pump utilization18
defects. The longstanding hyperglycemia can be exacer-
bated by a sudden increase in the demand for insulin While the most common cause of DKA in patients
during an intercurrent illness. Acute hyperglycemic with newly diagnosed DM is a delayed diagnosis, the
emergencies in children with type 2 DM include DKA most common cause in those with known DM is a lack
Bonadio 553

of insulin due to impaired delivery: either because of which can present with similar symptoms. The follow-
missed injections or, in patients who use an insulin ing conditions must be considered in the differential
pump, due to pump failure, prolonged pump disconnec- diagnosis of DKA:
tion without appropriate monitoring, or inappropriate
insulin-dosing reduction or discontinuation during an •• Other causes of metabolic acidosis with an anion
illness with reduced oral intake.19 In older children and gap (using the “MUDPILES” mnemonic: metha-
adolescents who self-manage their DM regimen and do nol intoxication, uremia, paraldehyde intoxica-
not receive adequate adult supervision, DKA commonly tion, isoniazid toxicity, iron intoxication, lactic
results from missed insulin doses. With patient educa- acidosis, ethylene glycol intoxication, salicylate
tion, careful supervision by parents or other responsible intoxication)
adults, and 24-hour access to professional telephone •• Asthma/respiratory insufficiency with respiratory
guidance, recurrent DKA should be preventable. acidosis
Stress due to an intercurrent febrile or vomiting ill- •• Sepsis
ness that results in relative insulin insufficiency is the •• Acute abdomen/appendicitis
common reason that precipitates DKA.20 Intercurrent •• Gastroenteritis with dehydration
failure to provide additional insulin in a timely fashion •• HHS
can lead to the development of ketoacidosis.
The diagnosis is usually straightforward in a patient
with known DM who presents with dehydration, meta-
Pathophysiology
bolic acidosis, and ketonuria.
Type 1 DM is a disease caused by destruction of pan-
creatic beta cells, leading to absolute insulin deficiency
Prehospital Care
resulting in abnormalities of carbohydrate, fat, and
protein metabolism. Relative or absolute deficiency of Prehospital care of suspected pediatric DKA involves
insulin causes DKA. With progressive beta cell failure, thorough assessment of airway, breathing, and circula-
a lack of insulin and increased levels of counterregula- tion; obtaining intravenous (IV) access; performing a
tory hormones (especially glucagon) result in unregu- bedside glucose measurement; and administering a con-
lated hepatic glucose production and decreased glucose servative fluid bolus volume (10 mL/kg) of normal
utilization in insulin-dependent tissues (muscle and saline (NS). Insulin should not be administered prior to
adipose tissue). Increased secretion of the counterregu- hospitalization. Expedited transport to a hospital ED,
latory hormones glucagon, epinephrine, norepineph- with frequent assessment of vital signs and neurologic
rine, cortisol, and growth hormone characterize DKA. status, is essential. In the patient with signs of respira-
This hormonal milieu causes a catabolic state, with tory insufficiency, securing the airway and providing
increased mobilization of long-chain fatty acids from adequate ventilation should take precedence.
adipose stores, augmented hepatic gluconeogenesis,
and increased oxidation of fatty acids, producing large
quantities of acetoacetic and beta-hydroxybutyric
ED Assessment/Management
acids. As shown in Figure 1, this cascade results in With only minor variation, outpatient management pro-
hyperglycemia and ketoacidosis. tocols are consistent, effective, and time-honored
Hyperglycemia and hyperketonemia induce an (Figure 2).
osmotic diuresis, resulting in total body fluid depletion Patients with DKA often present with nausea and
typically of the order 7% to 10% of body weight. vomiting. As DKA intensifies, patients become lethar-
Ketogenesis and poor tissue perfusion (resulting in lac- gic from dehydration, acidosis, and hyperosmolarity,
tic acidosis) from dehydration combine to produce met- and they can progress to coma. Other physical find-
abolic acidosis, which may be severe. If the metabolic ings typically include tachycardia, tachypnea, and
derangements are not arrested and reversed with rehy- Kussmaul respirations. Despite significant dehydra-
dration and replacement of insulin and electrolytes, fatal tion, paradoxical hypertension can be manifested at
dehydration and metabolic acidosis will ensue. some point during an episode of DKA in nearly 30%21
and may be due to a central nervous system (CNS)-
mediated mechanism causing abnormal hemodynamic
Differential Diagnosis regulation. Clinical risk factors for the development of
It is imperative to differentiate DKA from other causes hypertension include a more severe degree of acidosis,
of metabolic acidosis with an increased anion gap, all of having an underlying renal disease, and exhibiting
554 Clinical Pediatrics 62(6)

INSULIN DEFICIENCY

COUNTERREGULATORY HORMONE SECRETION [GLUCAGON,

CATECHOLAMINES, GROWTH HORMONE, CORTISOL]

LIPOLYSIS DECREASED GLUCOSE UTILIZATION BY INSULIN

DEPENDENT TISSUES [MUSCLE/ADIPOSE]

HEPATIC KETOGENESIS PRODUCTION OF

ACETOACETATE AND BETA-HYDROXYBUTYRATE HEPATIC GLUCONEOGENESIS

METABOLIC ACIDOSIS HYPERGLYCEMIA

OSMOTIC DIURESIS DEFICITS IN

+
DEHYDRATION [TISSUE FLUID VOLUME, Na , K+, Cl-, HCO3-,
HYPOPERFUSION/LACTIC ACIDOSIS] Ca+2, PO4-

Figure 1. Pathophysiology of DKA.

Abbreviation: DKA, diabetic ketoacidosis.

lower glasgow coma score (GCS) scores.21 Patients to the breath suggests the presence of acetone, formed
may also experience abdominal pain. Acidosis, along by the spontaneous decarboxylation of acetoacetate,
with poor splanchnic perfusion, can cause an intestinal that is excreted by the lungs. This mechanism removes
ileus, and patients frequently complain of abdominal about one fourth of the exogenous hydrogen ions gen-
pain that may mimic an acute abdomen. A fruity odor erated by hepatic ketogenesis.22
Bonadio 555

Admission/Hour #1:
• Assess vital signs, GCS score, circulatory status
• Establish IV access
• Perform EKG
• Measure lab tests: CBC, serum glucose/electrolytes/BUN/creatinine/Ca+2/PO4/osmolality; venous blood gas pH; urinalysis for
ketones; HgbA1c
• IV fluid 0.9% NaCl 10 mL/kg bolus infusion over 1 hour

