Case:

It’s your 5th straight night and your pediatric ED shift ends in one hour. A few stable patients came in during your shift, and you’re happy about that. No one wants kids to be sick. Your mind begins to wander. To be honest, the inner fat kid you’ve always been is really just craving pancakes at this point after you leave work.

An attending comes to visit from Adult ED and asks “Breakfast?!”  followed by “How was your night?” You say the Q word, “quiet”, because you’re not superstitious like that. You hear lightning strike outside as everyone’s mouth drops open, nearby staff turn to glare at you, a flock of birds begins to head to a different city, and the EMS notification phone rings twice in a row. Way to go.

A 14-year-old boy is brought into the ED by EMS with altered mental status. He has no past medical history. Dad says the patient has not been feeling well over the last few days and he has been having fevers as well as cough with green sputum. This morning he was difficult to arouse for school and Dad called 911.

The heart rate is 130 with a respiratory rate of 33 and a pressure of 98/62. The patient is responsive to verbal stimuli, there is a fruity odor on his breath, and he has dry mucous membranes. He has rapid, deep breathing, tachycardia, and the bedside fingerstick glucose reads: HIGH

The basic metabolic panel results appear on your monitor: Na 129 mmol/L, K 4.0 mmol/L, Cl 96, HCO3 7, BUN 39, Cr 1.1, and glucose 600. He has positive ketones in his urine and serum beta-hydroxybutyrate is twice normal limits.

 

What is the pathophysiology?

Diabetic ketoacidosis, or DKA, occurs when the body is relatively deficient in insulin leading to counter-regulatory hormone effects which produces hyperglycemia and ketonemia. Ketonemia results in a metabolic acidosis which leads to vasodilation, myocardial depression, and Kussmaul breathing for respiratory compensation. Fruity odor on the breath is caused by acetone, one of the ketone bodies produced in DKA (1,5).

The leading cause of mortality in patients less than 24 years old with diabetes is DKA, most often resulting from cerebral edema (2).

 

Why do diabetics get DKA?

Common causes of DKA are infection (most common), myocardial infarction, medication non-compliance, hyperthyroidism, occlusion of insulin pump catheter, pregnancy, substance abuse, and drugs that affect carbohydrate metabolism. These include corticosteroids, pentamidine, sympathomimetics, thiazides, and antipsychotic medications (1,5).

 

What are the clinical features?

 

Patients may present with tachycardia, hypotension, signs of dehydration, and electrolyte abnormalities secondary to osmotic diuresis. Hyperkalemia may be present despite likely total body potassium depletion but DKA may also present with either normal potassium or hypokalemia, both requiring potassium repletion. Those who present with normal or low potassium levels have marked whole body depletion and warrant aggressive repletion of potassium. Patients may experience nausea, vomiting, or abdominal pain which may help identify undiagnosed diabetics, especially children (5).

 

How is it diagnosed?

 

The diagnosis is made based on suspected insulin deficiency, ketones in the urine or serum, and metabolic acidosis. Though not mandatory, glucose greater than 250, bicarbonate less than 15 mmol/L, pH less than 7.3, and moderate ketonemia are consistent with the diagnosis of DKA. Ketone body formation produces metabolic acidosis with an increased anion gap (1,5).

Image courtesy of Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic Crises in Adult Patients With Diabetes. Diabetes Care 2009;32(7):1335–43.

 

You diagnose the patient with DKA. The nurse is proactive and asks if you would like to start insulin. You remember potassium cannot be too low before starting insulin but is it high enough? Do they need K repletion?

 

How is DKA treated?

 

Airway, breathing, and circulation should be stabilized followed by initiation of fluid resuscitation and subsequent intravenous fluid maintenance and electrolyte repletion. Normal saline versus lactated ringers versus other fluids in resuscitation is still debatable. See this post for more discussion.

 

Image courtesy of Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic Crises in Adult Patients With Diabetes. Diabetes Care 2009;32(7):1335–43.

