Diabetic Ketoacidosis Diagnosis and Management

Diagnostic Criteria

• The American Diabetes Association recommends checking blood glucose immediately to distinguish between diabetic ketoacidosis (DKA) and benign starvation ketosis, with glucose >250 mg/dL indicating a medical emergency requiring immediate DKA evaluation 1, 2
• DKA diagnostic criteria require all of the following: plasma glucose >250 mg/dL, arterial pH <7.30, serum bicarbonate <15 mEq/L, positive urine or serum ketones, and anion gap >10 mEq/L 1, 3, 4

Differential Diagnoses

• Diabetic ketoacidosis (DKA) presents with polyuria, polydipsia, weakness, abdominal pain, Kussmaul respirations, and altered mental status, and is precipitated by infection, insulin omission, or new-onset diabetes 1, 4
• Starvation ketosis occurs from reduced caloric intake, causing a shift to fat metabolism, and is characterized by normal or low blood glucose, serum bicarbonate usually not lower than 18 mEq/L, and blood ketones ranging 0.3-4 mmol/L 5, 4
• Hyperosmolar hyperglycemic state (HHS) is characterized by glucose typically >600 mg/dL, effective serum osmolality >320 mOsm/kg, and only small amounts of ketones present, with profound dehydration and altered mental status 1

Critical Testing Recommendations

• Blood beta-hydroxybutyrate measurement is strongly preferred over urine testing for all clinical decision-making, as urine dipsticks only detect acetoacetate and significantly underestimate total ketone concentration 2, 3, 5
• The American College of Clinical Endocrinologists recommends immediate evaluation for pathological ketosis in patients on SGLT2 inhibitors, with blood ketone action thresholds of <0.5 mmol/L (no intervention), 0.5-1.5 mmol/L (initiate sick-day rules), and ≥1.5 mmol/L (immediate medical attention) 2, 3, 5

Risk Stratification

• High-risk patients requiring immediate evaluation include those with known type 1 diabetes, history of prior DKA, currently on SGLT2 inhibitors, and presenting with unexplained hyperglycemia or symptoms of ketosis 1, 2, 5
• Lower-risk patients likely have starvation ketosis, characterized by no diabetes history, normal blood glucose, recent decreased oral intake, and no abdominal pain or altered mental status 5, 4

Ketone Detection and Diagnostic Considerations

Diagnostic Criteria and Testing

• Up to 30% of first morning urine specimens show positive ketones even in healthy individuals, and this increases during fasting states, according to the American Diabetes Association 6
• DKA diagnostic criteria require: glucose >250 mg/dL, arterial pH <7.30, serum bicarbonate <15 mEq/L, and anion gap >10 mEq/L, as recommended by the American Diabetes Association 7
• Blood beta-hydroxybutyrate measurement is strongly preferred over urine testing for all clinical decision-making, as stated by the American Diabetes Association 7
• The nitroprusside method (standard urine dipstick) does not measure beta-hydroxybutyrate, the predominant ketone body in DKA, according to the American Diabetes Association 7, 8, and 6

Medication-Related Considerations

• SGLT2 inhibitors decrease the hyperglycemia that typically accompanies DKA, making diagnosis more challenging, as noted by the American Diabetes Association 7 and 6
• SGLT2 inhibitors increase DKA risk and can cause euglycemic DKA where glucose may be <250 mg/dL despite severe ketoacidosis, as reported by the American Diabetes Association 6

Differential Diagnoses

• Individuals with alcoholic ketoacidosis will have positive urine ketone readings, but hyperglycemia is not usually present, according to the American Diabetes Association 6
• Positive urine ketone readings are found in up to 30% of first morning urine specimens from pregnant women (with or without diabetes), as stated by the American Diabetes Association 6
• Ketones can appear after hypoglycemic episodes, as noted by the American Diabetes Association 6