Hour #2:
• Establish that criteria for DKA are met
• Reassess vital signs, GCS score, circulatory status
• Repeat measure of serum glucose/electrolytes/pH
• IV fluid: 0.9% NaCl with KCl 20mEq/L + KPO4 [or K-acetate] 20mEq/L, at 1.5 times daily maintenance infusion rate
• Regular insulin at 0.1 units/kg/hour by continuous drip infusion

Hour #3:
• Reassess vital signs, GCS score, circulatory status
• Repeat measure of serum glucose/electrolytes/pH
• Continue hour #2 IV fluida and insulin drip infusion

If serum pH <7.35 and HCO3 < 20 mEq/L: tolerated, If serum pH >7.35 or HCO3 >20 mEq/L: Give oral fluids; if
Continue hour #2 IV fluida and insulin drip and resolution of abnormal findings, discharge

Hour #4:
• Assess vital signs, GCS score, circulatory status
• Repeat measure of serum glucose/electrolytes/pH
• Continue hour #3 IV fluida and insulin drip infusion

If serum pH <7.35 and HCO3 <20 mEq/L: Hospitalize If serum pH >7.35 or HCO3 >20 mEq/L: Give oral fluids; if
tolerated and resolution of abnormal findings,discharge

Figure 2. Clinical pathway for outpatient management of pediatric DKA.

Abbreviations: BUN, blood urea nitrogen; CBC, complete blood count; CCS, glasgow coma score; EKG, electrocardiogram; DKA, diabetic
ketoacidosis; HgbA1c, hemoglobin A1c; IV, intravenous.
a
Add D5 to IV fluids when serum glucose concentration declines to a level of 250 to 300 mg/dL.

Laboratory Abnormalities of DKA usually accompanies DKA.23 A factitious hyponatremia

can be due to hyperglycemia, where the serum sodium
Although significant depletion of total body electrolytes measured is artificially lowered by approximately 1.6 to
and minerals occurs with DKA, serum levels of these 2 mEq/L for every 100 mg/dL increase in serum glu-
components often appear to be only minimally altered cose.24 Factitious hyponatremia can also be due to
on presentation. hypertriglyceridemia, as lipids have low sodium content
Despite the profound total body sodium depletion and, thus, dilute the measured moiety of serum sodium.
(8–10 mEq/kg) due to osmotic diuresis and renal excre- Children with DKA experience a total body potas-
tion of sodium-buffered anions, mild hyponatremia sium deficit averaging approximately 5 mEq/kg. Most
556 Clinical Pediatrics 62(6)

of the potassium is lost from the intracellular pool and is quality of skin turgor, respiratory pattern, adequacy of
then excreted by the kidney due to the osmotic diuresis oral mucosa hydration, pulse strength, and extremity
of potassium-buffered anions. Volume depletion causes perfusion. Severe dehydration is usually associated with
secondary hyperaldosteronism, which promotes urinary greater degrees of hyperglycemia. Dehydration >10%
potassium excretion. Despite total body potassium is suggested by the presence of marked hyperglycemia,
depletion, the serum potassium level at presentation weak peripheral pulses, widened pulse pressure, and
may be normal, increased, or decreased, depending oliguria.
largely on the severity of total body potassium depletion Ideally, vascular access is obtained via the placement
and the degree of acidosis. In addition, both insulin of 2 large-bore IV lines, with 1 line used for serial blood
administration and correction of acidosis can induce sampling. Blood should be sampled to measure serum
translocation of potassium back into cells, decreasing glucose; electrolytes; BUN/creatinine; osmolality;
serum levels as therapy progresses. Clinicians in the ED venous blood gas (pH/[pCO2]); complete blood count;
must anticipate an early decline in the serum potassium HgbA1c; and calcium, phosphorus, and magnesium
concentration with ongoing therapy. An electrocardio- concentrations. Appropriate specimens for culture
gram (EKG) (reflecting the myocyte intracellular K+ (blood, urine, throat) should be obtained if indicated.
concentration) can assess for U-waves and flattened Some centers have reported that point-of-care measure-
T-waves indicative of intracellular K+ deficit.25 ment of blood beta-hydroxybutyrate concentrations is
Although the serum PO4 level may initially be useful for confirming the diagnosis of DKA and moni-
normal, total body depletion of this mineral is toring the effectiveness of the therapy. Urine should be
common (4 mEq/kg) due to the osmotic diuresis. serially assessed for ketonuria, and a pregnancy test
Hypophosphatemia can be exacerbated by insulin should be performed on menstruating-aged female
therapy, which promotes entry of phosphate into patients. Continuous cardiorespiratory monitoring can
cells.26 In addition, there are deficits of chloride (5–7 be used when clinically indicated. Oxygen should be
mEq/kg), Ca+2, and Mg. administered to patients with hypoxia, severe circula-
Dehydration with decreased glomerular filtration tory impairment, or shock.
causes elevated blood urea nitrogen (BUN) and creati-
nine concentrations. Leukocytosis (consistent with
Other Diagnostic Studies
infection) and elevated serum levels of inflammatory
markers typically occur from increased stress hormone As previously mentioned, an EKG should be performed
release (epinephrine and cortisol).3 Pediatric DKA is on presentation as it reflects the degree of aberration in
only infrequently associated with infection; in 1 series, intracellular myocyte potassium concentration. Also, up to
only 18% of patients had a proven viral infection, and one third of children with DKA have been reported to
only 13% had a documented bacterial infection.27 exhibit prolonged QT interval corrected for heart rate on
presentation; especially prevalent in those with more pro-
found degrees of metabolic acidosis.28 In addition, patients
General Goals of DKA Therapy with an altered mental status require an emergency cranial
The goals of DKA therapy are to correct dehydration computed tomography (CT) scan to assess for evidence of
and acidosis, reverse ketogenesis, normalize serum glu- CE and also to ensure no other intracranial process, such
cose levels, and prevent complications. After expedited as stroke or intracranial hemorrhage, is present. The CT
triage, the ED assessment of DKA includes a thorough scan should only be performed after instituting an effec-
clinical evaluation to confirm the diagnosis and deter- tive resuscitative therapy to reverse CE.
mine its cause, including a careful search for the under- Management decisions, prognosis, and disposition
lying infection. A detailed history of outpatient are based on the clinical response and laboratory profile
management is important in recurrent DKA as factors as therapy proceeds.
such as insulin omission, failure to adjust to sick-day
regimens, or insulin pump failure account for most epi- Treatment: Fluid and Electrolyte
sodes. Compliance issues may be improved with appro-
priate psycho-educational intervention. Therapy
Airway, breathing, and circulation take precedence. The objectives of the fluid and electrolyte therapy are to
Vital signs should be measured and regularly reassessed. restore intravascular volume, to improve glomerular fil-
The level of consciousness can be serially quantified tration with enhanced serum clearance of glucose and
using the GCS score. The degree of dehydration is esti- ketones, and replenish depleted electrolytes and
mated by assessing vital signs, capillary refill time, minerals.
Bonadio 557