IV insulin therapy should be started based on the anion gap, potassium level, and glucose level. When patients in DKA have potassium less than 3.3 mmol/L, potassium should be replaced prior to initiating insulin therapy as insulin administration will further lower extracellular potassium. At a potassium between 3.3 mmol/L and 5.2 mmol/L (or the upper limit of normal for a particular laboratory) insulin therapy should be started at 0.1 U/kg/hr. Patients in this potassium range should receive potassium repletion at 20 to 40 mEq with each liter of fluid. Frequent electrolyte and glucose evaluation should be obtained. Glucose often corrects faster than the metabolic gap acidosis. Therefore, at glucose of 200-250 mg/dL, insulin drip rate should be reduced and patient should receive D5 ½ NS intravenous fluid instead of the initial resuscitation fluid, with a glucose goal in the 200s. Continuation of insulin prevents continuation of ketone body production, so insulin infusion should be tapered and subcutaneous rapid-acting insulin schedule should begin when glucose is less than 200 and two of three criteria are met: venous pH greater than 7.30, bicarbonate greater than or equal to 15 mEq/L, and serum anion gap less than 12 mEq/L (5).

 

Cerebral edema is a complication of treatment that tends to occur 4 to 12 hours into treatment and seen predominantly in young children with new-onset diabetes. It should be suspected if neurologic status deteriorates after treatment. 1g/kg mannitol should be administered before diagnostic CT1 There is some evidence that neither the sodium chloride concentration of intravenous fluid, whether normal saline or hypertonic, nor the rate of infusion of fluids influence neurologic outcomes in children with diabetic ketoacidosis (3,4)

 

This is primarily a board review post and is not meant to be an up to date literature review.

 

References

  1. Kefer MP. Diabetic ketoacidosis. In: Tintinalli JE, Cline DM. Tintinallis emergency medicine manual. 7th ed. New York: McGraw-Hill Medical; 2012: 628-630
  2. Vella A. The diabetic child. In: Tintinalli JE, Cline DM. Tintinallis emergency medicine manual. 7th ed. New York: McGraw-Hill Medical; 2012: 380
  3. Shafi O, Kumar V. Initial Fluid Therapy in Pediatric Diabetic Ketoacidosis: A comparison of Hypertonic Saline Solution and Normal Saline Solution. Pediatric Endocrinology Diabetes and Metabolism 2018;24(2):56–64.
  4. Kuppermann N, Ghetti S, Schunk JE, et al. Clinical Trial of Fluid Infusion Rates for Pediatric Diabetic Ketoacidosis. New England Journal of Medicine 2018;378(24):2275–87.
  5. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic Crises in Adult Patients With Diabetes. Diabetes Care 2009;32(7):1335–43.

 

The following two tabs change content below.

aaurreco

Latest posts by aaurreco (see all)


1 Comment

Dr. Paul Pukurdpol · October 17, 2018 at 2:51 am

Excellent review of DKA and potassium management Dr. Aurrecoechea! I noticed that you cited a common adult DKA management algorithm in your discussion. Your case, however, was a 14 year old, and there are some important differences in the management of adult vs pediatric DKA. The most important difference is that, while bolus dosing of insulin is a common practice in adult DKA (often due to Type II DM), bolus dosing of insulin has been found to be harmful in children, as they are often exquisitely sensitive to insulin and are at greater risk of hypoglycemia.

In addition to being gentle with insulin, we should also be more gentle with fluid resuscitation, in order to prevent cerebral edema. A June 2018 NEJM article by Kuppermann et al from the PECARN group found that it was safe to give a “rapid” infusion of NS or 1/2 NS, however “rapid” was defined as two 10 cc/kg boluses in the first hour. That is to say ONE standard 20 cc/kg bolus of NS, followed by twice maintenance, constitutes “rapid” fluid resuscitation in pediatric DKA – way less than we are used to giving in adults. Notably, you could also go “slow” and just give one 10 cc/kg bolus, and outcomes were essentially unchanged, which is the route CHOP chose for their DKA protocol (see https://www.chop.edu/clinical-pathway/diabetes-type1-with-dka-clinical-pathway). Children just don’t seem to get as hypovolemic as adults do.

Leave a Reply

Avatar placeholder

Your email address will not be published. Required fields are marked *

%d bloggers like this: