Anion Gap Metabolic Acidosis

What's on this Page

Anion Gap Metabolic Acidosis

Description

Anion gap metabolic acidosis is a metabolic acidosis characterized by a buffering of nonchloride acids, resulting in a decrease in serum pH and bicarbonate without an elevation in serum chloride.⁽²⁾
An elevated anion gap can also be seen in the presence of other mixed acid-base disorders, with or without a change in blood pH.⁽²⁾

Definitions

An anion gap (AG) is the difference between serum sodium cations and chloride plus bicarbonate anions.^(1,2)View full sizeAG = [Na⁺] – {[Cl^–] + [HCO₃^–]}
Serum albumin typically accounts for the normal AG of 8-12 mEq/L.⁽¹⁾
The AG value is dependent on the type of measurement instrument used. Physicians should use their own laboratory’s reference range to determine what is normal.^(2,3)
An elevated anion gap suggests a possible accumulation of organic or inorganic acids due to endogenous or exogenous sources.^(1,3)
Acidemia is defined as having an arterial pH of < 7.38.⁽²⁾
Acidosis is the process that causes acidemia.⁽²⁾
Metabolic acidosis is defined as having a decrease in serum bicarbonate (HCO₃^–) to < 22 mmol/L as the cause of acidemia.⁽²⁾

Incidence and Prevalence

In a cohort of 411 patients with homozygous sickle cell disease, 42% had metabolic acidosis. Prevalence is higher among female patients than male patients (Clin J Am Soc Nephrol 2014 Apr;9(4):648 full-text).
64% of 851 patients admitted to intensive care unit (ICU) with suspected acid-base disorder reported to have acute metabolic acidosis
- based on retrospective cohort study
- 851 patients admitted to ICU with suspected acid-base disorder and all necessary variables required to diagnose an acid-base disorder were evaluated
- 548 (64%) had metabolic acidosis (standard base excess [SBE] < -2 mEq/L)
  - 19.2% with hyperchloremic metabolic acidosis
  - 37.2% with strong ion gap (SIG) metabolic acidosis
  - 43.6% with lactic acidosis
- Reference – Crit Care 2006 Feb;10(1):R22 full-text, commentary can be found in Crit Care 2006;10(3):413

Pathogenesis

An electrical neutrality in a solution requires the sum of cations and anions to be equal.⁽³⁾
Sodium cations, chloride anions, bicarbonate anions, and albumin anions constitute the quantitative majority of ions in an extracellular fluid compartment.⁽³⁾
Anion gap metabolic acidosis develops when the concentration of bicarbonate decreases relative to the levels of sodium and chloride due to the following processes:^(1,2)
- An increase in nonchloride acid production, such as in lactic acidosis, ketoacidosis, or ingestion of certain medications or toxic substances
- A decrease in acid excretion, such as in renal failure
- Cell lysis, such as during massive rhabdomyolysis
An elevated anion gap (see definition here) means that there is a significant presence of unmeasured anions (the measured anions are chloride and bicarbonate) due to the presence of an increased amount of organic or inorganic acids. The following are some of the unmeasured anions in anion gap metabolic acidosis:⁽¹⁾
- Lactate
- Ketones, such as beta-hydroxybutyrate and acetoacetate, which are the cause of ketoacidosis
- Formate, which is converted from formaldehyde, which is converted from methanol, in the body
- Hippurate or benzoate, from toluene ingestion
- Acetate and chloroacetate, from paraldehyde ingestion
- Glyoxylate and oxalate, which are converted from ethylene glycol in the body

Differential Diagnosis

Causes of Elevated Anion Gap

Overview

An increased anion gap suggests a possible accumulation of organic or inorganic acids due to endogenous or exogenous sources.^(1,3)
Commonly used mnemonics reflecting the causes of anion gap metabolic acidosis are as follows:
- GOLD MARK:⁽²⁾
  - G: glycols (ethylene and propylene)
  - O: 5-oxoproline (pyroglutamic acid)
  - L: L-lactate
  - D: D-lactate
  - M: methanol
  - A: aspirin
  - R: renal failure, or rhabdomyolysis
  - K: ketoacidosis
- CAT MUD PILES:⁽¹⁾
  - C: carbon monoxide, or cyanide
  - A: alcoholic ketoacidosis (starvation ketoacidosis)
  - T: toluene, which causes both normal anion gap and elevated anion gap acidosis
  - M: methanol, or metformin
  - U: uremia
  - D: diabetic ketoacidosis
  - P: propylene glycol, paraldehyde, or phenformin
  - I: isoniazid, or iron
  - L: lactic acidosis
  - E: ethylene glycol, or ethanol
  - S: salicylates

Ketoacidosis

Uncontrolled diabetes sup class=”ph sup”>(1,2)
Alcoholic ketoacidosis^(1,2)
Starvation ketoacidosis^(1,2)
Life-threatening ketoacidosis, which has been reported following a strict adherence to Atkins diet (maintaining ketonuria for 1 month) in a case report (Lancet 2006 Mar 18;367(9514):958)
Inactivating mutations in monocarboxylate transporter 1, which has been reported to be associated with an increased ketoacidosis severity in a genetic analysis of 96 patients (N Engl J Med 2014 Nov 13;371(20):1900)