Fluid administration IV should be continued until the Prospective studies have not shown clinical benefits
acidosis has resolved, and the patient can tolerate orally from PO4 replacement.31-33 Yet total body stores of PO4
administered fluids. Fluid repletion is the single most are depleted with DKA, and this is an essential substrate
important initial intervention in managing DKA. Due to for cardiac muscle. One theoretical concern against rou-
osmotic diuresis, DKA patients have a total body fluid tine replacement is that it may lead to symptomatic
deficit of 5% to 10% of body weight. Administration of hypocalcemia. By contrast, uncorrected hypophosphate-
IV fluids corrects dehydration and restores adequate mia may detrimentally lower the levels of 2,3-diphos-
renal perfusion and glomerular filtration, which phoglycerate in red blood cells and reduce tissue oxygen
enhances glucose excretion, ameliorates lactic acidosis, delivery.31 Rhabdomyolysis and hemolytic anemia have
and decreases the serum level of counter-regulatory been rarely reported in cases of severe hypophosphate-
hormones.28 mia during DKA. Regardless of whether PO4 is replaced,
serum Ca+2 concentration should be monitored.
The “2-bag” system of fluid administration has been
Rehydration Rate and Fluid shown to be advantageous in augmenting faster response
Composition times to modifying the IV fluid therapy during DKA
The first step in rehydration is a bolus infusion of NS. treatment. With the 2-bag system, 1 infusion bag con-
Assuming patients with DKA are hyperosmolar due to tains D10 NS, and the other bag contains NS—equal
hyperglycemia and ketonemia, a gradual decline in amounts are initially administered, resulting in a con-
serum glucose and osmolarity is desirable. Patients with centration equivalent of 5% dextrose, and the subse-
an adequate blood pressure should receive a bolus IV quent ratio of each bag’s contribution was adjusted to
infusion of 10 mL/kg of NS during the initial hours of maintain normoglycemia.34
therapy. The NS bolus infusion may decrease the blood
glucose 20% solely by dilution and improving renal per- Insulin Therapy
fusion, leading to enhanced urinary glucose excretion.
In general, a “slow and steady” approach to rehydra- Insulin must be administered to resolve DKA. Regular
tion is desirable. After the initial hour of IV rehydration, IV insulin administration should be continued until the
subsequent parenteral fluid administration should con- acidosis has resolved and the patient can tolerate orally
tinue with 0.9% NaCl to replenish total body Na+ defi- administered fluids. Insulin therapy counteracts the
cit. Some protocols recommend exclusive use of the effects of glucagon in the liver (suppressing gluconeo-
0.9% concentration during the first 6 hours of therapy, genesis and ketogenesis), inhibits lipolysis and proteoly-
converting to 0.45% thereafter. The total volume admin- sis, and enhances glucose utilization in muscle and
istered over the next 47 hours should equal the sum of adipose tissues. Only regular insulin is used to treat
maintenance fluid volume (1500 mL/m2 body surface DKA. Following the initial bolus infusion of NS, insulin
area/day) added to a replacement volume to replace the therapy should be initiated as a continuous infusion at a
fluid deficit based on the degree of dehydration.4 The rate of 0.1 units/kg/h. At that point, fluid repletion will
total fluid volume administered rarely exceeds 1.5 to 2 have improved peripheral perfusion and partially cor-
times the usual daily fluid requirement. A prospective rected the metabolic acidosis, both of which will enhance
study showed that greater-volume IV-fluid infusion insulin efficacy. Ideally, insulin will slowly decrease the
rates were associated with more rapid resolution of met- serum glucose level approximately 50 to 100 mg/dL/h,
abolic acidosis, but there was no difference in the over- with associated gradual decline in serum osmolality. The
all length of required hospital treatment.29,30 insulin infusion should not be interrupted (eg, during
Once the serum K+ concentration is determined and transport to CT or inpatient ward) until the metabolic
urine output is established, K+ supplementation should acidosis is resolved, as the half-life of IV insulin is only
be considered. Because about two-thirds of patients will 6 minutes in blood and 30 minutes in tissue.35 A bolus
develop hypokalemia in the course of treatment for infusion of insulin at the outset of therapy does not has-
DKA, K+ repletion should commence once the serum ten the resolution of DKA and is not recommended at
K+ level drops below 5 mEq/L. Further adjustments are any point in management, to avoid undesired rapid shifts
based on hourly serial measurement of serum K+. in serum osmolarity and due to the potential for devel-
Potassium administration (total concentration of 40 oping hypoglycemia and hypokalemia.36,37
mEq/L; approximately double the usual maintenance There appears to be no difference in the rate of
requirement) can be given as 1/2 KCl and 1/2 KPO4 (or serum glucose decline when comparing IV short-acting
K-acetate) formulations to avoid Cl- overload, which insulin infusion administered at 0.1 units/kg/h versus
can exacerbate acidosis.4 0.05 units/kg/h.38,39
558 Clinical Pediatrics 62(6)

Repeated subcutaneous administration of short-act- phases of ED assessment and management. It is essen-

