Pediatric DKA Diagnostic Criteria
In pediatric patients, diabetic ketoacidosis (DKA) is diagnosed when all three of the following criteria are met: blood glucose >250 mg/dL, venous pH <7.3, and bicarbonate <15 mEq/L, with moderate ketonuria or ketonemia. 1, 2, 3
Core Diagnostic Triad
The diagnosis requires the simultaneous presence of:
- Hyperglycemia: Blood glucose >250 mg/dL (13.9 mmol/L)
- Acidemia: Venous pH <7.3
- Ketosis: Bicarbonate <15 mEq/L with moderate ketonuria or ketonemia
Venous pH is sufficient for diagnosis and monitoring—arterial blood gases are generally unnecessary, as venous pH is typically only 0.03 units lower than arterial pH. 2, 3
Severity Classification
Once DKA is confirmed, classify severity based on the degree of acidosis 4:
- Mild DKA: Venous pH ≥7.2 to <7.3, bicarbonate ≥10 to <15 mmol/L
- Moderate DKA: Venous pH ≥7.1 to <7.2, bicarbonate ≥5 to <10 mmol/L
- Severe DKA: Venous pH <7.1, bicarbonate <5 mmol/L
Severe DKA carries higher morbidity and mortality risk and requires more intensive monitoring, potentially including central venous and intra-arterial pressure monitoring. 4
Essential Initial Laboratory Workup
After history and physical examination, obtain STAT 1, 3:
- Blood glucose
- Venous blood gases (not arterial)
- Complete electrolyte panel
- Blood urea nitrogen (BUN) and creatinine
- Calcium and phosphorous
- Urinalysis for ketones
- Complete blood count with differential
- Electrocardiogram
Calculate the anion gap: (Na+) - (Cl- + HCO3-). An anion gap >10-12 mEq/L supports the diagnosis. 3
Critical Diagnostic Pitfalls
Ketone Measurement Method Matters
β-hydroxybutyrate (β-OHB) measurement in blood is the preferred method for monitoring DKA. 1 The nitroprusside method (used in most urine and some serum ketone tests) only measures acetoacetic acid and acetone, missing β-OHB—the strongest and most prevalent acid in DKA. During treatment, β-OHB converts to acetoacetic acid, which can falsely suggest worsening ketosis when the patient is actually improving. Therefore, nitroprusside-based ketone measurements should not guide treatment decisions. 1, 3
Distinguishing DKA from Other Conditions
DKA must be differentiated from:
- Hyperosmolar hyperglycemic state (HHS): Blood glucose >600 mg/dL, venous pH >7.3, bicarbonate >15 mEq/L, with altered mental status or severe dehydration 1, 2, 3
- Other high anion gap metabolic acidoses: Lactic acidosis, salicylate ingestion, methanol/ethylene glycol poisoning, chronic renal failure 3
- Acute illnesses with overlapping symptoms: Pneumonia, asthma exacerbation, urinary tract infection, gastroenteritis, acute abdomen, CNS infection 5
Type 1 vs Type 2 Diabetes at Presentation
Up to 33% of children with type 2 diabetes have ketonuria at diagnosis, and 5-25% present with ketoacidosis. 6 This overlap makes immediate diabetes type classification difficult. Key discriminators include:
- C-peptide levels (fasting >0.47 nmol/L, 1-hour >0.98 nmol/L, 2-hour >2.03 nmol/L favor type 2)
- BMI (>20.95 kg/m² favors type 2)
- Systolic blood pressure (>110.5 mmHg favors type 2)
- Presence of fatty liver (favors type 2)
- Islet autoantibodies (positive favors type 1) 7
However, initial treatment should address the metabolic emergency regardless of diabetes type, with classification refined after metabolic stabilization. 8
Resolution Criteria
DKA is considered resolved when glucose <200 mg/dL, serum bicarbonate ≥18 mEq/L, and venous pH >7.3. 1, 2, 3 The anion gap should normalize to ≤12 mEq/L. 2
Special Considerations
Corrected Sodium
Calculate corrected sodium for hyperglycemia: For each 100 mg/dL glucose above 100 mg/dL, add 1.6 mEq/L to the measured sodium value. 1, 3 This helps assess true sodium status and guides fluid management.
Temperature
Patients may be normothermic or hypothermic despite infection due to peripheral vasodilation. Hypothermia is a poor prognostic sign. 3
Mental Status
Mental status varies from full alertness to coma. Altered consciousness is more common in severe DKA and should raise concern for cerebral edema, especially if it worsens during treatment. 4, 3