Lactic Acidosis

For type A L-lactic acidosis, a hypoxic lactic acidosis, possible causes are as follows:^(1,2)
- Septic shock
- Hypovolemic shock
- Acute mesenteric ischemia
- Cyanide poisoning
- Carbon monoxide poisoning
- Hypoxemia
For type B L-lactic acidosis, nonhypoxic lactic acidosis, possible causes are as follows:^(1,2)
- Thiamine deficiency
- Seizure
- Poor lactate clearance due to liver failure
- Glycogen storage diseases
- Malignancy such as Hodgkin Lymphoma (HL)
- Medication use, and toxic ingestions, including:
  - Propofol
  - Metformin
  - Niacin
  - Nucleoside reverse transcriptase inhibitors
  - Isoniazid toxicity
  - Iron toxicity
D-lactic acidosis, which is a rare acidosis caused by the generation of nonmetabolizable D-lactate by intestinal bacteria in patients with the following conditions:
- Short bowel syndrome^(1,2)
- Prior small bowel resection⁽¹⁾
- Blind loop syndrome sup class=”ph sup”>(1)
Other causes of lactic acidosis include the following:^(1,2)
- Aspirin intoxication
- Ethanol intoxication (different than alcoholic ketoacidosis)
- Methanol poisoning
- Ethylene glycol poisoning
- Propylene glycol intake
- Early toluene intake
- Paraldehyde poisoning

Other Causes

Underexcretion of organic acids, such as in advanced renal failure (uremia) sup class=”ph sup”>(1,2)
Cell lysis, such as in massive rhabdomyolysis ⁽²⁾
Penicillin-derived antibiotics, such as when ammonium is excreted in the urine with an anion other than chloride⁽²⁾
Pyroglutamic acid (5-oxoproline)⁽²⁾

Acidosis Without Elevated Anion Gap

Hyperchloremic metabolic acidosis with a normal anion gap, that is, when a decrease in bicarbonate ions corresponds with an increase in chloride ions⁽²⁾
Respiratory acidosis
For all causes of acidosis unrelated to elevated anion gap, see also the following topics:
- Hyperchloremic Metabolic Acidosis
- Respiratory Acidosis

Mimics

Pseudo-anion gap metabolic acidosis can occur in the setting of venous blood gas drawn from a heparinized tube (NDT Plus 2008 Apr;1(2):94 full-text).

Serum Anion Gap (AG) in Conditions Other Than Metabolic Acidosis

Serum AG can be used in the differential diagnosis of metabolic acidosis and other acid-base disorders.
- An elevated serum AG (> the upper limit of normal range) may indicate various conditions other than metabolic acidosis, including metabolic alkalosis, respiratory alkalosis, hyperphosphatemia, and IgA paraproteinemia.
- A low serum AG (< the lower limit of normal range) is less frequent and is typically due to laboratory error, but it may also indicate hypoalbuminemia, monoclonal IgG gammopathy, polyclonal gammopathy, intoxication with lithium or bromide, and possibly hypercalcemia.
- A negative serum AG is uncommon and is typically due to laboratory error, but it may also be suggestive of multiple myeloma or intoxication with bromide or iodide.
- Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
A serum AG is the difference between serum sodium (measured cation) and measured anions.⁽²⁾View full sizeAG = [Na+] – {[Cl-] + [HCO3-]}
Normal serum AG values:
- Normal serum AG values can vary due to the differences in methods of measurement, as well as individual variability.
- The following average values and ranges of AG in healthy individuals have been reported:
  - Average AG of 11 mEq/L (11 mmol/L), with a range of 6-16 mEq/L (6-16 mmol/L)
  - Average AG of 12 mEq/L (12 mmol/L), with a range 8-16 mEq/L (8-16 mmol/L)
  - Average AG of 15 mEq/L (15 mmol/L), with a range 10-20 mEq/L (10-20 mmol/L)
- Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
The use of own laboratory’s reference range to determine the normal values of AG is recommended due to the dependency on the type of measurement instrument used.^(2, 3)
Elevated serum AG:
- An elevated serum AG is more common compared to low serum AG (Clin J Am Soc Nephrol 2007 Jan;2(1):162).
- Causes of elevated serum AG other than metabolic acidosis (which is the most common) may include metabolic alkalosis, respiratory alkalosis, hyperphosphatemia, IgA paraproteins, and laboratory errors.
  - Metabolic alkalosis:
    - Metabolic alkalosis is an increase in pH (pH > 7.42) due to a primary elevation in serum concentration of bicarbonate (> 26 mEq/L [26 mmol/L]), which may occur due to loss of gastric fluid or diuretics use.
    - Metabolic alkalosis has been reported to be associated with an increase in serum AG of about 4-6 mEq/L (4-6 mmol/L) mostly due to an increase in albumin concentration, in the absence of disorders which might increase it such as organic acidosis or renal failure.
    - Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
  - Respiratory alkalosis:
    - Respiratory alkalosis is characterized by a pH of > 7.42 and a partial pressure of arterial carbon dioxide (PaCO₂) of < 38 mm Hg.
    - The serum AG does not appear to change significantly, but small increases of about 3 mEq/L (3 mmol/L) of AG, for a chronic reduction in PaCO₂ of 20 mm Hg, have been reported.
    - Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
  - Severe hyperphosphatemia:
    - Severe hyperphosphatemia appears to be an important factor in a marked elevation of serum AG.
    - A serum phosphate concentration of 19-23 mg/dL due to exogenous phosphate administration has been reported to be associated with a serum AG of 50 mEq/L (50 mmol/L).
    - Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
  - IgA paraproteins:
    - IgA paraproteins are typically anions (Clin J Am Soc Nephrol 2007 Jan;2(1):162).
    - The impact of IgA myeloma on serum AG has been reportedly conflicting (Clin J Am Soc Nephrol 2007 Jan;2(1):162).
    - IgA monoclonal gammopathy associated with elevated serum AG
      - based on case-control study
      - 242 patients (median age 63-71 years) with monoclonal gammopathies matched to 40 controls (mean age 56 years) without monoclonal gammopathy
      - 109 patients (45%) had IgG, 64 (26.4%) had IgA, 21 (8.6%) had IgM, and 48 (20%) had light chain monoclonal gammopathy
      - 87 patients (36%) had hypoalbuminemia (albumin < 3.6 g/dL) requiring correction
      - normal serum AG range defined as 10-15 mEq/L (10-15 mmol/L)
        