ing insulin has been shown to effectively resolve DKA tial to frequently (hourly) reassess vital signs, neuro-
metabolic abnormalities in those with mild-moderate logic status, and progression of abnormal clinical
DKA. One such regimen successfully administered 0.15 findings as DKA resolves. It enables the inpatient team
units/kg of subcutaneous regular insulin every 2 hours.40 to seamlessly continue protocol therapy. A bedside data
flow sheet should be used to enter hour-by-hour clinical
information and should include
NaHCO3 Therapy
Despite the sometimes profound metabolic acidosis, •• Vital signs
sodium bicarbonate (NaHCO3) therapy is not recom- •• GCS score
mended for patients with DKA.41-44 Controlled trials •• Laboratory values: serum glucose, pH/pCO2,
have shown no clinical benefit from NaHCO3 adminis- electrolytes, Ca+2, PO4-, qualitative urine
tration.45,46 Paradoxical cerebral acidosis can occur due ketones (until negative)
to the formation and diffusion of CO2 from the systemic •• Medications administered
circulation into the CNS.47 Rapid administration of •• Amount of insulin administered
NaHCO3 can cause intracellular influx of potassium and •• Fluid volume input/output.
induce acute hypokalemia.6,46,47 The NaHCO3 therapy
has been associated with CE, the most common cause of
Complications of DKA
mortality in children with DKA.6 There are 2 critical
instances in which NaHCO3 supplementation should Potential complications of DKA and their management
only be considered: (1) severe acidosis (serum pH < are listed in Table 2. The most frequent ones are hypo-
7.0) associated with myocardial depression and circula- glycemia and hypokalemia, the risks of which can be
tory insufficiency4 and (2) cardiac dysrhythmia due to minimized with frequent monitoring of serum levels and
hyperkalemia. adequate and timely replacement of glucose and K+,
As with treating hyperglycemia and hyperosmolarity, respectively. Acute kidney injury occurs in a relatively
the metabolic acidosis of DKA should be gradually large proportion of children with DKA. In 2 studies of
resolved with measured repletion of fluid and insulin pediatric DKA episodes, up to two-thirds of the patients
supplementation. Although HCO3 losses are large in had evidence of acute kidney injury, with most manifest-
DKA, the body replenishes a substantial amount during ing this disorder on presentation likely due to more
DKA treatment, as insulin stimulates HCO3 generation severe degrees of dehydration and acidosis.48,49 Finally,
from ketone metabolism. there have been case reports of pediatric DKA patients
who demonstrated transient myocardial dysfunction
associated with elevated serum troponin levels.50 A pro-
Ongoing Therapy posed mechanism is myocardial stunning due to severe
An important caveat in DKA management is that, typi- metabolic acidosis, resulting in an acute rise in intracel-
cally, the serum glucose concentration decreases to the lular calcium levels and damage to myocytes.51
normal range before ketosis and acidosis have resolved.4
To prevent hypoglycemia, 5% dextrose should be added
Cerebral Edema
to IV fluids when the serum glucose level declines to a
level of 250 to 300 mg/dL. If the serum glucose concen- Of all the complications of DKA, CE is the most serious.
tration continues to decline below the range of 150 to Clinically overt CE, although rare (<1% of DKA cases),
200 mg/dL despite the addition of 5% dextrose, the is a potentially devastating consequence of DKA ther-
appropriate maneuver is to administer a higher rate of apy, accounting for 60% to 90% of all DKA deaths.41
glucose infusion (in the form of >5% dextrose solution) The associated mortality rate of CE is 20% to 25%, and
instead of decreasing the rate of insulin infusion. approximately one fourth of survivors suffer lasting
Continued insulin infusion is essential to inhibit gluco- neurologic complications.6,52
neogenesis and ketogenesis, promote peripheral glucose Risk factors for developing DKA-associated CE
utilization, and correct metabolic acidosis. include53-55

•• Younger patient age

Monitoring •• New-onset DM
Meticulous clinical monitoring is the cornerstone of suc- •• Prolonged duration of symptoms
cessful DKA management, beginning in the initial •• Hypocapnia
Bonadio 559

Table 2. Potential Complications of DKA.

• Cerebral edema
 <1% of cases
 Associated with rapid rehydration or with overly aggressive rehydration or correction of acidosis/hyperglycemia
 Treatment: mannitol, 0.25–1 g/kg IV; 3% hypertonic NaCl, 5–10 mL/kg IV—some experts recommend initial therapy
with mannitol, followed by hypertonic NaCl if there is a lack of response; consider intubation with hyperventilation if
respiratory insufficiency is present.
• Hypoglycemia
 Due to the insufficient serum glucose concentration for insulin infusion administered
 Treatment: add 5%–10% dextrose to IV fluids when the serum glucose level is 250–300 mg/dL
• Hypokalemia
 Due to renal losses
 Exacerbated with resolution of dehydration and acidosis
 Treatment: add K+ to IV fluids when serum potassium concentration is <5.0 mEq/L and urine output is established.
• Cardiac dysrhythmia
 Due to hyperkalemia, hypokalemia, or hypocalcemia
 Treatment: correct specific imbalance
• Pulmonary edema/acute respiratory distress syndrome
 Due to increased pulmonary capillary permeability
 Treatment: administer oxygen and a diuretic
• Pancreatitis
 Typically subclinical (may be due to hypertonicity or hypoperfusion)
 Consider measuring serum amylase/lipase
• Rhabdomyolysis
 Due to muscle proteolysis
 Measure serum CPK
 Treatment: hydration and alkalinization of the urine
• Thromboembolism
 Enhanced by subclinical endothelial injury, hypofibrinolysis, and platelet aggregation, as well as elevated levels of
procoagulant factors
•• Acute renal failure

Abbreviations: CPK, creatine phosphokinase; DKA, diabetic ketoacidosis; IV, intravenous.

•• Relatively greater degree of hyperglycemia, aci- rapidly than that in the CNS, resulting in intracranial
dosis, or elevated BUN on presentation fluid shifts and CE. Evidence from clinical studies, how-
•• NaHCO3 therapy ever, suggests that this mechanism may not play a cen-
•• Attenuated rise in serum sodium concentrations tral role; various case reports have documented
during therapy DKA-induced symptomatic and even fatal CE in patients
•• Relatively greater fluid volumes administered in prior to receiving therapy.55-58 A more recent postulated
the initial 4 hours mechanism for DKA-related CE involves acidosis-acti-
•• Insulin administration in the initial hour of vated cell membrane sodium-hydrogen exchangers. An
therapy elevated hydrogen ion concentration promotes intracel-
lular sodium and water influx, with consequent
The mechanisms underlying the development of this edema.59,60 In addition, the ketone bodies beta-hydroxy-
disorder are poorly understood. There is evidence to butyrate and acetoacetate may affect vascular permea-
suggest that various aspects of DKA therapy may con- bility and contribute to edema formation.61
tribute to its development. Magnetic resonance imaging CE-associated DKA usually manifests during the ini-
findings suggest that DKA-related CE is vasogenic (dis- tial 3 to 12 hours of therapy, although onset can be ear-
rupted blood-brain barrier), not cytotoxic.56 A long- lier or later.52 The symptoms of CE may even be present
standing theory involves the interplay between CNS prior to initiating the DKA therapy.62-65 Nearly two third
hyperosmolarity and therapeutic fluid resuscitation in of cases occur within the first 7 hours of treatment, with
the susceptible patient. It postulates that CNS-protective the remaining cases occurring up to 30 hours after fluid
generation of idiogenic osmols helps to minimize the resuscitation and initiation of insulin treatment.4,66
osmolarity gradient between serum and CNS. During Subclinical CE, as detected on intracranial imaging,
treatment, serum hyperosmolarity may decrease more has been reported in majority of children with DKA,
560 Clinical Pediatrics 62(6)