Diabetic Ketoacidosis Diagnosis and Management

Primary Diagnostic Consideration

• The simultaneous presence of ketones and hematuria most commonly indicates diabetic ketoacidosis (DKA) with concurrent urinary tract infection, as infection precipitates approximately 50% of DKA cases and urinary tract infections are among the most frequent bacterial triggers, according to the American Diabetes Association 9
• Infection is the most common precipitating factor for DKA, with urinary tract infections being particularly frequent, and the presence of hematuria alongside ketones should trigger bacterial cultures of urine, blood, and other sites as indicated, as recommended by the American Diabetes Association 9
• Patients with DKA and infection can be normothermic or even hypothermic due to peripheral vasodilation, and hypothermia, if present, is a poor prognostic sign, as stated by the American Diabetes Association 9

Alternative Causes of Combined Ketones and Hematuria

• Alcohol-related trauma can cause hematuria in patients with alcoholic ketoacidosis, which presents with positive urine ketones but hyperglycemia is not usually present, according to the American Diabetes Association 9
• Alcohol-induced coagulopathy can cause bleeding and hematuria in patients with alcoholic ketoacidosis, as reported by the American Diabetes Association 9

Critical Diagnostic Criteria

• The diagnostic approach for DKA requires complete DKA diagnostic criteria, including plasma glucose >250 mg/dL, arterial pH <7.30, serum bicarbonate <15 mEq/L, positive ketones, and anion gap >10 mEq/L, as defined by the American Diabetes Association 9

Guideline Recommendations for Ketone Testing and Interpretation

Preferred Diagnostic Modality

• The American Diabetes Association recommends using blood β‑hydroxybutyrate measurement rather than urine dipstick testing for all clinical decisions because urine dipsticks detect only acetoacetate and substantially underestimate total ketone levels, missing the predominant ketone body in ketoacidosis. 10

Reliability of Urine Ketone Tests

• The American Diabetes Association states that urine ketone tests are unreliable for diagnosing or monitoring treatment of ketoacidosis. 10

Indications for Ketone Testing in Type 1 Diabetes

• The American Diabetes Association advises that individuals with type 1 diabetes should perform ketone testing when any of the following occur:

Factors Leading to False‑Positive Results

• The American Diabetes Association notes that sulfhydryl‑containing medications such as captopril can cause false‑positive urine ketone results. 10

Factors Leading to False‑Negative Results

• The American Diabetes Association warns that urine ketone test strips left exposed to air for prolonged periods may yield false‑negative results. 10
• Additionally, highly acidic urine—such as that occurring after large intake of ascorbic acid—can produce false‑negative ketone readings. 10

Physiologic (Benign) Ketonuria

• The American Diabetes Association reports that physiologic ketonuria is common and generally harmless; for example, up to 30 % of first‑morning urine samples from pregnant individuals may test positive for ketones. 10

Ketonuria: Etiology, Pathophysiology, and Diagnostic Strategies

Physiological (Benign) Causes

• Carbohydrate‑deprivation states such as starvation, prolonged fasting, or very low‑carbohydrate diets (<50 g/day) trigger hepatic fat oxidation, producing β‑hydroxybutyrate concentrations of 0.3–4 mmol/L while maintaining normal blood glucose and serum bicarbonate ≥ 18 mEq/L. 11
• Up to 30 % of first‑morning urine samples from asymptomatic healthy adults are positive for ketones, especially after an overnight fast. 12
• Approximately 30 % of first‑morning urine samples from pregnant individuals (with or without diabetes) contain ketones, reflecting a normal metabolic adaptation to increased energy demands. 12

Pathological Causes

• Diabetic ketoacidosis (DKA) is defined by the simultaneous presence of: blood glucose > 250 mg/dL, arterial pH < 7.30, serum bicarbonate < 15 mEq/L, detectable urine or blood ketones, and an anion gap > 10 mEq/L. 12
• SGLT2‑inhibitor therapy raises the risk of euglycemic DKA, in which severe ketoacidosis occurs with blood glucose < 250 mg/dL, complicating timely diagnosis. 12
• Alcoholic ketoacidosis results from ethanol metabolism depleting hepatic glycogen and NAD⁺, leading to lipolysis and ketogenesis with positive urine ketones but typically without hyperglycemia. 12
• After hypoglycemic episodes, counter‑regulatory hormone surges can stimulate lipolysis, producing transient ketonuria during recovery. 12

Diagnostic Considerations

• Standard urine dipsticks based on nitroprusside detect only acetoacetate and miss β‑hydroxybutyrate, the predominant ketone in DKA, thereby under‑estimating total ketone burden. 12
• Quantitative blood β‑hydroxybutyrate measurement is strongly preferred over urine dipstick testing for all clinical decision‑making regarding ketosis and DKA. 12
• During DKA treatment, β‑hydroxybutyrate levels fall while acetoacetate may rise, making urine dipstick results unreliable for monitoring therapeutic response. 12
• False‑negative urine ketone results can occur when test strips are exposed to air for prolonged periods or when urine is highly acidic (e.g., after large ascorbic‑acid intake). 12

Guidelines for Evaluation and Management of Trace Ketones in Urine

Diagnostic Criteria for Diabetic Ketoacidosis (DKA)

• The diagnostic work‑up for suspected DKA should include arterial pH, serum bicarbonate, anion gap, and preferably a blood beta‑hydroxybutyrate measurement. 13
• DKA is confirmed when all of the following are present: glucose > 250 mg/dL, pH < 7.30, bicarbonate < 15 mEq/L, positive ketones, and anion gap > 10 mEq/L. 13

Frequency of Physiologic Ketonuria

• Up to 30 % of first‑morning urine specimens from healthy individuals are ketone‑positive, most commonly after an overnight fast. 13

Populations Requiring Heightened Evaluation

• Individuals with type 1 diabetes or a prior episode of DKA should be evaluated more closely when trace ketones are detected. 13
• Febrile or acutely ill diabetic patients are at high risk because infection precipitates approximately 50 % of DKA cases. 13

Limitations of Urine Dipstick Testing

• Standard urine dipsticks detect only acetoacetate and miss beta‑hydroxybutyrate, the predominant ketone body in pathological ketosis; their sensitivity for mild‑to‑moderate ketosis can be as low as 35‑52 %. 13
• Because of this limitation, blood beta‑hydroxybutyrate measurement is strongly preferred over urine testing for all clinical decision‑making. 13

Blood Ketone Thresholds and Management Actions

• 0.5–1.5 mmol/L blood beta‑hydroxybutyrate: initiate “sick‑day” measures (oral hydration, supplemental short‑acting insulin with carbohydrate intake, frequent glucose/ketone monitoring). 13
• ≥1.5 mmol/L blood beta‑hydroxybutyrate: seek immediate medical attention; intravenous insulin therapy is typically required. 13

Management When Only Urine Testing Is Available

• Patients prone to ketosis who present with unexplained hyperglycemia or DKA‑compatible symptoms (e.g., abdominal pain, nausea, vomiting) should follow sick‑day rules and obtain medical advice if urine ketone levels rise. 13
• Urine ketone results should not be used to monitor DKA treatment response, because beta‑hydroxybutyrate declines first while acetoacetate may paradoxically increase, leading to misleading urine readings. 13

Clinical Guidance on Trace Urinary Ketones in Adults

Prevalence and Physiologic Context

• Up to one‑third of first‑morning urine samples from asymptomatic healthy adults contain trace ketones, especially after overnight fasting. 14
• Approximately one‑third of pregnant individuals (with or without diabetes) show positive first‑morning urine ketones, reflecting normal metabolic adaptation during pregnancy. 14
• Low‑carbohydrate diets (< 50 g carbohydrate per day) or prolonged fasting can produce modest blood β‑hydroxybutyrate levels (≈ 0.3–4 mmol/L) while maintaining normal glucose and serum bicarbonate ≥ 18 mEq/L, resulting in benign ketonuria. 14