        elevated serum AG defined as > 15 mEq/L (15 mmol/L)
        
        low serum AG defined as < 10 mEq/L (10 mmol/L)
      - comparing elevated vs. low serum AG
        
        in patients with IgA monoclonal gammopathy, 31% vs. 9% (p = 0.002)
        
        in patients with IgG monoclonal gammopathy, 7% vs. 22% (p = 0.02)
        
        no significant difference in patients with IgM or light chain monoclonal gammopathies
      - Reference – Clin J Am Soc Nephrol 2011 Dec;6(12):2814 full-text
  - Laboratory errors may also cause elevated serum AG.
    - Elevated serum AGE may result from overestimations of sodium levels and underestimations of chloride and bicarbonate levels.
    - When the serum is separated and exposed to air for > 1 hour, bicarbonate levels will decrease and the serum AG will increase.
    - Reference – Clin J Am Soc Nephrol 2007 Jan;2(1):162
Low or negative serum AG:
- Low serum AG values have been reported in in 0.8%-3% of measurements, and 90% of them has been reported to be due to laboratory errors.
- Negative serum AG values have been reported in about 0.12% of measurements.
- The causes leading to low or negative serum AG may include laboratory errors, hypoalbuminemia, IgG paraproteins, increased unmeasured cations, such as calcium and magnesium, lithium and bromide toxicity, and aspirin poisoning.
  - Laboratory errors are the most frequent cause of low or negative serum AG.
    - Some are random, nonreproducible laboratory errors.
    - Some reproducible artifacts affecting determination of serum AG include underestimation of serum sodium, overestimation of serum chloride, and overestimation of serum bicarbonate.
      - Underestimation of serum sodium may occur in the presence of hypernatremia or hypertriglyceridemia.
      - Overestimation of serum chloride may occur in the presence of hypertriglyceridemia; however, it is rare if serum chloride is determined by using direct ion-selective electrodes.
      - Overestimation of serum bicarbonate (up to 2-3 mEq/L [2-3 mmol/L]) may occur if serum is not separated from cellular elements immediately following specimen collection.
    - Reference – CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143
  - Hypoalbuminemia is the second most common cause of low serum AG.
    - Hypoalbuminemia may lead to a low serum AG but not a negative serum AG.
    - Low levels of albumin may result in a decrease in unmeasured anions.
      - For every 1 g/dL change in serum albumin, it has been reported that there is a change in the serum AG of 2.3-2.5 mEq/L (2.3-2.5 mmol/L).
      - The formula to correct for low albumin is as follows:View full sizecorrected AG = uncorrected AG + {2.5 × (normal albumin – observed albumin)}
        
        AG is in mEq/L (mmol/L).
        
        Albumin level is in g/dL.
        
        See DynaMed calculator for Anion Gap in Hypoalbumin States.
    - Reference – CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143
  - IgG paraproteins:
    - IgG paraproteins are typically cations (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - IgG can increase serum chloride levels without altering sodium and bicarbonate levels, resulting in a low or negative AG (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - Conditions associated with IgG paraprotein that may cause a low or negative serum AG include monoclonal IgG gammopathy, polyclonal gammopathy, and multiple myeloma (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - The serum AG has been reported to be lower by > 50% in patients with polyclonal IgG gammopathy compared to normal control individuals (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - IgG monoclonal gammopathy associated with low serum AG
      - based on case-control study
      - 242 patients (median age 63-71 years) with monoclonal gammopathies matched to 40 controls (mean age 56 years) without monoclonal gammopathy
      - 109 patients (45%) had IgG, 64 (26.4%) had IgA, 21 (8.6%) had IgM, and 48 (20%) had light chain monoclonal gammopathy
      - 87 patients (36%) had hypoalbuminemia (albumin < 3.6 g/dL) requiring correction
      - normal serum AG range defined as 10-15 mEq/L (10-15 mmol/L)
        
        elevated serum AG defined as > 15 mEq/L (15 mmol/L)
        
        low serum AG defined as < 10 mEq/L (10 mmol/L)
      - comparing low vs. elevated serum AG
        
        in patients with IgG monoclonal gammopathy, 22% vs. 7% (p = 0.02)
        
        in patients with IgA monoclonal gammopathy, 9% vs. 31% (p = 0.002)
        
