The symptoms of DKA usually develop rapidly over 1 day or less. DKA may be the initial presentation in up to 25% of people with newly diagnosed diabetes. Hyperglycemia is a key diagnostic criterion for DKA; however, a wide range of plasma glucose levels can be present on admission, and approximately 10% of DKA patients present with glucose <250 mg/dL (“euglycemic DKA”). Hyperosmolar hyperglycemic state (HHS) is often discussed as a separate condition. However, DKA and HHS represent two points on the spectrum of metabolic derangements in diabetes. In contrast to DKA, HHS may evolve insidiously over days to weeks. Symptoms of hyperglycemia in both DKA and HHS include polyuria, polydipsia, weakness, and weight loss.
Important factors to consider in the patient's past or current medical history include infection, myocardial infarction, pancreatitis, stroke, acromegaly, hyperthyroidism, and Cushing syndrome, as these may be precipitants or risk factors for DKA. In euglycemic DKA, pregnancy, starvation, concomitant alcohol use, and sodium-glucose cotransporter 2 (SGLT-2) inhibitors have all been implicated in its etiology.
It is essential to take a full medication history, in particular looking for corticosteroids, thiazides, pentamidine, sympathomimetics, second-generation antipsychotic agents, and immune checkpoint inhibitors, as these affect carbohydrate metabolism and may participate in the development of hyperglycemic crises. Cocaine abuse may be an independent risk factor associated with recurrent DKA. SGLT-2 inhibitors (e.g., canagliflozin, dapagliflozin, empagliflozin), used for glycaemic control of type 2 diabetes, have been the subject of an FDA warning about a risk for DKA.
Physical signs of volume depletion include dry mucous membranes, poor skin turgor, tachycardia, hypotension, and, in severe cases, shock. DKA patients may exhibit nausea, vomiting, Kussmaul respiration, acetone breath, and, occasionally, abdominal pain. Abdominal pain may correlate with the degree of acidosis in patients with DKA, and it may be confused with an acute abdominal crisis. Mild hypothermia may be observed in some patients, due to peripheral vasodilation. Hyperthermia is not usual, even in the presence of infection. Mental status may be altered in DKA, varying from alert in mild DKA to stupor and/or coma in severe DKA. In HHS, mental obtundation and coma are more frequent. Focal neurologic signs (hemianopia and hemiparesis) and seizures may also be features in HHS.
Initial laboratory evaluation
Plasma glucose is typically >250 mg/dL with presence of acidosis and ketonemia. However, a wide range of plasma glucose levels can be present on admission, and approximately 10% of DKA patients present with glucose <250 mg/dL (euglycemic DKA).
Positive for glucose and ketones. Other potential findings include leukocytes and nitrites in the presence of infection, and myoglobinuria and/or hemoglobinuria in rhabdomyolysis.
Arterial and venous blood gases
Arterial blood gas (ABG) shows a metabolic acidosis, which is essential for the diagnosis of DKA. Arterial pH measurement is necessary for diagnosis of DKA, but venous pH is recommended for monitoring treatment, due to the pain and risk of infection in obtaining frequent arterial samples. A venous pH sample is usually 0.03 units lower than arterial pH, and this difference should be considered.
The pH varies from 7.00 to 7.30, and the arterial bicarbonate ranges from <10 mEq/L in severe DKA to >15 mEq/L in mild DKA.
Capillary or serum ketones (beta-hydroxybutyrate)
There are three main ketones that are produced in DKA that can be measured: acetone, acetoacetate, and beta-hydroxybutyrate (BOHB).
In early DKA, the acetoacetate concentration is low, but it is a major substrate for ketone measurement by many laboratories (nitroprusside reaction method). Therefore, serum ketone measurement by usual laboratory techniques has a high specificity, but low sensitivity for the diagnosis of DKA; hence a negative test for serum ketones does not exclude DKA. Acetone is rarely measured due to its volatile nature. Conversely, BOHB is an early and abundant ketoacid that can be the first signal of the development of DKA. Point-of-care BOHB testing is widely available and is highly sensitive and specific for the diagnosis of DKA.
During the treatment of DKA, BOHB is converted to acetoacetate, which is detected by the nitroprusside method. Therefore, the increase in acetoacetate during the treatment of DKA may mistakenly indicate a worsening of ketonemia.
Another potential source of error in detecting ketone bodies is the patient's medications. Some drugs, such as the ACE inhibitor captopril, contain sulfhydryl groups that can react with the reagent in the nitroprusside test and give a false-positive result. Therefore, clinical judgement and other biochemical tests will be required in patients who are receiving such medications.
