ED Management of DKA

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).

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).

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).

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.

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)