        no significant difference in patients with IgM or light chain monoclonal gammopathies
      - Reference – Clin J Am Soc Nephrol 2011 Dec;6(12):2814 full-text
  - Increased unmeasured cations, such as calcium and magnesium:
    - Unmeasured cations may reduce serum AG only if they accumulate in the serum along with chloride or bicarbonate ions.
    - Hypercalcemia that is associated with primary hyperparathyroidism has been reported to reduce serum AG to a greater magnitude compared to hypercalcemia that is associated with malignancy.
    - Hypermagnesemia may not affect serum AG, but magnesium administration as a chloride or bicarbonate salt may reduce serum AG.
    - Reference – CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143
  - Lithium toxicity:
    - Lithium is commonly prescribed for the treatment of bipolar disorder (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - Lithium concentrations of > 4 mEq/L (4 mmol/L) have been reported to decrease serum AG (CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143).
    - A negative serum AG of -2 mEq/L (-2 mmol/L) due to lithium intoxication (lithium level 14.5 mEq/L [14.5 mmol/L]) has been reported in a case report of a 65-year-old male adult (CMAJ 2007 Mar 27;176(7):921 full-text).
  - Bromide toxicity:
    - Bromide toxicity is a common cause of negative serum AG after laboratory errors.
    - Bromide-containing medications have largely been removed, but bromide may still be present in some sedatives, some herbal medications, and the treatment of myasthenia gravis as pyridostigmine bromide.
    - An increased bromide concentration may lead to low or negative serum AG as it interferes with chloride measurement, causing spurious increase of its concentration.
    - Bromide has been reported to lead to negative serum AG of up to -60 mEq/L (-60 mmol/L).
    - Reference – CMAJ 2007 Mar 27;176(7):921 full-text, Am J Kidney Dis 2016 Jan;67(1):143
  - Aspirin poisoning (salicylate level 84.6 mg/mL) has been reported to be associated with negative serum AG of -47 mEq/L (-47 mmol/L) in a case report of a 47-year-old female adult (Am J Kidney Dis 2016 Jan;67(1):143).

History and Physical

Clinical Presentation

The clinical presentation of anion gap metabolic acidosis varies by its underlying cause.
- The symptoms and signs of lactic acidosis related to shock and tissue hypoperfusion include the following:
  - Tachycardia
  - Tachypnea
  - Hypotension
  - Decreased urine output
  - Altered mental status
  - References
    - Postgrad Med 2000 Mar;107(3):249
    - N Engl J Med 2014 Dec 11;371(24):2309 full-text, editorial can be found in N Engl J Med 2015 Mar 12;372(11):1078
- Symptoms and signs of diabetic ketoacidosis include the following:
  - Acetone on breath (fruity smell)
  - Polyuria
  - Polydipsia
  - Nausea and vomiting
  - Abdominal pain
  - Weakness
  - Lethargy
  - Kussmaul respirations (deep respirations)
  - Altered mental state, including coma, severe cases
  - Reference – BMJ 2015 Oct 28;351:h5660, correction can be found in BMJ 2015 Nov 2;351:h5866
- Clinical presentation of alcoholic ketoacidosis:
  - Symptoms and signs of alcoholic ketoacidosis are typically nonspecific, and they can include the following:
    - Abdominal pain and vomiting, often due to alcohol-induced gastritis or pancreatitis (N Engl J Med 2018 Jan 11;378(2):202)
    - Nausea (Emerg Med Clin North Am 2014 May;32(2):293)
    - Diarrhea (Oxf Med Case Reports 2016 Mar;2016(3):31)
  - Alcoholic ketoacidosis typically happens after a period of bingeing, followed by the cessation of drinking (Emerg Med Clin North Am 2014 May;32(2):293).
- Symptoms and signs of methanol poisoning are as follows:
  - Signs and symptoms of the central nervous system, such as coma
  - Decreased vision
  - Gastrointestinal signs and symptoms, such as abdominal pain
  - Pulmonary dysfunction
  - Parkinson-like symptoms, which are rare
  - Reference – N Engl J Med 2018 Jan 18;378(3):270, correction can be found in N Engl J Med 2019 Jan 10;380(2):202
- Symptoms and signs of ethylene glycol intoxication, due to oxalate crystal formation in lungs, heart, and kidneys, are as follows:
  - Neurologic dysfunction, which typically happens ≤ 12 hours after exposure
  - Cardiac and pulmonary dysfunction, which typically happens 12-24 hours after exposure
  - Acute kidney injury, which typically happens 48-72 hours after exposure
  - Reference – N Engl J Med 2018 Jan 18;378(3):270, correction can be found in N Engl J Med 2019 Jan 10;380(2):202
- Symptoms and signs of salicylate poisoning are as follows:
  - Tinnitus⁽¹⁾
  - Altered mental status⁽¹⁾
  - Tachycardia⁽¹⁾
  - Tachypnea (Emerg Med J 2002 May;19(3):206 full-text)
  - Hyperthermia⁽¹⁾
  - Pulmonary edema⁽¹⁾
- Symptoms and signs of carbon monoxide (CO) poisoning:
  - Clinical presentation of CO poisoning can range from no symptoms or mild, flu-like symptoms to coma and death, depending on the severity of exposure.
  - Some nonspecific symptoms following acute CO poisoning can include headache (most common), fatigue, shortness of breath, chest pain, nausea and/or vomiting, myalgias, dizziness, confusion, memory loss, impaired concentration or language, ataxia, affective changes, such as anxiety or depression, parkinsonism, and loss of consciousness.
  - Additional symptoms in patients with CO poisoning due to methylene chloride may include skin or eye irritation and cough or airway irritation.
  - Some nonspecific symptoms following chronic CO poisoning (intermittent, long-term, subacute exposures to CO lasting > 24 hours) can include fatigue, mood disorder and emotional distress, memory problems, sleep problems, vertigo, neuropathy, paresthesias, abdominal pain, and diarrhea.
- Common symptoms and signs of propylene glycol poisoning are as follows:
  - Cardiac arrhythmias
  - Convulsions
  - Coma
  - Extreme agitation
  - Other severe neuropsychiatric signs and symptoms
  - Reference – Clin Toxicol (Phila) 2015;53(7):589