Typically increased due to volume depletion.
Sodium: serum sodium is usually low due to osmotic reflux of water from the intracellular to extracellular space in the presence of hyperglycemia. Total sodium deficit is 7 to 10 mEq/kg. Hypernatremia in the presence of hyperglycemia indicates profound volume depletion. Alternately, in the presence of high serum chylomicron concentration, pseudonormoglycemia and pseudohyponatremia may occur.
Calculation of the corrected sodium: the corrected serum sodium level should be evaluated as this is used to guide appropriate fluid replacement. The equation for conventional units is: corrected sodium (mEq/L) = measured sodium (mEq/L) + 0.016 (glucose [mg/dL] - 100).
Potassium: total potassium deficit is 3 to 5 mEq/kg. Serum potassium is usually elevated due to extracellular shift of potassium caused by insulin insufficiency, hypertonicity, and acidemia, but the total body potassium concentration is low due to increased diuresis. Therefore, low potassium level on admission indicates severe total-body potassium deficit.
Chloride: usually low. The total chloride deficit is 3 to 5 mEq/kg.
Magnesium: usually low. The total body deficit of magnesium is usually 1 to 2 mEq/kg.
Calcium: usually low. Total body calcium deficit is usually about 1 to 2 mEq/kg.
Phosphate: despite the total body phosphate deficit averaging 1.0 mmol/kg, serum phosphate is often normal or increased at presentation, but decreases with insulin therapy.
The calculated serum anion gap in mEq/L (serum sodium - [serum chloride + bicarbonate]) gives an estimate of the unmeasured anions in plasma, which in DKA are ketoacids. The anion gap is typically more than 10 to 12 mEq/L in DKA. [ Anion Gap ]
Normalization of the anion gap reflects correction of the ketoacidosis as these anions are removed from the blood.
Rhabdomyolysis is common in cocaine users with concurrent DKA, and creatine phosphokinase levels should be assessed in known or suspected cocaine users who present with DKA.
In rhabdomyolysis, pH and serum osmolality are usually mildly elevated and plasma glucose and ketones are normal. Myoglobinuria and/or hemoglobinuria are detected on urinalysis.
Measured to exclude lactic acidosis. Lactate levels are normal in DKA but elevated in lactic acidosis.
Liver function tests (LFTs)
Usually normal, and are used to screen for an underlying hepatic precipitant. Abnormal LFTs indicate underlying liver disease such as fatty liver, or other conditions such as congestive heart failure.
Serum amylase and lipase
Amylase is elevated in the majority of patients with DKA, but this may be due to nonpancreatic sources such as parotid glands.
Serum lipase is usually normal and may be beneficial in differentiating pancreatitis in patients with elevated amylase level. However, mildly elevated serum lipase level in the absence of pancreatitis has also been reported in patients with DKA.
This is variable in DKA.
CBC with differential
Leukocytosis is present in hyperglycemic crises and correlates with blood ketone levels. However, leukocytosis >25,000/microliter may indicate infection and requires further evaluations.
Used to exclude myocardial infarction (MI) as a precipitant or to look for cardiac effects of electrolyte disturbances (usually of potassium). Evidence of MI includes Q waves or ST segment changes. Evidence of hypokalemia (U waves) or hyperkalemia (tall T waves) may be present.
A high index of suspicion for MI should be maintained as diabetic patients often present with atypical symptoms.
Indicated to exclude pneumonia. In pneumonia, may show typical changes of pneumonia including infiltration, consolidation, effusions, and cavitation.
Blood, urine, or sputum cultures
Should be obtained if there are signs of infection such as chills, constitutional upset (e.g., fatigue, confusion, anxiety), or symptoms and signs of specific infections. The most common precipitating infections are pneumonia and urinary tract infections. Patients are usually normothermic or hypothermic due to peripheral vasodilation so fever may not be seen.
Usually normal, but are elevated if MI is the precipitant. A high index of suspicion should be maintained as diabetic patients often present with atypical symptoms.
Typical deficits in mild DKA
Total water (L): 6
Water (mL/kg): 100
Na+ (mEq/kg): 7 to 10
Cl- (mEq/kg): 3 to 5
K+ (mEq/kg): 3 to 5
PO4 (mmol/kg): 5 to 7
Mg++ (mEq/kg): 1 to 2
Ca++ (mEq/kg): 1 to 2.
Notes: Deficits based on per kg of body weight.
Venepuncture and phlebotomy: animated demonstration
How to perform an ECG: animated demonstration
Radial artery puncture animated demonstration
Femoral artery puncture animated demonstration
Use of this content is subject to our disclaimer