Alcoholic Ketoacidosis Workup

Diagnosis of alcoholic ketoacidosis (AKA) requires arterial blood gas (ABG) measurement and serum chemistry assays. Usual laboratory findings include the following ^[10] :

• The arterial pH is less than 7.3, and the serum bicarbonate level is less than 15 mEq/L

• The calculated anion gap is greater than 14 mmol/L

• The partial pressure of carbon dioxide is decreased, secondary to compensatory hyperventilation

• In some patients, severe vomiting results in chloride depletion and metabolic alkalosis, with consequently higher pH values than those found in patients with diabetic ketoacidosis ^{[4, 11]}

Arterial blood gas (ABG) measurement may show a low pCO₂ level, low bicarbonate level, and normal partial pressure of oxygen (pO₂) level. The pattern is consistent with a metabolic acidosis with a respiratory compensation.

Serum pH levels may be misleading because the patient with AKA often has a mixed acid-base disorder. In addition to metabolic acidosis due to ketone formation, a metabolic alkalosis may be present due to vomiting and volume depletion. ^[7] A respiratory alkalosis may be present secondary to hyperventilation. The possibility of a double or triple acid-base disorder means serum pH levels may be near normal despite a severe acid-base disturbance.

A compensatory respiratory alkalosis alone cannot correct the pH to normal, because the drive for compensation decreases as the pH approaches normality. This implies that a significant noncompensatory metabolic alkalosis also must be present if the pH is near the normal range.

Venous blood gas measurements correlate very well with arterial measurements. One should consider using venous blood gas measurements in lieu of arterial blood gas measurements. ^[23]

All patients with AKA have ketonuria and most have ketonemia. In AKA, the average ratio of hydroxybutyric acid (β-OH) to acetoacetic acid (5:1) tends to be higher than that which occurs in diabetic ketoacidosis (3:1). ^{[5, 12, 11]} The nitroprusside reaction (Acetest) may be negative or only weakly positive for serum ketones in AKA because nitroprusside reacts with acetone and acetoacetic acid, but not with β-OH. ^{[5, 24]} Direct serum measurements of β-OH should be used when available.

With initial therapy, ketone formation shifts toward the production of acetoacetic acid. Measured ketone levels rise, although β-OH levels decrease.

A study by Ahlström et al found beta-hydroxybutyrate analysis to be an important element in determining death by AKA. The investigators determined that after Sweden introduced such analysis on a more widespread basis in the practice of forensic autopsy, the number of deaths in the autopsy population that were attributed to AKA rose from 3-10 cases annually between 2013 and 2015 to 66 cases in 2019. ^[25]

The hallmark of AKA is ketoacidosis without marked hyperglycemia; the serum glucose level may be low, normal, or slightly elevated. ^[4] This finding can help distinguish AKA from diabetic ketoacidosis (DKA). Serum glucose levels above 300 mg/dL usually indicate DKA, unless AKA has developed in a diabetic patient. ^[11]

The anion gap is elevated. Lactate levels may be elevated. An elevated lactate level (usually does not exceed 3 mmol/L) may result from dehydration or seizure or could be the direct metabolic effect of alcohol.

Hyponatremia and hypokalemia are common laboratory findings in patients with AKA. Vomiting and extracellular volume depletion may cause hyponatremia. Hypokalemia is often associated with hypomagnesemia.

Hypomagnesemia may be caused by poor nutrition, decreased renal absorption of magnesium, or nasogastric suctioning. Serum magnesium levels are not reliable indicators of total body magnesium stores, however. Due to the linked excretion between potassium and magnesium, the presence of hypokalemia is a strong indicator of hypomagnesemia and can be used as a surrogate test to determine if magnesium replacement is needed.

True hypocalcemia associated with hypomagnesemia may be present. Concomitant pancreatitis also may contribute to true hypocalcemia. Factitious hypocalcemia can result from a markedly decreased serum albumin level following prolonged malnutrition with alcoholism.

Phosphate levels may be low, normal, or elevated. Ethanol-enhanced urinary excretion, emesis, and antacid use may contribute to hypophosphatemia in people who have chronic alcoholism.

Hyperuricemia is commonly observed; it results from decreased renal perfusion, tissue catabolism, competitive inhibition of renal uric acid excretion by ketone bodies, and direct ethanol enhancement of adenine nucleotide degradation. In a study of alcoholic Japanese men aged 40 years or older, Yokoyama et al suggested that the development of hyperuricemia was associated with quicker ethanol and acetaldehyde metabolism by those who had both the ADH1B*2 allele and the ALDH2*1/*1 genotype and with greater levels of ketosis. ^[26]

Anemia may be present secondary to nutritional deficiencies, alcoholic bone marrow suppression, or GI bleeding. The hematocrit (Hct) may be falsely elevated from hemoconcentration in the presence of intravascular volume depletion. Other findings are a decreased white blood cell (WBC) macrocytosis (mean corpuscular volume [MCV] 100-110 fL). Thrombocytopenia may be present due to chronic liver disease.

Liver and pancreatic function test results, including hepatic enzymes (eg, serum glutamic-oxaloacetic transaminase [SGOT], lactate dehydrogenase [LDH], alkaline phosphatase), total bilirubin, and pancreatic amylase and lipase levels, may be elevated because of associated illnesses (eg, alcohol-induced hepatitis, pancreatitis).

Alcohol level may be low or zero due to anorexia and decreased drinking in the preceding 1-3 days. Blood alcohol levels do not typically change the management of AKA and are therefore not often necessary.

Free fatty acid levels are usually markedly elevated, which is secondary to increased lipolysis. Insulin levels are low, glucagon levels are high. Cortisol and catecholamine levels are markedly elevated, and modest elevations of growth hormone are common.

A study by Kim et al indicated that in males with AKA, a high delta neutrophil index (DNI) value predicts poor outcomes. The cumulative survival rates for patients with an initial DNI level of 4.5% or greater were lower than for those whose DNI value was less than 4.5%. ^[27]

Because of the high risk of aspiration pneumonia in people with alcoholism, consider obtaining a chest radiograph. Esophageal rupture may occur with prolonged retching, resulting in pneumomediastinum or in subdiaphragmatic air.

Consider obtaining an urgent abdominal series in patients with significant vomiting and abdominal pain. These symptoms may indicate obstruction, perforation of a viscus, and/or pancreatitis.