History

Medication History

Ask about the use of following medications:
- Aspirin^(1,2)
- Metformin^(1,2)
- Isoniazid^(1,2)
- Iron supplements^(1,2)
- Propofol, an anesthetic^(1,2)
- Niacin supplements⁽²⁾
- Antiretrovirals, such as nonnucleoside reverse-transcriptase inhibitors^(1,2)
Review the use of recent medication infusions that use propylene glycol as the delivery vehicle, such as the following:⁽¹⁾
- Lorazepam
- Diazepam
- Phenytoin
- Etomidate
- Nitroglycerin
- Esmolol

Past Medical History (PMH)

Ask about diabetes and kidney failure.⁽¹⁾
Ask about the following conditions that predispose lactic acidosis:
- Sepsis⁽²⁾
- Heart failure⁽²⁾
- Seizure disorder⁽²⁾
- Liver disease^(1,2)
- Malignancy⁽¹⁾
- Glycogen storage disease⁽¹⁾
- Previous small bowel obstruction, short bowel syndrome, or blind bowel loop, which are risk factors for D-lactic acidosis⁽¹⁾

Social History (SH)

Patients may have the following history:⁽¹⁾
- Significant alcohol use/abuse
- Exposure to smoke, which can contain carbon monoxide, or cyanide
- Exposure to chemicals, such as toluene
- Ingestion of methanol (‘wood alcohol’), which can be found in the following fluids:
  - Paints
  - Solvents
  - Antifreeze
  - Fuel for outdoor stoves
  - Shellac
  - Varnish
  - Automotive fluids
- Ingestion of ethylene glycol, which can be found in the following fluids:
  - Antifreeze
  - Brake fluid
  - Deicing solution

Diagnostic Testing

Testing Overview

Evaluate the clinical presentation to assess the possible underlying causes of the acid-base disorder.
Measure arterial blood gas and serum electrolytes to determine if metabolic acidosis is present (pH < 7.38 and bicarbonate [HCO₃^–] < 22 mEq/L [22 mmol/L]).
Estimate the expected respiratory compensation to determine if there is another primary respiratory acidosis or alkalosis process in addition to metabolic acidosis.
Measure serum anion gap.
- If the serum anion gap is normal, see Hyperchloremic Metabolic Acidosis for additional information on diagnosis and management.
- If serum anion gap is elevated, causes of acidosis are likely the following:
  - Lactic acidosis
  - Ketoacidosis
  - Acidosis caused by medications, toxins, or poisons
  - Acidosis related to renal failure (uremia)
  - See Differential Diagnosis for a list of possible underlying causes.
Further evaluation in patients with elevated anion gap:
- Calculate the serum osmolal gap to determine if methanol, ethylene glycol, or propylene glycol poisoning is present, especially if there is a suspicion of ingestion or if patients have altered mental status.
- Measure serum lactate level to determine if lactic acidosis is present.
- Measure serum ketone level to determine if ketoacidosis is present.
- Estimate the glomerular filtration rate to determine if acidosis is related to renal failure.^(1,2)
- Calculate Delta-Delta gradient and ratio to determine if there is metabolic alkalosis or hyperchloremic metabolic acidosis in addition to anion gap metabolic acidosis.

Initial Testing

Blood Gas Analysis

Check the arterial blood gas (ABG) to determine if respiratory compensation or mixed acid-base disorder is present.⁽²⁾
- Metabolic acidosis is characterized by pH < 7.38, bicarbonate (HCO₃^–) < 22 mEq/L (22 mmol/L).
- Determine if respiratory compensation is present.
  - Use predicted partial pressure of arterial carbon dioxide (PaCO₂) = 1.5 × (HCO₃^–) + 8 ± 2 mm Hg (Winter’s formula) to determine respiratory compensation.
  - If observed PaCO₂ is lower than the predicted, respiratory alkalosis is also present.
  - If observed PaCO₂ is higher than predicted, respiratory acidosis is also present.
EVIDENCE SYNOPSIS: Venous blood gas may be useful for determining pH and bicarbonate levels, but it may not be as accurate for determining partial pressure of oxygen and carbon dioxide, which are important when evaluating mixed acid base disturbances, especially if there is a suspected respiratory component.
- venous pH and venous HCO₃^– appear to correlate well with arterial values, while normal venous PCO₂ may help rule out elevated arterial PCO₂ levels for patients in the emergency department
  - based on systematic review of observational studies
  - systematic review of 19 studies evaluating peripheral arterial and venous blood gas values in adult patients in the emergency department
  - conditions that patients had included chronic obstructive pulmonary disease (COPD) and other respiratory conditions, diabetic ketoacidosis, uremia, and tricyclic antidepressant overdose
  - comparing venous to arterial values
    - mean difference in pH was -0.033 units (95% CI -0.039 to -0.027) with narrow 95% limits of agreement in analysis of 15 studies with 1,805 patients
    - mean difference in HCO₃^– was 1.03 (95% CI 0.56-1.50) with narrow 95% limits of agreement in analysis of 12 studies with 1,481 patients
    - mean difference in PCO₂ was 4.41 mm Hg (95% CI 2.55-6.27) with broad 95% limits of agreement in analysis of 13 studies with 1,628 patients
  - 4 studies evaluated using peripheral venous PCO₂ to rule out elevated arterial PCO₂
    - venous PCO₂ of > 45 mm Hg had 100% sensitivity for detecting arterial PCO₂ of > 50 mm Hg in analysis of 2 studies
    - venous PCO₂ of < 45 mm Hg had 100% negative predictive value for arterial PCO₂ of > 50 mm Hg in 1 study
  - Reference – Eur J Emerg Med 2014 Apr;21(2):81
  - similar results can be found in systematic review of 6 studies evaluating arterial and venous blood gas values in patients with COPD or respiratory distress (Eur J Emerg Med 2010 Oct;17(5):246)
- venous pH and venous HCO₃^– levels appear to correlate well with arterial values but not PCO₂ and PO₂ in intensive care unit (ICU) patients
  - based on 3 cross-sectional studies
  - 110 ICU patients had 168 matched samples of central venous and arterial blood gas taken within 5 minutes of each other
    - mean difference between arterial and central venous levels were
      - -0.03 for pH (95% CI -0.07 to 0.01)
      - 0.52 mmol/L for HCO₃^– (95% CI -1.81 to 2.58)
      - 0.08 mmol/L for lactate (95% CI -0.27 to 0.42)
      - 0.19 mmol/L for base excess (95% CI -1.86 to 2.24)
  - Reference – Emerg Med J 2006 Aug;23(8):622 full-text
  - 40 ICU patients had 221 paired central venous and arterial blood gas levels checked
    - mean difference between arterial and central venous levels were
      - 0.027 for pH (95% CI -0.028 to 0.081)
      - -0.8 mEq/L for HCO₃^– (95% CI -4.0 to 2.4)
      - -3.8 mm Hg for PCO₂ (95% CI -12.3 to 4.8)
    - Reference – Clin J Am Soc Nephrol 2010 Mar;5(3):390 full-text
  - 45 ICU patients aged 15-80 years had peripheral venous and arterial blood gas levels obtained
    - correlation coefficient for pH, HCO₃^–, and base excess all exceeded 0.9, showing high correlation
    - PCO₂ (r = 0.83), not considered highly correlated
    - partial pressure of oxygen (PO₂) had low correlation (r = 0.21)
    - Reference – Anesth Essays Res 2013 Sep-Dec;7(3):355 full-text
- arterial and venous blood gas values do not appear to have close correlation in mechanically ventilated patients
  - based on cross-sectional study
  - 102 patients in the ICU on mechanical ventilation had arterial and venous blood gas checked simultaneously
  - venous and arterial pH, PCO₂, HCO₃^–, base excess, and PO₂ did not have close correlation by Pearson test of correlation (close correlation considered if r ≥ 0.9)
  - Reference – Tanaffos 2012;11(4):30 full-text
Check carboxyhemoglobin levels if carbon monoxide exposure is suspected.⁽¹⁾

