Pediatric DKA

The presentation, assessment & prehospital management of diabetic ketoacidosis in children



Marc-David Munk, MD, MPH | | Thursday, October 23, 2008


Î Describe the clinical presentation of a pediatric patient with DKA.

Î Understand the physiologic changes taking place in a pediatric patient with DKA, particularly changes in acid-base status, dehydration and electrolyte shifts.

Î List common conditions predisposing a child to an episode ofpediatric DKA.

Î Explain the treatment priorities for a pediatric patient with DKA, including calculating pediatric resuscitation and maintenance fluid rates

Î Discuss the phenomenon of cerebral edema inpediatric DKA.

Paramedics are called to the home of a seven-year-old female for a reported asthma attack. According to the patient's father, the family returned from vacation the day before, and his usually healthy daughter wasn't feeling well. She hadn't been hungry for most of the day, but her father did note that she was drinking large amounts of water. When they came home, she went right to bed. This morning, she was difficult to rouse and was breathing very heavily. Fearing an asthma attack, he called 9-1-1.

Assessing the girl, the paramedics find a tachycardic, tachypneic seven year old. She has classicKussmaul breathing and is not responding to verbal stimulus. Suspecting a new-onset diabetes, the crew takes a glucometer reading. The glucose level is 530 mg/dL; they start an IV and begin to administer normal saline, per protocol.

Two hours later at the hospital, the patient is looking much better and is considerably more alert when she mentions to her father that she has a headache. Fifteen minutes later, she becomes unresponsive. The emergency physician suspects cerebral edema, immediately administers mannitol and rushes the patient to the CT scanner to confirm the diagnosis. She is admitted to the pediatric ICU where she steadily improves. One week later, she is discharged home with the diagnosis of new-onset type 1 diabetes mellitus, diabetic ketoacidosis (DKA) and cerebral edema secondary to DKA.


In children, diabeticketoacidosis generally results from an untreatedinsulin deficiency. DKA in children is typically secondary to untreated or undertreated type 1 diabetes mellitus, although DKA can also be seen in children with type 2 diabetes.

Type 1 diabetes is usually diagnosed during childhood or young adulthood. In type 1 diabetes, the body is unable to produce insulin because of an autoimmune destruction of the pancreas (the organ that produces insulin).

Type 2 diabetes is usually diagnosed in adulthood. In this form of diabetes, the pancreas makes insulin, but insulin production is insufficient to keep blood sugars normal. Alternatively, the body may be ˙resistantÓ to insulin; again, high blood sugar levels result. The termsinsulin-dependent andnon-insulin-dependent diabetes are no longer used but once referred to type 1 and type 2 diabetes, respectively. Insulin is administered to almost all patients with type 1 diabetes. Many patients with type 2 diabetes who have failed to respond to other forms of medication may also require insulin.

Diabetic ketoacidosis

When the body lacks insulin, it cannot process freely available glucose into usable energy, giving rise to the common saying that diabetes is ˙famine in the midst of plenty.Ó In the quest to provide energy to cells, a complex chain of metabolic processes ensues, including the production ofketones and acids. Ketones and acidemia in diabetics thus define DKA, althoughketosis is not strictly needed to make the diagnosis.

According to the 2004 consensus statement from the European Society for Pediatric Endocrinology, DKA can be diagnosed when a child demonstrates both hyperglycemia (defined as a blood glucose of > 200 mg/dL) and metabolicacidosis (defined as a venous pH < 7.30 and/or serum bicarbonate < 15 meq/L). In addition to hyperglycemia and acidosis, other abnormalities usually seen in children with DKA include ketosis, dehydration, electrolyte imbalances andhyperosmolarity. The clinical manifestations of DKA are related to the degree of hyperosmolarity, volume depletion and acidosis.