even in those without neurologic signs or symptoms. 3% hypertonic NaCl at 5 to 10 mL/kg IV over 30 min-
Studies have demonstrated subtle cognitive functioning utes. A repeat dose can be given if necessary. With respi-
alterations including deficits in memory, attention, and ratory insufficiency requiring intubation, aggressive
verbal intelligence quotient in children with type 1 DM hyperventilation should be avoided as it has been associ-
who experience >1 episode of DKA versus those with- ated with worse neurologic outcomes. Cranial CT imag-
out a history of DKA.67 ing for suspected CE (those who display signs of altered
A randomized, controlled trial of children with DKA mental status, as evidenced by the diminishing GCS
compared differing volumes of initial fluid bolus admin- score) is recommended to confirm the diagnosis and to
istration (20 mL/kg vs 10 mL/kg) with several concen- rule out other etiologies of DKA-induced altered mental
trations of NaCl (0.45% vs 0.9%). They found no status (such as CNS thromboses or hemorrhage). Yet it
significant differences in the rate of decline in mental is important to emphasize that the diagnosis of CE is
status or clinically apparent brain injury during the treat- clinical, and if signs are present, immediate therapeutic
ment for DKA, or in neurocognitive function after intervention takes precedence over imaging. Bedside
recovery from DKA, between groups and concluded sonographic measurement of the optic nerve sheath
that the rate of rehydration fluid administration did not diameter as a surrogate marker for increased intracranial
significantly influence neurologic outcomes of children pressure has been shown to directly correlate with the
with DKA.66 Another study compared 2 different rehy- severity of DKA.70
dration protocols in children with DKA to assess the risk
of associated magnetic resonance imaging (MRI)-
Long-Term Cognitive and Cardiac
documented subclinical CE.68 From a total of 18 patients,
8 received more-rapid IV hydration (20 mL/kg initial Sequelae of DKA
volume of 0.9% NS IV bolus infusion, two third of fluid A study assessing DKA impact on subsequent cognitive
deficit replaced over initial 24 hours, with urine output abilities found that a single episode of moderate-severe
volume replacement) versus 10 who received slower IV DKA was associated with subtle memory decline, and
hydration (10 mL/kg initial volume of 0.9% NS IV bolus repeated episodes of DKA were associated with lower
infusion, fluid deficit evenly replaced over initial 48 IQ.71 It is reported that severe DKA initiates the synthe-
hours, and no urine output replacement). There was no sis of autoantibodies to cardiac antigens, which can lead
significant difference in rates of MRI changes consistent to the development of cardiomyopathy in young patients
with subclinical CE between the groups. with DKA.72
Up to 40% of initial brain imaging studies in children
with DKA and CE are normal.69 Therefore, CE compli-
cating DKA is a clinical diagnosis, and its presumptive Monitoring Progress
diagnosis should be based on clinical assessments. Signs Efficacy of DKA therapy is monitored by following
and symptoms of DKA-associated CE include hourly trends in vital signs, fluid balance, neurologic
status (GCS scores), and serum glucose/acid-base/elec-
•• Headache trolyte balances. Inhibition of ketogenesis, essential to
•• Vomiting resolving DKA, is reflected by decreasing serum glu-
•• Altered mentation/fluctuating level of consciousness cose and increasing serum pH/HCO3 levels with a nar-
•• Abnormal motor or verbal response to pain rowing of the anion gap. It is important to note that
•• Decorticate or decerebrate posturing resolving hyperglycemia alone does not denote resolu-
•• Cranial nerve palsy (especially cranial nerves III, tion of DKA; there must be a concomitant resolution of
IV, and VI) metabolic acidosis to signify the cessation of ketogene-
•• Abnormal neurogenic respiratory pattern (eg, sis. For this to occur, adequate quantities of insulin and
grunting, tachypnea, Cheyne-Stokes respiration, glucose must be administered, to arrest lipolysis and
apneusis) ketogenesis and decrease levels of counterregulatory
•• Bradycardia hormones.
•• Hypertension
Controversies and Cutting Edge:
If clinically indicated, a therapy to decrease intracra-
nial pressure should commence prior to confirmatory
Rate of Insulin Administration
imaging. The IV fluid infusion rate should be decreased Several prior studies compared lower-dose (0.05 units/
by 30%, and immediate infusion of mannitol should kg/h) versus standard-dose (0.1 units/kg/h) IV insulin
have commenced at 0.25 to 1 g/kg IV over 20 minutes or infusion for the initial treatment of DKA in children
Bonadio 561