High‑Risk Populations Requiring Immediate Evaluation

• Adults with known type 1 diabetes or a prior episode of DKA should test for ketones during acute illness, stress, persistent hyperglycemia (> 300 mg/dL), pregnancy, or any symptoms suggestive of ketoacidosis. 14
• Patients receiving SGLT2‑inhibitor therapy are at increased risk for euglycemic DKA, where severe ketoacidosis may occur despite glucose levels below the usual diagnostic threshold. 14
• Febrile or acutely ill diabetic patients—particularly those with infections, which precipitate roughly half of DKA cases—require prompt ketone assessment. 14
• Any individual with unexplained hyperglycemia or symptoms compatible with DKA (abdominal pain, nausea, vomiting, rapid breathing, altered mental status) should undergo urgent evaluation. 14

Limitations of Urine Dipstick Testing

• Standard urine dipsticks detect only acetoacetate and miss β‑hydroxybutyrate, the predominant ketone body in DKA, leading to substantial underestimation of total ketone burden. 14
• Blood β‑hydroxybutyrate measurement is strongly preferred over urine testing for all clinical decisions regarding ketosis and DKA. 14
• Urine ketone results are unreliable for diagnosing or monitoring DKA treatment because β‑hydroxybutyrate declines early during therapy while acetoacetate may paradoxically rise. 14
• The sensitivity of urine dipsticks for mild‑to‑moderate ketosis can be as low as 35–52 %. 14

Blood Ketone Thresholds and Recommended Actions

• β‑hydroxybutyrate < 0.5 mmol/L – No specific intervention required.
• β‑hydroxybutyrate 0.5–1.5 mmol/L – Initiate “sick‑day” management: oral hydration, short‑acting insulin with carbohydrate intake, and frequent glucose/ketone monitoring.
• β‑hydroxybutyrate ≥ 1.5 mmol/L – Seek immediate medical care; intravenous insulin therapy is typically indicated.

Diagnostic Criteria for Diabetic Ketoacidosis (DKA)

• Plasma glucose > 250 mg/dL
• Arterial pH < 7.30
• Serum bicarbonate < 15 mEq/L
• Positive urine or serum ketones
• Anion gap > 10 mEq/L
  [Each criterion is supported by the same reference.] [14]

Common Sources of False‑Positive and False‑Negative Urine Results

• False‑positive: Sulfhydryl‑containing medications such as captopril can produce spurious ketone readings. 14
• False‑negative: Prolonged exposure of test strips to air or highly acidic urine (e.g., after large ascorbic acid intake) can diminish detection accuracy. 14

Practical Use of Urine Ketone Testing When Blood Assays Are Unavailable

• Ketone testing supplies should be readily accessible in outpatient and clinic settings for immediate assessment of at‑risk patients. 14
• Urine ketone results should not be used to monitor response to DKA treatment because of their unreliability in reflecting true ketone clearance. 14
• In patients prone to ketosis who present with unexplained hyperglycemia or DKA‑compatible symptoms, follow sick‑day rules and seek medical advice if urine ketone levels rise. 14

Evaluation and Management of Large Ketonuria in Diabetes

Triage and Diagnostic Criteria

• Large ketonuria combined with plasma glucose > 250 mg/dL defines a medical emergency that warrants immediate work‑up for diabetic ketoacidosis (DKA). [15][16]
• DKA is confirmed only when all of the following are present: plasma glucose > 250 mg/dL, arterial pH < 7.30, serum bicarbonate < 15 mEq/L, positive ketones (urine or blood), and anion gap > 10 mEq/L. 15
• If any of the above criteria are absent, large ketonuria alone does not constitute DKA. 15

Limitations of Urine Ketone Testing

• Standard urine dipsticks detect only acetoacetate and miss β‑hydroxybutyrate, the predominant ketone body in DKA, leading to under‑estimation of total ketone burden. 15
• Consequently, “large” ketonuria on urine may actually under‑represent the severity of ketoacidosis. 15
• Blood β‑hydroxybutyrate measurement is strongly preferred for all clinical decision‑making, including diagnosis and monitoring. 15
• Urine ketone results are unreliable for tracking treatment response because β‑hydroxybutyrate falls while acetoacetate may rise, causing persistently positive dipsticks despite improvement. 15
• Up to 30 % of healthy individuals exhibit trace‑to‑moderate ketonuria after fasting; this physiologic finding is benign and does not require intervention. 15