Ketones are the body's attempt to provide fuel for itself. In uncontrolled diabetes, there is a paucity of insulin; thus depleted of usable energy, the body produces glucagon and other stress hormones. These hormones stimulate the body to use glycogen for fuel, as well as proteins and fats. Lipolysis (fat burning), in turn, produces ketone bodies that are a usable, albeit inefficient, energy source, as well as ketoacids (e.g., acetoacetate and acetone). These acids are responsible for the fruity smell noted on the breath of diabetics with hyperglycemia. These ketoacids contribute to total body acidemia.

Worsening this acidosis, children with high glucose levels haveosmotic diuresis, which helps rid the body of excess glucose but results in dehydration, tissue hypoperfusion and ensuing lactic acidosis. A significant metabolic acidosis from ketoacids and lactic acid can thus be seen.

Children with DKA have potassium regulation problems as well. Most often, these patients have a total body potassium deficiency, although they may appear to have normal or even high potassium levels when measured. This is because the serum acids compete with potassium ions and push them from inside to outside the cell, where they are measured. When the acidosis is corrected and the potassium returns into cells, serum potassium levels can plunge to dangerously low levels.

Patients with DKA are bothhypoatremic andhypokalemic. Fluctuations of sodium and potassium affect every organ system. Most concerning are potassium's effects on the heart, where levels that are too high or too low can cause severe arrhythmias.

Finally, high levels of glucose and fluctuations in electrolyte levels cause overall high serum osmolarity levels. We'll talk about this in a moment, but differences in the tonicity between different parts of the body are dangerous because water shifts occur when electrolyte and glucose gradients between one area of the body and another are created. These water shifts have traditionally been thought to contribute to brain edema, although evidence for this is not absolutely clear.


The most difficult part of making a diagnosis of DKA is initially suspecting the condition in a child with symptoms but no history of diabetes. Symptoms of DKA are the presenting complaint in approximately 25% of new cases of type 1 diabetes. In many cases of DKA, there is a preceding trigger, usually infection; other risk factors include the onset of puberty or poor medication compliance in known diabetics.

Frequently, children will give a recent history of fatigue and malaise, nausea/vomiting, abdominal pain, thirst and frequent urination. Making a firm diagnosis of DKA in the prehospital setting is difficult, given the unavailability of lab testing. However, serum glucose can be measured and, in combination with such clinical findings as ketotic breath and Kussmaul breathing, may suggest the diagnosis.


In the child with DKA, multiple metabolic issues need to be corrected. Treatment goals include the management of dehydration, co-existent infectious triggers, hyperglycemia, potassium and other electrolyte abnormalities, acid-base abnormalities and osmolar gradients.

All children with suspected DKA should be placed on a cardiac monitor, if available, due to the arrhythmogenic effects of potassium and other electrolyte abnormalities. Any child in shock or with compromised circulation should receive supplemental oxygen. IV or IO access needs to be obtained. Children should receive no feedings or liquids by mouth, and during interfacility transports a Foley catheter should be placed to assess urine output status.

The initial step in management of a child with DKA is to ensure adequacy of ventilation. Remember that Kussmaul breathing is a normal physiologic response to acidosis, and should not by itself be considered evidence of respiratory distress.

The next priority should be correction of any hypotension and dehydration. Fluid repletion in moderate to severe DKA should start with 10 mL/kg over one hour, with the option of a second 10 mL/kg run if evidence of hemodynamic instability remains. The initial rehydration fluid of choice is normal saline solution. These children are often tachycardic and hypotensive from dehydration, and these indicators should guide fluid bolus therapy. All boluses of fluid in children need to be carefully measured; overhydrating too quickly can potentially cause serious problems.

After the initial fluid run, subsequent fluid drip rates are based on the patient's calculated volume deficit. It is critical that the initial boluses and subsequent fluid volumes be delivered using an IV pump orƒat a minimumƒwith very careful attention to drip rates using a microdrip set.

Although most emergency providers will have a child with DKA for only a short time, interfacility transfers may last several hours. In this case, knowing how to calculate rehydration rates is important. Most rehydration recommendations call for volume deficits to be corrected over 24Ï48 hours. Figuring out the volume to deliver over this period can be confusing, but generally the administered volume must accomplish two things: 1) correction of the child's dehydration and 2) satisfaction of the child's routine hourly fluid needs.