with IDDM and noted equivalent efficacy.38,39 A possi- during this therapeutic interval. With this method of
ble advantage to lower-dose infusion is a more gradual “abbreviated” DKA treatment, more than 90% of
decline in hyperglycemia and, therefore, of serum osmo- selected patients will experience resolution of metabolic
larity, which could potentially decrease the risk of CE. acidosis and hyperglycemia.
Clinicians should always ensure that adequate amount Metabolic parameters on admission accurately pre-
of insulin is supplemented to effect resolution of keto- dict the outcome of patients with DKA during the initial
genesis or else DKA will continue. 3 hours of therapy. A study of 63 consecutive children
presenting to an ED with DKA found the following
results75:
Special Circumstances
Patients' demographics identifying a greater risk of 1. Of patients with admission serum pH >7.2 or
DKA-related CE include younger patient age, prolonged HCO3 >10 mEq/L, 94% experienced resolu-
symptoms, and presenting laboratory profile with (1) a tion of metabolic acidosis and hyperglycemia
greater degree of acidosis, (2) a greater elevation in within 3 hours of initiating the therapy, toler-
BUN, (3) hypocapnia, and (4) a blunted rise in serum ated oral feeding, and were discharged from
sodium concentration with the ongoing therapy. If any the outpatient setting. The rate of DKA relapse
of these factors is present, special attention to monitor- or other complication within 48 hours of dis-
ing and gradual therapy correction of DKA are charge was only 3%.
important. 2. By contrast, of patients with admission serum
pH <7.2 and HCO3 <10 mEq/L, 92% had per-
sistence of metabolic acidosis after 3 hours of
Hospital Management and therapy, necessitating hospitalization for fur-
Disposition ther treatment. The rate of DKA relapse or
The hospitalized child with DKA requiring a contin- other complications within 48 hours of dis-
uous insulin IV infusion should receive initial care in charge from the hospital was 4% (similar to
an intensive care unit setting, with experienced nurs- those nonadmitted patients).
ing staff trained in DKA monitoring and manage- 3. Of patients with initial serum pH >7.2 or HCO3
ment. Written guidelines for DKA management >10 mEq/L who were successfully treated as
should be followed to safeguard against error. outpatients, the admission serum glucose con-
Frequent and timely evaluation of biochemical trends centration was predictive of the duration of out-
is essential. A pediatric endocrinologist should be patient therapy necessary to resolve metabolic
consulted to help guide management and monitor for acidosis; 83% with serum glucose <500 mg/dL
complications. required 2 hours of therapy, whereas 78% of
Social factors can contribute to the risk of DKA patients with a serum glucose >500 mg/dL
relapse requiring repeat hospitalization. Odds for a required 3 hours of therapy. The longer duration
30-day readmission to the hospital for DKA have been of therapy associated with more marked hyper-
shown to be increased for female adolescents and glycemia may reflect a greater degree of dehy-
patients with depression and psychosis and lessened for dration, thus necessitating more prolonged fluid
those with private insurance or admitted to a teaching repletion to restore glomerular filtration.
hospital.73 In addition, each episode of DKA enhances
the odds of DKA recurrence.74
Summary
The goals of treating children with DKA are to (1) per-
Outpatient Management form rehydration to restore perfusion and increase glo-
The outpatient management of mild DKA75 outlined in merular filtration; (2) replace mineral and electrolyte
Figure 2 uses all the previously mentioned therapeutic deficiencies and provide insulin to arrest ketogenesis
measures for restoring glucose homeostasis and inter- and gluconeogenesis; and (3) increase cellular glucose
mediary metabolism to a normal state. With this metabolism in insulin-dependent peripheral tissues. The
approach, patients receive standard therapeutic mea- key to effective management of pediatric DKA includes
sures and close monitoring for up to 3 hours in an outpa- accurate diagnosis, timely therapy, and gradual correc-
tient setting. Based on the average deficits accrued by tion guided by meticulous monitoring. Children with
patients with DKA, approximately one-third of the fluid mild DKA can be safely and effectively managed with 2
deficit and one-half of the sodium deficit are replenished to 3 hours of therapy administered in an outpatient
562 Clinical Pediatrics 62(6)

setting, resulting in discharge with little risk of relapse 10. Neu A, Willasch A, Ehehalt S, et al. Ketoacidosis at onset
or other complications provided they have parental sup- of type 1 diabetes mellitus in children—frequency and
port and stable home situations. The metabolic parame- clinical presentation. Pediatr Diabetes. 2003;4(2):77-80.
ters of acid-base status on admission are accurate in 11. Klingensmith G, Tamborlane W, Wood J, et al. Pediatric
diabetes consortium. Diabetic ketoacidosis at diabetes
distinguishing those children with DKA who are candi-
onset: still an all too common threat in youth. J Pediatr.
dates for outpatient management.
2013;162(2):330-334.
12. Rewers A, Klingensmith G, Davis C, et al. Diabetic keto-
Author Contributions acidosis at onset of diabetes: the SEARCH for diabetes in
The author takes sole responsibility for the production of this youth study. Pediatrics. 2008;121(5):1258-1266.
manuscript. 13. Komulainen J, Kulmala P, Savola K, et al. Clinical, auto-
immune, and genetic characteristics of very young children
Declaration of Conflicting Interests with type 1 diabetes. Diabetes Care. 1999;22(12):1950-
1955.
The author(s) declared no potential conflicts of interest with respect 14. Bui H, To T, Stein R, Fung K, Daneman D. Is diabetic
to the research, authorship, and/or publication of this article. ketoacidosis at disease onset a result of missed diagnosis?
J Pediatr. 2010;156(3):472-477.
Funding 15. Rosilio M, Cotton J, Wieliczko M. Factors associated
The author(s) received no financial support for the research, with glycemic control. A cross-sectional nationwide
authorship, and/or publication of this article. study in 2,579 French children with type 1 diabetes.
the French Pediatric Diabetes Group. Diabetes Care.
ORCID iD 1998;21(7):1146-1153.
16. Smith CP, Firth D, Bennett S, Howard C, Chisholm P.
William Bonadio https://orcid.org/0000-0002-9389-6322 Ketoacidosis occurring in newly diagnosed and estab-
lished diabetic children. Acta Paediatr. 1998;87(5):537-
References 541.
1. Grinstein G, Muzumdar R, Aponte L, Vuguin P, Saenger 17. Rewers A, Chase H, Mackenzie T. Predictors of acute
P, DiMartino-Nardi J. Presentation and 5-year follow- complications in children with type 1 diabetes. JAMA.
up of type 2 diabetes mellitus in African- American 2002;287(19):2511-2518.
and Caribbean-Hispanic adolescents. Horm Res. 18. Hanas R, Linblad B, Lindgren F. Predisposing conditions
2003;60(3):121-126. and insulin pump use in a 2-year population study of pedi-
2. Glaser N, Jones K. Non-insulin dependent diabetes mel- atric ketoacidosis in Sweden. Diabetes. 2005;54(suppl
litus in Mexican-American children and adolescents. West 1):A455.
J Med. 1998;168(1):11-16. 19. American Diabetes Association. Type 2 diabetes in chil-
3. Wolfsdorf J, Craig ME, Daneman D, et al. Diabetic keto- dren and adolescents. Diabetes Care. 2000;23(3):381-
acidosis. Pediatric Diabetes. 2007;8(1):28-42. 389.
4. Wolfsdorf J, Glaser N, Sperling MA American Diabetes 20. Guillermo E, Umpierrez M, Smiley D, et al. Narrative
Association. Diabetic ketoacidosis in infants, children, review: ketosis-prone type 2 diabetes mellitus. Ann Intern
and adolescents: a consensus statement from the American Med. 2006;144(5):350-357.
Diabetes Association. Diabetes Care. 2006;29(5):1150- 21. DePiero A, Kuppermann N, Brown K, et al. Hypertension
1159. during diabetic ketoacidosis in children. J Pediatr.
5. Glaser N, Kuppermann N. The evaluation and management 2020;223:156-163.e5. doi:10.1016/j.jpeds.2020.04.066.
of children with diabetic ketoacidosis in the emergency 22. Lévy-Marchal C, Papoz L, de Beaufort C, et al. Clinical
department. Pediatr Emerg Care. 2004;20(7):477-481; and laboratory features of type 1 diabetic children at the
quiz482-4. time of diagnosis. Diabet Med. 1992;9(3):279-284.
6. Glaser N, Barnett P, McCaslin I, et al. Risk factors for 23. Katz M. Hyperglycemia—induced hyponatremia—cal-
cerebral edema in children with diabetic ketoacidosis. N culation of expected serum sodium depression. N Engl J
Engl J Med. 2001;344(4):264-269. Med. 1973;289(16):843-844.
7. Hanas R, Lindblad B, Lindgren F. Diabetic ketoacido- 24. Malone JI, Brodsky SJ. The value of electrocardiogram
sis and cerebral edema in Sweden—a 2-year population monitoring in diabetic ketoacidosis. Diabetes Care.
study. Diabet Med. 2007;24(10):1080-1085. 1980;3(4):543-547.
8. Marcin J, Glaser N, Barnett P, et al. Clinical and thera- 25. Riley MS, Schade DS, Eaton RP. Effects of insulin infu-
peutic factors associated with adverse outcomes in sion on plasma phosphate in diabetic patients. Metabolism.
children with DKA-related cerebral edema. J Pediatr. 1979;28(3):191-194.
2003;141(6):793-797. 26. Flood RG, Chiang VW. Rate and prediction of infection
9. Yoon JW, Jun HS. Autoimmune destruction of pancreatic in children with diabetic ketoacidosis. Am J Emerg Med.
beta cells. Am J Ther. 2005;12(6):580-591. 2001;19(4):270-273.
Bonadio 563