High‑Risk Populations Requiring Immediate Evaluation

• Patients with type 1 diabetes are at highest risk for rapid DKA progression. 16
• Individuals with a prior history of DKA should be assessed urgently when large ketonuria is present. 16
• Use of SGLT2 inhibitors can precipitate euglycemic DKA (glucose < 250 mg/dL); normal glucose does not exclude DKA in these patients. [15][16]
• Pregnant patients with pre‑gestational diabetes have a ≈ 2 % risk of developing DKA during pregnancy, which may present euglycemically; immediate evaluation is mandatory. 16
• Acute illness, fever, or infection precipitate roughly 50 % of DKA episodes; such patients need prompt assessment regardless of glucose level. 16
• Presence of classic DKA symptoms—nausea, vomiting, abdominal pain, Kussmaul respirations, or altered mental status—warrants urgent work‑up. 16

Immediate Management Steps for Suspected DKA

• Obtain arterial blood gas, serum bicarbonate, anion gap, and comprehensive electrolytes without delay. 16
• Measure blood β‑hydroxybutyrate (preferred over urine ketones) as soon as possible. 16
• Actively search for precipitating factors (e.g., infection, medication non‑adherence, new stressors). 16
• Do not discontinue basal insulin even if the patient is not eating; continuation helps prevent worsening ketosis. 16
• If DKA diagnostic criteria are met, initiate intravenous fluids and insulin therapy promptly. 16
• Pregnant patients with any suspicion of DKA should seek immediate medical attention due to significant feto‑maternal risk. 16

Guidelines for Ketone Testing in Diabetes Management

1. Indications for Ketone Testing Based on Glucose Levels and Clinical Context

• Ketone testing is recommended when random blood glucose is unexpectedly elevated or exceeds 300 mg/dL (16.7 mmol/L), especially in individuals with type 1 diabetes, a prior episode of diabetic ketoacidosis (DKA), or who are receiving SGLT2‑inhibitor therapy. 17
• Type 1 diabetes patients should check ketones whenever blood glucose persistently exceeds 300 mg/dL, during acute illness or physiological stress, or when symptoms suggestive of ketosis (e.g., abdominal pain, nausea, vomiting) appear. 17
• Patients with a history of DKA should test for ketones with any unexplained hyperglycemia or when symptoms indicative of ketoacidosis develop. 17, 18
• Individuals treated with SGLT2 inhibitors require heightened vigilance because these agents can precipitate euglycemic DKA, where severe ketoacidosis occurs even with glucose levels below 250 mg/dL. 17

2. Symptom‑Driven Ketone Testing (Independent of Glucose Level)

• Immediate ketone testing is advised when patients experience nausea or vomiting. 19
• Immediate testing is also indicated for unexplained fatigue or weakness. 18

3. Preferred Method of Ketone Measurement

• Blood β‑hydroxybutyrate measurement is strongly preferred over urine ketone testing for all clinical decision‑making. 17
• Standard urine dipsticks detect only acetoacetate and miss β‑hydroxybutyrate, the predominant ketone body in DKA, leading to substantial under‑estimation of total ketone burden. 17

4. Limitations of Urine Ketone Testing

• Up to 30 % of healthy individuals may have positive urine ketones after an overnight fast, reducing test specificity. 17
• Urine ketone results cannot reliably rule out DKA in symptomatic patients because false‑negative results occur when β‑hydroxybutyrate is elevated. 17
• Urine ketones should not be used to monitor DKA treatment, as β‑hydroxybutyrate may fall while acetoacetate rises, producing misleadingly persistent positive results. 17
• False‑positive urine ketone results can arise from sulfhydryl‑containing medications (e.g., captopril). 17
• False‑negative results may occur if test strips are exposed to air or when urine is highly acidic. 17

5. Use of Urine Ketones When Blood Testing Is Unavailable

• In settings lacking blood β‑hydroxybutyrate measurement, urine ketones may be employed as a screening tool for patients with unexplained hyperglycemia or ketosis‑related symptoms. 17, 18
• If urine ketones are positive, clinicians should implement sick‑day management rules and seek medical advice promptly. 18
• A negative urine ketone test has a high negative predictive value and can help rule out DKA, but it is poor for confirming disease severity. 17