Here's an easy way to determine rehydration rates: Determine the fluid needed to correct the child's dehydration caused by the DKA. Then figure out the maintenance fluid rateƒthe fluid required to keep any child hydrated if not eating or drinking. The sum of these two hourly volumes is the hourly fluid rate.

Example: To start, you need to calculate the child's water deficit. Because determining percent dehydration in a child with DKA is clinically difficult, most sources assume that a child with DKA has anywhere from 7Ï10% deficit by weight. In the case of a 50 kg child, a 10% deficit _ 50 kg = 5 kg. Because 1 kg of water measures 1 L, a 5 L deficit exists.

Next, subtract the amount of fluid you gave in your initial 10 mL/kg run, in this case 500 mL. You have 4.5 L to replace.

How fast this is replaced is a matter of local preference. Some pediatric centers correct this deficit over 24 hours. Other centers prefer to rehydrate over 36 or 48 hours. In the above example, assuming a 24-hour period of correction, the hourly rate would be: 4,500 mL/24 hours = 187.5 mL/hour.

But we also need to calculate the child's maintenance fluid rate. A rule of thumb for this is to:

Î Give 100 mL/kg/day for each kg of the first 10 kg of body weight;

Î Give 50 mL/kg/day for each kg of the second 10 k of body weight; and

Î Give 20 mL/kg/day for each kg after that.

In the case of the 50-kg child: The first 10 kg _ 100 mL/kg = 1,000 mL. The second 10 kg _ 50 mL/kg = 500 mL. The remaining 30 kg _ 20 mL/kg = 600 mL.

Thus, the maintenance needs for this child are 2,100 mL over 24 hours, or 88 mL per hour.

Finally, add the rehydration hourly volume, and the maintenance volume is 187.5 mL/hr + 88 mL/hr = 275.5 mL/hr.

During interfacility transports, patients may come with insulin drips. Insulin is administered to correct hyperglycemia, reverse ketosis and acidosis, and prevent worsening dehydration. When running a drip is impossible, subcutaneous insulin and intramuscular insulin have been used to treat DKA. But this approach is far from ideal, mainly because insulin absorption from the tissues is erratic, particularly in severely ill children with impaired peripheral perfusion.

A few pointers when running insulin drips: Usually, an insulin drip will be set at anywhere from 0.05 to 0.10 unit/kg per hour. This means it needs to be run on an infusion pump to control the rate of administration. Before the drip is run, it's important to flush the insulin through the tubing circuit because the plastic in the IV line will initially bind insulin, decreasing the delivered amount. The drip should be piggybacked into the patient's intravenous line as close to the venous site as possible.

When running an insulin drip, it's important to do regular (half-hour or hourly) glucometer checks to ensure that the patient doesn't become hypoglycemic. If the blood glucose drops to less than 200 mg/dL, medical command should be contacted. Usually, medical command will request that dextrose be added to fluids to protect against hypoglycemia. D5 by itself is too hypotonic to use as a resuscitation fluid, and the administered fluid will usually be D5NS or even D5LR, per protocol. These fluids can be mixed in a pinch in the back of an ambulance using D50.

Remember that D50 is 50% dextrose; to get to a concentration of around 5% (D5) you will need to add 100 mL, or two 50 mL amps of D50 to a full 1 L bag of normal saline. Doing so will give you a dextrose solution containing 4.5% dextrose, or D4.5. (To make a 5% solution exactly, you would withdraw 100 cc of saline from the 1 L bag before adding the D50, but, in a stressful situation, this is time consuming and increases the risk of error; in most cases D4.5 is just fine for a few hours.) Remember, this is not D5W, or 5% dextrose in water; it's now 5% dextrose in 0.9% saline.