27. Owen OE, Licht JH, Sapir DG. Renal function and 41. Lever E, Jaspan JB. Sodium bicarbonate therapy in severe
effects of partial rehydration during diabetic ketoacidosis. diabetic ketoacidosis. Am J Med. 1983;75(2):263-268.
Diabetes. 1981;30(6):510-518. 42. Okuda Y, Adrogue HJ, Field JB, Nohara H, Yamashita
28. Perez M, Quigley L, Medar S, Clark BC. 163: QTC pro- K. Counterproductive effects of sodium bicarbon-
longation in hospitalized pediatric patients with diabetic ate in diabetic ketoacidosis. J Clin Endocrinol Metab.
ketoacidosis. Crit Care Med. 2020;48:64. doi:10.1097/01. 1996;81(1):314-320.
ccm.0000619012.77906.d8. 43. Hale PJ, Crase J, Nattrass M. Metabolic effects of bicar-
29. Waldhäusl W, Kleinberger G, Korn A, Dudczak R, bonate in the treatment of diabetic ketoacidosis. Br Med J
Bratusch-Marrain P, Nowotny P. Severe hyperglycemia: (Clin Res Ed). 1984;289(6451):1035-1038.
effects of rehydration on endocrine derangements and 44. Morris L, Murphy M, Kitabchi A. Bicarbonate ther-
blood glucose concentration. Diabetes. 1979;28(6):577- apy in severe diabetic ketoacidosis. Ann Intern Med.
584. doi:10.2337/diab.28.6.577. 1986;105(6):836-840.
30. Bakes K, Haukoos JS, Deakyne SJ, et al. Effect of vol- 45. Green SM, Rothrock SG, Ho JD, et al. Failure of adjunc-
ume of fluid resuscitation on metabolic normalization in tive bicarbonate to improve outcome in severe pediatric
children presenting in diabetic ketoacidosis: a random- diabetic ketoacidosis. Ann Emerg Med. 1998;31(1):41-48.
ized controlled trial. J Emerg Med. 2016;50(4):551-559. 46. Ohman JL, Jr, Marliss EB, Aoki TT, Munichoodappa CS,
doi:10.1016/j.jemermed.2015.12.003. Khanna VV, Kozak GP. The cerebrospinal fluid in dia-
31. Becker DJ, Brown DR, Steranka BH, Drash AL. betic ketoacidosis. N Engl J Med. 1971;284(6):283-290.
Phosphate replacement during treatment of diabetic keto- 47. Soler N, Bennett M, Dixon K, FitzGerald MG, Malins
sis. Am J Dis Child. 1983;137(3):241-246. JM. Potassium balance during treatment of diabetic keto-
32. Fisher JN, Kitabchi AE. A randomized study of phosphate acidosis with special reference to the use of bicarbonate.
therapy in the treatment of diabetic ketoacidosis. J Clin Lancet. 1972;2(7779):665-667.
Endocrinol Metab. 1983;57(1):177-180. 48. Hursh B, Ronsley R, Islam N, Mammen C, Panagiotopoulos
33. Burghen GA, Etteldorf JN, Fisher JN, Kitabchi AQ. C. Acute kidney injury in children with type 1 diabetes
Comparison of high-dose and low-dose insulin by con- hospitalized for diabetic ketoacidosis. JAMA Pediatr.
tinuous intravenous infusion in the treatment of diabetic 2017;171:e170020.
ketoacidosis in children. Diabetes Care. 1980;3(1):15-20. 49. Myers S, Glaser N, Trainor J, et al. Frequency and risk
34. Poirier MP, Greer D, Satin-Smith M. A prospective factors of acute kidney injury during diabetic ketoacidosis
study of the “two-bag system” in diabetic ketoacidosis in children and association with neurocognitive outcomes.
management. Clin Pediatr (Phila). 2004;43(9):809-813. JAMA Netw Open. 2020;3:e2025481.
doi:10.1177/000992280404300904. 50. Sakou, II, Soldatou A, Seretis A, Karanasios E, Paltoglou
35. Skjaervold NK, Lyng O, Spigset O, Aadahl P. G, Karavanaki K. Markedly elevated troponin and
Pharmacology of intravenous insulin administration: NT-proBNP and myocardial dysfunction in an adoles-
implications for future closed-loop glycemic control by cent with severe diabetic ketoacidosis: a case report. Clin
the intravenous/intravenous route. Diabetes Technol Ther. Pediatr Endocrinol. 2022;31(3):192-198.
2012;14(1):23-29. doi:10.1089/dia.2011.0118. 51. Botta K, Mugisha A, Berdaoui M, et al. Acute coronary
36. Chua H, Schneider A, Bellomo R. Bicarbonate in diabetic syndrome and diabetic keto acidosis: the chicken or the
ketoacidosis—a systematic review. Ann Intensive Care. egg? Ann Transl Med. 2019;7:397.
2011;1(1):1-23. 52. Marcin JP, Glaser N, Barnett P, et al. Factors associated
37. Lindsay R, Bolte RG. The use of an insulin bolus in with adverse outcomes in children with diabetic ketoacido-
low dose insulin infusion for pediatric diabetic keto- sis-related cerebral edema. J Pediatr. 2002;141(6):793-797.
acidosis. Pediatr Emerg Care. 1989;5(2):77-79. 53. Roberts J, Vavilala M, Schenkman KA, Shaw D, Martin
doi:10.1097/00006565-198906000-00001. LD, Lam AM. Cerebral hyperemia impaired cerebral
38. Puttha R, Cooke D, Subbarayan A, et al.Low dose (0.05 autoregulation associated with diabetic ketoacidosis in
units/kg/h) is comparable with standard dose (0.1 units/ critically ill children. Crit Care Med. 2006;34(8):2217-
kg/h) intravenous insulin infusion for the initial treatment 2223.
of diabetic ketoacidosis in children with type 1 diabetes-an 54. Edge JA, Hawkins MM, Winter DL, Dunger DB. The risk
observational study. Pediatr Diabetes. 2010;11(1):12-17. and outcome of cerebral oedema developing during dia-
39. Nallasamy K, Jayashree M, Singhi S, Bansal A. Low- betic ketoacidosis. Arch Dis Child. 2001;85(1):16-22.
dose vs standard dose insulin in pediatric diabetic keto- 55. Lawrence SE, Cummings EA, Gaboury I, Daneman D.
acidosis: a randomized clinical trial. JAMA Pediatr. Population-based study of incidence and risk factors
2014;168(11):999-1005. doi:10.1001/jamapediat- for cerebral edema in pediatric diabetic ketoacidosis. J
rics.2014.1211. Pediatr. 2005;146(5):688-692.
40. Razavi Z, Maher S, Fredmal J. Comparison of subcuta- 56. Figueroa RE, Hoffman WH, Momin Z, Pancholy A,
neous insulin aspart and intravenous regular insulin for Passmore GG, Allison J. Study of subclinical cerebral
the treatment of mild and moderate diabetic ketoacidosis edema in diabetic ketoacidosis by magnetic resonance
in pediatric patients. Endocrine. 2018;61(2):267-274. imaging T2 relaxometry and apparent diffusion coeffi-
doi:10.1007/s12020-018-1635-z. cient maps. Endocr Res. 2005;31(4):345-355.
564 Clinical Pediatrics 62(6)