Blood Tests

Overview

Check serum electrolytes and albumin to determine if serum anion gap is elevated or normal, correcting for low albumin.
Calculate delta-delta gradient and/or ratio to determine if other concomitant acid-base disorders are present along with anion gap metabolic acidosis.
Check serum osmolality, electrolytes, blood urea nitrogen (BUN), glucose, and alcohol to calculate serum osmolal gap to determine the presence of unmeasured solute in the serum.
Check for serum lactate and ketones to determine if lactic acidosis and/or ketoacidosis are present.

Serum Anion Gap

Check serum electrolytes and serum albumin to measure serum anion gap (AG) to determine if AG is elevated or normal.
- AG is the difference between serum sodium (measured cation) and measured anions.⁽²⁾View full sizeAG = [Na⁺] – {[Cl^–] + [HCO₃^–]}
- AG value is dependent on the type of measurement instrument used. Physicians should use their own laboratory’s reference range to determine what is normal.^(2,3)
- Ideally, one should compare the new anion gap to the patient’s baseline anion gap if it is known.⁽³⁾
- The reference ranges based on venous bicarbonate values may be 1-2 mEq/L higher than those based on arterial values.⁽³⁾
- See DynaMed calculator for Anion Gap.
- The formula to correct for low albumin is as follows:^(2,3)View full sizecorrected AG = uncorrected AG + {2.5 × (normal albumin – observed albumin)}
  - Anion gap is in mmol/L.
  - Albumin is in g/dL.
  - See DynaMed calculator for Anion Gap in Hypoalbumin States.
- AG > 10 mmol/L above the upper limit of the reference value is very suggestive of organic acidosis. A minor increase is less helpful for diagnosis.⁽²⁾

Delta-delta Gradient and/or Ratio

Calculate the Delta-Delta gradient or Delta-Delta ratio.
- Delta-Delta gradient or Delta-Delta ratio are used to determine if other concomitant acid-base disorders are present along with anion gap metabolic acidosis. For example, hyperchloremic metabolic acidosis from shock and dehydration can have both hyperchloremia and hyponatremia, which can lead to the presence of a high anion gap.^(1,2,3)
- Delta-Delta gradient (delta gap):⁽²⁾
  - The formula to calculate Delta-Delta gradient if there is ketoacidosis is as follows:View full sizedelta-delta gradient = change in anion gap – change in bicarbonate
  - The formula to calculate Delta-Delta gradient if there is lactic acidosis is as follows:View full sizedelta-delta gradient = 0.6(change in anion gap) – change in bicarbonate
  - If delta-delta gradient is from -5 to +5, simple anion gap metabolic acidosis is present.
  - If delta-delta gradient is > +5, metabolic alkalosis or respiratory acidosis is also present.
  - If delta-delta gradient is < -5, hyperchloremic metabolic acidosis or respiratory alkalosis is also present.
  - See DynaMed calculator for Anion Gap Delta Delta Gradient.
  - CLINICIANS’ PRACTICE POINT: If the baseline anion gap is not known, use the highest normal reference value as the baseline. If the baseline bicarbonate value is not known, use the lowest normal reference value as the baseline.
- Delta-Delta ratio (delta ratio):⁽¹⁾
  - Delta-Delta ratio is the ratio of the change of anion gap to the change of bicarbonate.
  - If delta ratio is 0.8-1.2, simple anion gap metabolic acidosis is present.
  - If delta ratio is 0.3-0.7, hyperchloremic metabolic acidosis or respiratory alkalosis is also present.
  - If delta ratio is > 1.2, metabolic alkalosis or respiratory acidosis is also present.
  - See also DynaMed calculator for Anion Gap Delta Delta Ratio.