Keeping the insulin drip running when a child is in danger of becoming hypoglycemic may seem counterintuitive. With insulin drips, it's important to remember that hyperglycemia is, in fact, only one of a number of underlying metabolic issues that need to be corrected.

A frequent mistake is to follow the glucose concentration as a measure of improvement of DKA. When insulin is started, the first variable to correct will be hyperglycemia, but such issues as acidemia and electrolyte abnormalities will still persist because they correct less rapidly. Keep the drip going if the patient isn't frankly hypoglycemic, but add dextrose to the drip.

One final word about insulin drips:Be sure that the insulin drip has been mixed with short-acting insulin (usually regular insulin). Longer acting insulin, such as 70/30, lispro insulin and insulin glargine, has, on occasion, been accidentally used in a drip, and this form of insulin is not appropriate for DKA resuscitation.

In general, insulin will not be started until dehydration has been corrected, and often not until potassium has been administered. This is because when insulin is begun and acidosis is corrected, potassium will go from the extracellular space into the intracellular space. This sudden drop in extracellular potassium can lead to significant clinical hypokalemia, including cardiac arrhythmia as a complication. Often, just initial fluid rehydration will result in a drop in blood glucose due to dilution.

Although patients with DKA are by definition acidemic, multiple studies have shown that bicarbonate therapy should not be used in DKA. Trials have shown that administering bicarbonate will slow the rate of recovery of ketosis, worsen hypokalemia, worsen hyperosmolarity and may be a risk factor for cerebral edema. Only in cases of cardiac arrest or imminent arrest should bicarbonate be considered as a stop-gap measure to allow improved cardiac function in the setting of severe acidosis.

Cerebral edema

Cerebral edema is the most feared complication of DKA. Occurring in 0.3Ï1% of DKA cases, cerebral edema is primarily responsible for the current 0.15Ï5% mortality rate of patients with DKA. Although it has been reported in adults in a few isolated cases, cerebral edema due to DKA is an almost exclusively pediatric condition.

Cerebral edema most commonly occurs in the first 24 hours of rehydration. Often, onset will be subtle. Patients may complain of headache, and bradycardia may be notedƒboth of which are signs of increasing intercerebral pressure. Changes in neurological status, including restlessness, irritability, drowsiness and incontinence, will also be noted. Later symptoms include seizures,papilledema and respiratory arrest. Cerebral edema is a potentially catastrophic finding because it often leads to permanent neurologic sequelae.

Currently, the etiology of cerebral edema is unclear and quite controversial. It was traditionally believed that cerebral edema is related to overly aggressive management of DKA, including too-rapid correction of hyperglycemia and overly rapid administration of fluids. However, these assumptions have been called into question with recent trials.

In a large study published in 2001, the development of cerebral edema was significantly associated only with lower initial partial pressures of arterial carbon dioxide and higher initial blood urea nitrogen levels. The only treatment variable associated with cerebral edema was the use of bicarbonate. Other studies have found an association with overly aggressive hyperventilation (to a pCO < 22 mmHg). These findings suggest that severity of illness is more likely a predictor of cerebral edema than the treatment the patient received; however, cautious therapy as a means of mitigating the threat of cerebral edema is endorsed by every expert consensus statement.

Whether the rate of fluid administration is related to the development of cerebral edema remains unclear. Efforts by researchers to identify the ˙smoking gunÓ causing cerebral edema are still underway.

Currently, three prevailing theories attempt to explain the phenomenon of cerebral edema: 1) Some authors feel that damage of the blood/brain barrier endothelial lining leads to an increase in the amount of fluid/water diffusion across the blood/brain barrier once therapy is begun. 2) Overly aggressive fluid rehydration causes patients to become hyponatremic, leading to fluid shifts from the blood to the brain. 3) Damage to brain cells from dehydration and/or a lack of perfusion leads to brain injury and swelling of injured brain tissue.

