Pediatric Diabetic Ketoacidosis (DKA) Workup

Updated: Dec 22, 2020
  • Author: William H Lamb, MD, MBBS, FRCP(Edin), FRCP, FRCPCH; Chief Editor: Dale W Steele, MD, MS  more...
  • Print

Approach Considerations

The following lab studies are indicated in patients with diabetic ketoacidosis:

  • Blood glucose

  • Blood gases

  • Potassium

  • Sodium

  • Blood urea and creatinine

  • Bicarbonate - Usually available from blood gas analysis

  • Capillary blood ketone

  • Glycosylated hemoglobin (HbA1c)

  • Serum osmolarity

Additional laboratory studies are often measured but have limited value in management without specific indication

  • Full blood count

  • Cultures

  • Insulin

  • Phosphate, calcium, and magnesium levels

  • Amylase levels

  • Lipids

Imaging studies

Perform head computed tomography (CT) scanning if coma is present or develops. Concurrently, initiate appropriate measures to manage cerebral edema. Perform chest radiography if clinically indicated.


Check the patient’s consciousness level hourly for up to 12 hours, especially in a young child with a first presentation of diabetes. The Glasgow coma scale (see the image below) is recommended for this purpose.

Pediatric Diabetic Ketoacidosis. Glasgow Coma Scal Pediatric Diabetic Ketoacidosis. Glasgow Coma Scale, modified for age of verbal response.

The normal maximum score on the Glasgow coma scale is 15. A score of 12 or less implies significant impairment of consciousness. A falling score may signify the development of cerebral edema.


Laboratory Studies

Blood glucose

Capillary blood samples analyzed on any modern blood glucose meter are acceptable for monitoring changes in blood glucose levels as treatment progresses, but measure at least 1 whole blood glucose at presentation.

Check blood glucose at least hourly during the initial stages of treatment (more frequently if blood glucose levels fall quickly or if changes to insulin infusion rates are made).

Blood gases

Traditionally, arterial blood samples are used; however, free-flowing capillary or venous samples are as reliable as the arterial samples for monitoring acidosis, are much easier to collect, and are less traumatic for the child. [41]

The severity of diabetic ketoacidosis can be defined by blood gas results, as follows:

  • Mild diabetic ketoacidosis - pH level of less than 7.3, bicarbonate level of less than 15 mmol/L

  • Moderate diabetic ketoacidosis - pH level of less than 7.2, bicarbonate level of less than 10 mmol/L

  • Severe diabetic ketoacidosis - pH level of less than 7.1, bicarbonate level of less than 5 mmol/L


Initial blood potassium levels are usually normal or high, despite considerable deficits of total body potassium. This is because the acidosis encourages leakage of intracellular potassium. Insulin drives potassium back into the cells, and levels may drop very quickly with treatment.

Frequent checks of potassium levels (ie, every 1-2 h), together with electrocardiographic monitoring, may be required in the first hours of therapy.


Measured sodium values are likely to be low because of the dilutional effect of hyperglycemia. True sodium levels can be calculated by adding 1.6 mEq/L sodium for every 100 mg/dL glucose (ie, 1 mmol/L sodium for 3 mmol/L glucose).

Sodium levels should rise with treatment. Failure of sodium levels to rise is associated with an increased risk of cerebral edema.

Blood urea and creatinine

Some creatinine assays can be affected by the presence of ketones, thus giving falsely elevated results. Under these circumstances, blood urea may give a better measure of dehydration.

Capillary blood ketone

Capillary blood ketone can be measured using a handheld meter; the level is always elevated at presentation of diabetic ketoacidosis (>2 mmol/L). Two studies have proposed using serial measurements as a way of indicating the resolution of diabetic ketoacidosis when the pH level is more than 7.3 and the sequential capillary blood ketone level is less than 1 mmol/L. [42, 43]


Insulin levels may be indicated in children with recurrent diabetic ketoacidosis, as an absence of measurable insulin can be used to emphasize that omission is probable. Caution is needed because not all assays measure the newer analogue insulins; insulin antibody levels can also affect the result.

Additional studies

Other lab studies include the following:

  • Bicarbonate (usually available from blood gas analysis) - This reflects the degree of acidosis

  • Glycosylated hemoglobin (HbA1c) - High results are expected in a patient with newly diagnosed diabetes and in patients with an established diagnosis who have poor compliance with treatment. A normal result may be seen where acute failure of insulin delivery from an insulin pump results in rapid onset DKA.

  • Urine - Check all urine for glucose and ketones for at least 24 hours, particularly if capillary blood ketones are not available

  • Full blood count - The white blood cell (WBC) count is usually elevated, even in the absence of infection

  • Culture - Perform blood culture and other cultures if clinically indicated in the presence of fever or symptoms (eg, urine, throat swab)

  • Amylase - Blood amylase levels often are elevated in diabetic ketoacidosis and can be misleading in the presence of abdominal pain

  • Serum osmolarity - This is usually elevated

  • Phosphate, calcium, and magnesium - These levels are invariably reduced but without obvious clinical significance

  • Lipids - Extremely high triglyceride levels are sometimes present; this causes an artificial lowering of other blood values, such as those for glucose, sodium, and potassium, but abnormal lipids are almost an inevitable consequence of DKA



Ideally, insert a good-sized venous cannula into each arm, the first for fluid, electrolyte, and insulin replacement and the second for regular sampling.

Arterial cannulation is appropriate for patients who require mechanical ventilation or for those who need intensive care for conditions such as coma, shock, or severe acidosis.

Insert a nasogastric tube and aspirate the gastric contents for all patients with impaired consciousness and for children with repeated vomiting.

Consider urinary catheterization for children with impaired consciousness. This allows accurate calculation of urinary losses, particularly in the early hyperosmolar phases of diabetic ketoacidosis in which osmotic diuresis can lead to massive urinary losses, even in the presence of dehydration.

Manage cerebral edema with intubation and mechanical ventilation in addition to osmotic diuresis.



Electrocardiography (ECG) is a useful adjunct to monitor potassium status. Characteristic changes appear with extremes of potassium status. Characteristic changes of hypokalemia as represented on ECG (see the image below) are as follows:

  • Apparent prolongation of QT interval

  • ST segment depression

  • Flat or diphasic T waves

  • Prominent U waves

  • Prolongation of PR interval

  • Sinoatrial block

    Pediatric Diabetic Ketoacidosis. A graphical repre Pediatric Diabetic Ketoacidosis. A graphical representation of the electrocardiographic changes of hypokalemia.

Hyperkalemia may develop due to overcorrection of potassium loss, with electrocardiographic changes occurring as follows (see the image below):

  • Broadening of the QRS

  • Peaked T waves

  • Prolonged PR interval

  • Disappearance of P wave

  • Diphasic QRS complex

  • Asystole

    Pediatric Diabetic Ketoacidosis. A graphical repre Pediatric Diabetic Ketoacidosis. A graphical representation of the electrocardiographic changes of hyperkalemia (due to overcorrection of potassium loss).