Serum Lactate and Ketones

Measure serum lactate.^(1,2)
- Lactic acidosis is defined as an elevated serum lactate in the presence of metabolic acidosis.
- Consider free-flowing arterial sample for the greatest accuracy for measuring serum lactate.
- High lactate levels (> 10 mEq/L) may be seen with cyanide poisoning.
Check for serum ketones (beta-hydroxybutyrate and acetoacetate).⁽¹⁾
- If there is alcoholic and starvation ketoacidosis, the ratio of beta-hydroxybutyrate to acetoacetate is about 7:1, and the serum blood glucose is normal or low.
- If there is diabetic ketoacidosis, the ratio of beta-hydroxybutyrate to acetoacetate is about 3:1 and the serum blood glucose is elevated.
- Metabolic acidosis with no ketones may be seen in methanol and ethylene glycol poisoning.

Serum Osmolal Gap (OG)

Measure serum osmolality to calculate the serum OG.^(1,2)
- OG helps determine the presence of unmeasured solutes in the serum.⁽¹⁾View full sizeOG = (measured – calculated) serum osmolality
- Calculated serum osmolality = (2 × Na) + glucose/18 + BUN/2.8 + ethanol/4.6.
- Osmolal gap of > 10 mOsm/kg is normal.
- The combination of both elevated anion gap and high osmolal gap may indicate poisoning with methanol, ethylene glycol, or propylene glycol (in hospitalized patients).
See DynaMed calculator for Osmolal Gap Calculator.

Urine Studies

Urine studies in patients with suspected ethylene glycol intoxication:⁽¹⁾
- Urinalysis may reveal calcium oxalate crystals.
- Urine may be fluorescent under the Wood lamp.

Management

Management Overview

Management should focus on reversing the underlying cause of acidosis.⁽¹⁾
Consider dialysis for the removal of toxins or drugs contributing to lactic acidosis, or if there are severe, life-threatening acidosis or renal failure (⁽¹⁾,N Engl J Med 2014 Dec 11;371(24):2309).