57. Deeb L. Development of fatal cerebral edema during 67. Cameron FJ, Scratch SE, Nadebaum C, et al. Neurological
outpatient therapy for diabetic ketoacidosis. Pract Diab. consequences of diabetic ketoacidosis at initial presenta-
1989;6:212-213. tion of type 1 diabetes in a prospective cohort study of
58. Glasgow AM. Devastating cerebral edema in diabetic keto- children. Diabetes Care. 2014;37(6):1554-1562.
acidosis before therapy. Diabetes Care. 1991;14(1):77- 68. Glaser NS, Wootton-Gorges SL, Buonocore MH, et al.
78. Subclinical cerebral edema in children with diabetic keto-
59. Couch RM, Acott PD, Wong GW. Early-onset fatal acidosis randomized to 2 different rehydration protocols.
cerebral edema in diabetic ketoacidosis. Diabetes Care. Pediatrics. 2013;131(1):e73-e80.
1991;14(1):78-79. 69. Muir AB, Quisling RG, Yang MC, Rosenbloom AL.
60. Fiordalisi I, Harris GD, Gilliland MG. Prehospital cardiac Cerebral edema in childhood diabetic ketoacidosis: natu-
arrest in diabetic ketoacidemia: why brain swelling may ral history, radiographic findings, and early identification.
lead to death before treatment. J Diabetes Complications. Diabetes Care. 2004;27(7):1541-1546.
2002;16(3):214-219. 70. Şık N, Erbaş İM, Demir K, Yılmaz D, Duman M. Bedside
61. Kitabchi AE, Nyenwe EA. Hyperglycemic crises in dia- sonographic measurements of optic nerve sheath diameter
betes mellitus: diabetic ketoacidosis and hyperglycemic in children with diabetic ketoacidosis. Pediatr Diabetes.
hyperosmolar state. Endocrinol Metab Clin North Am. 2021;22(4):618-624. doi:10.1111/pedi.13188.
2006;35(4):725-751, viii. 71. Ghetti S, Kuppermann N, Rewers A, et al. Cognitive func-
62. Smedman L, Escobar R, Hesser U, Persson. Sub-clinical tion following diabetic ketoacidosis in children with new-
cerebral edema does not occur regularly during treatment onset or previously diagnosed type 1 diabetes. Diabetes
for diabetic ketoacidosis. Acta Paediatr. 1997;86:1172- Care. 2020;43(11):2768-2775. doi:10.2337/dc20-0187.
1176. 72. Hoffman WH, Sharma M, Cihakova D, et al. Cardiac
63. Isales C, Min L, Hoffman W. Acetoacetate and beta- antibody production to self-antigens in children and ado-
hydroxybutyrate differentially regulate endothelin-1 lescents during and following the correction of severe dia-
and vascular endothelial growth factor in mouse brain betic ketoacidosis. Autoimmunity. 2016;49(3):188-196.
microvascular endothelial cells. J Diabet Complications. 73. Bhatt P, Dave M, Amponsah JK, et al. Etiologies, trends,
1993;13(2):91-97. and predictors of 30-day pediatric readmissions after hos-
64. Hoffman WH, Steinhart CM, el Gammal T, Steele S, pitalizations for diabetic ketoacidosis in the United States.
Cuadrado AR, Morse PK. Cranial CT in children and Pediatr Diabetes. 2020;21(6):969-978. doi:10.1111/
adolescents with diabetic ketoacidosis. AJNR Am J pedi.13059.
Neuroradiol. 1988;9(4):733-739. 74. Hammersen J, Tittel S, Warncke K, et al. Previous dia-
65. Glaser NS, Wootton-Gorges SL, Buonocore MH, et al. betic ketoacidosis as a risk factor for recurrence in a large
Frequency of sub-clinical cerebral edema in children with prospective contemporary pediatric cohort: results from
diabetic ketoacidosis. Pediatr Diabetes. 2006;7(2):75-80. the DPV initiative. Pediatric Diabetes. 2021;22:455-462.
66. Kuppermann N, Ghetti S, Schunk J, et al. Clinical trial doi:10.1111/pedi.13185.
of fluid infusion rates for pediatric diabetic ketoacido- 75. Bonadio WA, Gutzeit M, Losek J, et al. Outpatient
sis. N Engl J Med. 2018;378:2275-2287. doi:10.1056/ management of diabetic ketoacidosis. Am J Dis Child.
NEJMoa1716816. 1988;142(4):448-450.