Diagnosing cerebral edema is critical. Patients with this condition have a mortality of around 24%. Thirty-five percent of survivors have neurologic complications, including motor deficits, visual impairment, short-term memory loss, speech problems and convulsions. In one large published case series with 69 patients, the outcome was death in 64%, severe disability in 13%, mild disability in 8.6% and intact survival in only 14.5%.

During therapy for DKA, it's important that the patient be monitored for changes in mental status, particularly between four and 12 hours after the initiation of DKA therapy (the peak time for initial presentation). Children undergoing therapy for DKA should be observed carefully, and limited neuro checks need to be performed often. At a minimum, mental status, the presence of headache, changes in vision and developing nausea should be noted.

Treatment should be initiated if there is any concern that cerebral edema is developing. In any patient with suspected cerebral edema due to DKA, IV fluids should be stopped immediately. The head of the bed should be elevated to at least a 30_ angle. Mannitol, an osmotic diuretic, has been used to reduce cerebral edema and is currently the standard of care. Although mannitol has been shown to have beneficial effects in multiple case reports, convincing randomized trial evidence is still lacking. Current expert consensus is that the drug should be administered in cases of suspected cerebral edema. In consultation with medical command, IV mannitol should be given (0.25Ï1.0 g/kg over 20 minutes) in patients with signs of cerebral edema; this may be repeated in two hours if there is no initial response. In cases refractory to mannitol, 3% hypertonic saline has also been used successfully to treat patients with significant cerebral edema.

Careful observation of children with cerebral edema is key, because intubation and ventilation may become necessary from worsening intracerebral pressure. In the case of an intubated child with cerebral edema, hyperventilation isnot currently recommended. A large, multi-center study found that hyperventilation to low CO levels was associated with a poor outcome. Theories abound as to why this is so, but many clinicians feel that decreased cerebral blood flow from hyperventilation causes cerebral ischemia and, likely, worsened brain injury.


Pediatric endocrinologists have an old adage about DKA: Leave them a little bit dry and a little bit sweet, and give them lots of time. Children with DKA do not develop the condition overnight, and they can't be made metabolically normal overnight either.

Diabetic ketoacidosis itself is a fairly common condition. Many non-specific findings, such as changes in thirst, hunger and urination, may be the cardinal signs of a significant problem. DKA is often the first presentation of diabetes in children, and the diagnosis can often be suspected on the basis of clinical presentation and a simple glucometer reading in the field. Any child in whom diabetes is even suspected should have a glucometer reading taken.

Children with DKA may appear extremely ill. Treatment of suspected DKA should begin in the field with a cautious normal saline bolus. However, the initial administration of fluid must be extremely carefully managed and should be strictly limited to 10 cc/kg until discussed with medical command. Although no clear correlation can be drawn between the rate of IV fluid administration and the development of cerebral edema, the standard of care today is to proceed cautiously with correction of dehydration and metabolic derangements. Significant attention needs to be paid to IV drip rates.

Crews undertaking interfacility transport of children with DKA must be alert to the complicated physiologic changes underway in a child receiving therapy. In particular, electrolyte imbalances are likely, and hypoglycemia is possible. Crews should be familiar with calculating saline and insulin drip rates.

Last, prehospital providers must be diligent about performing frequent neurologic assessments in children with DKA. Any worsening mental status should raise concern for the possibility of cerebral edema, which should be treated immediately with head-of-bed elevation, mannitol IV and possible intubation.

Marc-David Munk, MD, MPH, is chief resident in emergency medicine at the University of Pittsburgh. He is a command physician with Pittsburgh EMS and a flight physician with STAT MedEvac.


1.Dunger DB, Sperling MA, Acerini CL. European Society for Paediatric Endocrinology/Lawson Wilkins Pediatric Endocrine Society consensus statement on diabetic ketoacidosis in children and adolescents.Pediatrics 2004;113:e133-e140.

2. Muir AB, Quisling RG, Yang MCK. Cerebral edema in childhood diabetic ketoacidosis: Natural history, radiographic findings, and early identification.Diabetes Care 2004;27:1541-1546.