6. Criteria for Confirming DKA in the Emergency Setting

• DKA is confirmed only when all of the following are present:

Ketone Testing Guidelines for Diabetes Management

Primary Glucose Thresholds for Ketone Assessment

• Individuals at risk for ketosis (type 1 diabetes, prior DKA, or on SGLT2 inhibitors) should perform blood ketone testing when random blood glucose exceeds 250 mg/dL (13.9 mmol/L) or when any unexplained hyperglycemia is accompanied by ketosis‑related symptoms (e.g., abdominal pain, nausea). American Diabetes Association recommendation. 17
• The 250 mg/dL (13.9 mmol/L) glucose level serves as the diagnostic cutoff for diabetic ketoacidosis (DKA) when paired with metabolic criteria (acidosis, low bicarbonate, elevated anion gap). American Diabetes Association guideline. 19

Additional Clinical Triggers Beyond the Glucose Cutoff

• In high‑risk patients, any unexplained hyperglycemia—even below 250 mg/dL—warrants immediate ketone testing, particularly for those receiving SGLT2 inhibitors who may develop euglycemic DKA. American Diabetes Association guidance. 17
• The presence of ketosis‑related symptoms (abdominal pain, nausea, vomiting, Kussmaul respiration) mandates ketone testing regardless of glucose level. American Diabetes Association recommendation. 17

High‑Risk Populations Requiring Lower Testing Thresholds

• SGLT2‑inhibitor users: Because euglycemic DKA can occur with glucose as low as 150–200 mg/dL, clinicians should maintain a low threshold for ketone testing and not rely solely on glucose values. Lancet Diabetes & Endocrinology evidence. 20
• Type 1 diabetes or prior DKA history: During acute illness, stress, or any deterioration in glycemic control, a lower glucose threshold for ketone testing should be applied. American Diabetes Association recommendation. 17

Preferred Testing Modality

• Blood β‑hydroxybutyrate measurement is strongly preferred over urine ketone dipsticks for clinical decision‑making because urine tests detect only acetoacetate and miss the predominant ketone body in DKA. American Diabetes Association guidance. 17
• When urine ketones are negative, 10–13 % of patients still have elevated blood ketones that require clinical intervention. American Diabetes Association observation. 17
• Urine ketone results should not be used to monitor DKA treatment response, as β‑hydroxybutyrate declines while acetoacetate may rise, leading to misleadingly positive urine tests despite improvement. American Diabetes Association recommendation. 17
• Trace or mild ketonuria in diabetic patients should not be dismissed as benign, because it may indicate impending DKA, unlike the physiologic ketonuria seen in up to 30 % of healthy individuals after fasting. American Diabetes Association caution. 17

DKA Diagnostic Criteria (to be applied after ketone detection)

• Confirm DKA when blood pH < 7.30, serum bicarbonate < 15 mEq/L, and anion gap > 10 mmol/L are present alongside hyperglycemia and elevated blood β‑hydroxybutyrate. American Diabetes Association criteria. [21][19]

Practical Implementation Summary

• Step 1: Measure random blood glucose in any diabetic patient presenting with illness or symptoms.
• Step 2: If glucose ≥ 250 mg/dL or any unexplained hyperglycemia with ketosis symptoms, obtain a blood β‑hydroxybutyrate level.
• Step 3: If blood ketone testing is unavailable, a urine dipstick may be used with the understanding of its limitations.
• Step 4: Apply the DKA diagnostic criteria (pH, bicarbonate, anion gap) when ketones are elevated.
• Step 5: Initiate appropriate management based on ketone level and clinical status (e.g., sick‑day rules for 0.5–1.5 mmol/L, urgent medical care for ≥1.5 mmol/L).

For patients on SGLT2 inhibitors: consider testing ketones at glucose > 150 mg/dL if symptomatic, and maintain a high index of suspicion for euglycemic DKA. Lancet Diabetes & Endocrinology evidence. 20