How is diabetic ketoacidosis diagnosed?

Diagnosis of Diabetic Ketoacidosis

Diabetic ketoacidosis (DKA) is diagnosed by the presence of hyperglycemia (blood glucose >250 mg/dL), metabolic acidosis (venous pH <7.3, serum bicarbonate <18 mEq/L), and elevated blood β-hydroxybutyrate (β-OHB), which is the preferred ketone measurement method. 1

Diagnostic Criteria

The diagnosis requires three key components measured simultaneously:

1. Hyperglycemia

• Blood glucose >250 mg/dL is the traditional threshold 1, 2, 3
• However, euglycemic DKA can occur, particularly in patients using SGLT2 inhibitors, where glucose may be only mildly elevated or even normal 1, 3

2. Metabolic Acidosis

• Venous pH <7.3 (venous pH is typically 0.03 units lower than arterial pH and is adequate for monitoring) 1
• Serum bicarbonate <18 mEq/L 1, 2
• Elevated anion gap (calculated as [Na⁺] - [Cl⁻ + HCO₃⁻]) 4
• Arterial blood gases are generally unnecessary; venous pH and anion gap suffice for diagnosis and monitoring 1

3. Ketone Body Elevation

Blood β-hydroxybutyrate (β-OHB) measurement is the gold standard for diagnosis and should be used preferentially over other ketone testing methods. 1

• β-OHB >3.0 mmol/L has the highest diagnostic performance (sensitivity 99.87%, specificity 92.89%) 5
• Research suggests optimal cut-off values of 6.3 mmol/L for β-OHB in confirmed DKA cases 6

Critical Testing Methodology

Preferred: Blood β-Hydroxybutyrate

• Specific β-OHB measurement in blood is mandatory for accurate DKA diagnosis 1
• β-OHB is the predominant ketone body in DKA and directly reflects the severity of ketoacidosis 1
• Point-of-care capillary β-OHB meters (such as Precision-Xtra) correlate highly with serum values (r=0.99) 5

Avoid: Nitroprusside-Based Testing

• Nitroprusside reagent tests (urine dipsticks, serum ketone strips) should NOT be used for DKA diagnosis or monitoring 1
• These methods only detect acetoacetate and acetone, missing β-OHB entirely 1
• During treatment, acetoacetate may increase as β-OHB decreases, falsely suggesting worsening ketosis 1
• Urine ketone testing is particularly unreliable and should not be used for clinical decision-making in DKA 1

Initial Laboratory Panel

Obtain the following tests immediately upon suspicion of DKA:

• Blood glucose (capillary or serum) 1, 7
• Venous blood gas for pH and bicarbonate 1, 7
• Serum electrolytes (sodium, potassium, chloride) to calculate anion gap 1, 7
• Blood β-hydroxybutyrate (specific measurement) 1
• Blood urea nitrogen and creatinine to assess renal function and hydration 1, 7
• Serum osmolality to differentiate from hyperosmolar hyperglycemic state 1
• Complete blood count, urinalysis, and electrocardiography 2, 3

Differential Diagnosis Considerations

Distinguishing DKA from Other Ketotic States

Alcoholic Ketoacidosis:

• Plasma glucose typically <250 mg/dL, often hypoglycemic 4
• History of recent alcohol cessation with poor oral intake 4
• High anion gap metabolic acidosis with elevated β-OHB 4

Starvation Ketosis:

• Serum bicarbonate usually not lower than 18 mEq/L (key differentiator) 4, 7
• Mildly elevated glucose or hypoglycemia 7
• β-OHB elevated but typically lower than in DKA 7
• Less pronounced anion gap elevation 7

Hyperosmolar Hyperglycemic State (HHS):

• Blood glucose typically >600 mg/dL 1
• Venous pH >7.3, bicarbonate >15 mEq/L 1
• Minimal or absent ketosis 1
• More profound alteration in mental status 1

Monitoring During Treatment

• Recheck blood glucose, electrolytes, venous pH, and β-OHB every 2-4 hours until clinical improvement 1, 7
• Serial arterial blood gases are unnecessary; venous pH adequately tracks acidosis resolution 1
• DKA resolution criteria: glucose <200 mg/dL, serum bicarbonate ≥18 mEq/L, venous pH >7.3 1, 7

Common Pitfalls to Avoid

• Do not rely on urine ketone testing for diagnosis or treatment monitoring—it misses β-OHB and can be misleading during therapy 1, 7
• Do not assume normal glucose excludes DKA—euglycemic DKA occurs, especially with SGLT2 inhibitor use 1, 3
• Do not use nitroprusside-based blood ketone tests for monitoring treatment—they show paradoxical increases as β-OHB falls 1
• Be aware that up to 30% of first morning urine specimens in pregnant women show positive ketones without DKA 1
• False-positive urine ketone results can occur with sulfhydryl drugs like captopril 1