3.Glaser N, Barnett P, McCaslin I. Risk factors for cerebral edema in children with diabetic ketoacidosis.The New England Journal of Medicine 2001;344:264-269.

4.Marcin JP, Glaser N, Barnett P. Factors associated with adverse outcomes in children with diabetic ketoacidosis-related cerebral edema.The Journal of Pediatrics 2002;141:793-797.

5.Edge JA, Hawkins MM, Winter DL. The risk and outcome of cerebral oedema developing during diabetic ketoacidosis.Archives of Disease in Childhood 2001;85:16-22.

6.Rosenbloom AL. Intracerebral crises during treatment of diabetic ketoacidosis.Diabetes Care 1990;13:22-33.

7.Roberts MD, Slover RH, Chase HP. Diabetic ketoacidosis with intracerebral complications.Pediatric Diabetes 2001;2:109-114.

Suggested Reading

Duck and Wyatt, 1998. Duck SC, Wyatt DT. Factors associated with brain herniation in the treatment of diabetic ketoacidosis.The Journal of Pediatrics 1998;113:10-14.

Felner and White, 2001. Felner EI, White PC. Improving management of diabetic ketoacidosis in children.Pediatrics 2001;108:735-740.

Fiordalisi, et al., 2002 Fiordalisi I, Harris GD, Gilliland MGF. Prehospital cardiac arrest in diabetic ketoacidemia: Why brain swelling may lead to death before treatment.Journal of Diabetes and Its Complications 2002;16:214-219.

Jeha and Haymond, 2006.Jeha G, Haymond MW: ˙Treatment and complications of diabetic ketoacidosis in children.Ó UpToDate Online. Accessed March 7, 2006.

Young, 2006.Young GM: ˙Pediatrics: Pediatric DKA.Ó Accessed March 6, 2006.

This continuing education activity is approved by the Center for Emergency Medicine, an organization accredited by the Continuing Education Board for Emergency Medical Services (CECBEMS), for 1.5 hours credit for First Responder, Basic and Advanced providers. If you have any comments regarding the quality of this program and/or your satisfaction with it, please contact CECBEMS by mail at CECBEMS, 5111 Mill Run Road, Dallas, TX 75244; by phone at 972/387-2862; by fax at 972/716-2007; or by

Connect: Have a thought or feedback about this? Add your comment now
Related Topics: Special Patients, Training

What's Your Take? Comment Now ...

Featured Careers & Jobs in EMS


Get JEMS in Your Inbox


Fire EMS Blogs

Blogger Browser

Today's Featured Posts


EMS Airway Clinic

The Evolution of Civilian High Threat Medical Guidelines

How mass killing events have proven a need for new guidelines.
More >

Multimedia Thumb

Baltimore Man Rescued from Building Collapse

Rowhouse collapse traps a worker in the basement area.
Watch It >

Multimedia Thumb

A Night with Wisconsin’s Busiest Medic Unit

Ride along one night with the paramedics of MED 5.
Watch It >

Multimedia Thumb

Patient Dies in West Virginia Ambulance Rollover

Marion County Rescue Squad crew is injured in collision.
Watch It >

Multimedia Thumb

Rescue Volunteers in Syria

White Helmets group at work during fighting.
More >

Multimedia Thumb

Boulder Pins Colorado Hiker

Wilderness EMS team frees trapped hiker.
More >

Multimedia Thumb

North Dakota Oilfield Medics

Tactics used in offshore platforms tailored to the remote areas.
More >

Multimedia Thumb

Field Bridge Xpress ePCR on iPad, Android, Kindle Fire

Sneak peek of customizable run forms & more.
Watch It >

Multimedia Thumb

Braun Ambulances' EZ Door Forward

Helps to create a safer ambulance module.
Watch It >

Multimedia Thumb

VividTrac offered by Vivid Medical - EMS Today 2013

VividTrac, affordable high performance video intubation device.
Watch It >

More Product Videos >