Management of Metabolic Acidosis

Sodium Bicarbonate for Metabolic Acidosis

Oral or IV sodium bicarbonate is often used as a temporizing measure, but the evidence for its benefits is limited (Nephrol Dial Transplant 2015 Jul;30(7):1104).
Sodium bicarbonate is given over a period of several minutes to hours unless very extreme acidemia is present (N Engl J Med 1998 Jan 1;338(1):26).
Serious potential risks of sodium bicarbonate:⁽¹⁾
- It may lead to hypokalemia and hypocalcemia from intracellular shifting, increasing the risk of arrhythmia.
- It may lead to generation of carbon dioxide and worsening of respiratory acidosis.
- It can lead to hypernatremia and contribute to volume expansion.
- It may increase lactate production by increasing pyruvate dehydrogenase activity.
Bicarbonate is generally not recommended or associated with improved outcome in patients with lactic acidosis, diabetic ketoacidosis, or alcohol ketoacidosis.
Bicarbonate may be useful in the following patients:
- Select patients with chronic kidney disease or acute kidney injuries
- Patients with methanol and ethylene glycol poisoning
- Select patients with salicylate poisoning
- Patients with cyanide poisoning who have shock, seizures, or arrhythmias
sodium bicarbonate may not reduce composite risk of death and organ failure, but may reduce need for renal replacement therapy in adults admitted to intensive care unit (ICU) with severe metabolic acidemia (level 2 [mid-level] evidence)
- based on randomized trial without blinding
- 400 adults (mean age 65 years) admitted to ICU within 48 hours of severe metabolic acidemia were randomized to 1 of 2 groups
  - sodium bicarbonate 0.5 mEq/mL IV with target arterial pH > 7.3 (each infusion 125-250 mL in 30 minutes, maximum 1,000 mL in 24 hours)
  - no sodium bicarbonate
- severe acidemia defined as pH ≤ 7.2, partial pressure of arterial carbon dioxide (PaCO₂) ≤ 45 mm Hg, and sodium bicarbonate ≤ 20 mmol/L
- all patients had total Sequential Organ Failure Assessment score ≥ 4 points or arterial lactate ≥ 2 mmol/L
- 46% had acute kidney injury
- 81% were on invasive mechanical ventilation and 78% received vasopressors
- primary outcome was composite of all-cause death at 28 days and ≥ 1 organ failure at 7 days
- 23% in control group were given sodium bicarbonate at median 7 hours after randomization mainly as salvage therapy
- 97% completed 28-day follow-up and were included in analyses
- comparing sodium bicarbonate vs. no sodium bicarbonate
  - primary outcome in 66% vs. 71% (not significant)
  - 28-day mortality 45% vs. 54% (p = 0.07)
  - ≥ 1 organ failure at 7 days in 62% vs. 69% (not significant)
  - renal replacement therapy during ICU stay in 35% vs. 52% (p = 0.0009, NNT 6)
  - median renal replacement therapy-free days during ICU stay 19 days vs. 8 days (p = 0.015)
- no significant differences in invasive mechanical ventilation-free days, vasopressor-free days, length of ICU stay, or ICU-acquired infections
- no life-threatening complications reported, but metabolic alkalosis, hypernatremia, and hypocalcemia occurred more frequently in sodium bicarbonate group
- see here for results in subgroup of patients with acute kidney injury
- Reference – BICAR-ICU trial (Lancet 2018 Jul 7;392(10141):31), editorial can be found in Lancet 2018 Jul 7;392(10141):3
oral sodium bicarbonate may not improve physical function in adults ≥ 60 years old with advanced chronic kidney disease and mild acidosis (level 2 [mid-level] evidence)
- based on randomized trial with high loss to follow-up
- 300 adults ≥ 60 years old (mean age 74 years, 71% male, 96% White) with advanced chronic kidney disease and mild acidosis (serum bicarbonate < 22 mmol/L) were randomized to sodium bicarbonate 500 mg orally 3 times daily vs. placebo for 24 months
- patients with serum bicarbonate < 22 mmol/L at 3 months had sodium bicarbonate or placebo dose increased to 1 g 3 times daily
- primary outcome was Short Physical Performance Battery (SPPB) score at 12 months, score range 0-12 points, with higher score indicating better lower limb strength and balance, and minimal clinically important difference defined as 1 point
- baseline measures, comparing sodium bicarbonate group vs. placebo group
  - mean SPPB score 8 points vs. 8.1 points
  - mean 6-minute walk distance 304 meters vs. 317 meters
- 50% in both groups had adherence ≤ 80%
- 73% completed 12-month follow-up
- comparing sodium bicarbonate vs. placebo at 12 months
  - mean SPPB score 8.3 points vs. 8.8 points (not significant)
  - mean 6-minute walk distance 294 meters vs. 336 meters (no p value reported)
  - adverse event in 86.1% vs. 89.1% (no p value reported)
  - death in 9.9% vs. 7.4% (no p value reported)
  - dialysis or renal transplantation in 21.7% vs. 22.3% (no p value reported)
  - ≥ 1 fall in 32.2% vs. 26.4% (no p value reported)
- Reference – BiCARB trial (Health Technol Assess 2020 Jun;24(27):1 full-text), commentary can be found in Ann Intern Med 2020 Nov 17;173(10):JC56
higher amounts of sodium bicarbonate therapy associated with increased mortality in trauma patients with severe mixed metabolic and respiratory acidosis undergoing emergency surgery (level 2 [mid-level] evidence); bicarbonate therapy associated with increased pH but also appears to increase arterial partial pressure of carbon dioxide (level 3 [lacking direct] evidence)
- based on retrospective cohort study
- 225 trauma patients undergoing emergency surgery with severe acidosis (pH ≤ 7.1) (most patients had a combined metabolic and respiratory acidosis) were evaluated
  - 56% of patients were classified as early deaths
    - 28% died in operating room
    - 28% died ≤ 48 hours post surgery
  - 44% of patients were classified as early survivors
    - 11% died > 48 hours post surgery (mean 15 days after surgery)
    - 33% survived beyond hospitalization
- 85% of patients received sodium bicarbonate therapy 2-8 vials (50 mEq per vial) during emergency trauma surgery
  - mortality in
    - 93% of 56 patients receiving ≥ 7 vials (p < 0.001 vs. patients receiving 1-2 vials)
    - 70% of 87 patients receiving 3-6 vials (no p value reported)
    - 51% of 49 patients receiving 1-2 vials
  - in subset of 73 patients receiving 2-8 vials (50 mEq per vial) of sodium bicarbonate in whom treatment and hemodynamics remained relatively consistent, sodium bicarbonate associated with
    - increased arterial pH (from 6.99 to 7.14, p < 0.001)
    - increased arterial partial pressure of carbon dioxide (from 45 to 51 mm Hg, p < 0.001)
    - increased end-title partial pressure of carbon dioxide (from 18 to 25 mm Hg, p < 0.001)
    - increased arterial bicarbonate levels (from 10.5 to 16.6 mEq/L, p < 0.001)
- Reference – J Trauma Acute Care Surg 2013 Jan;74(1):45

Veverimer For Metabolic Acidosis

Veverimer (TRC101) (not available in the United States) is an oral nonabsorbed polymer that selectively binds to hydrochloric acid and removes it from the gastrointestinal tract, effectively increasing the serum bicarbonate levels (Lancet 2019 Aug 3;394(10196):396).
veverimer associated with functional improvement and increased bicarbonate levels without increase in adverse events in patients with CKD and metabolic acidosis (level 2 [mid-level] evidence)
- based on randomized trial with possible bias due to manufacturer involvement
- 196 patients with chronic kidney disease (estimated GFR 20-40 mL/minute/1.73 m²) and metabolic acidosis (serum bicarbonate 12-20 mmol/L) were randomized to veverimer 6 g/day as oral suspension in water vs. placebo (microcrystalline cellulose) for 40 weeks
- veverimer is an oral nonabsorbed polymer that selectively binds and eliminates hydrochloric acid from gastrointestinal tract
- veverimer manufacturer (Tricida) was involved in design, conduct, analysis, interpretation of data, and preparation of manuscript
- comparing veverimer vs. placebo
  - adverse events in 81% vs. 80% (no p value reported)
  - serious adverse events in 2% vs. 5% (no p value reported)
  - renal system adverse events in 8% vs. 15% (no p value reported)
  - premature treatment discontinuation in 3% vs. 10% (no p value reported)
  - increase in serum bicarbonate concentration (≥ 4 mmol/L or normalization) in 63% vs. 38% (p = 0.0015)
- veverimer associated with improved patient-reported physical functioning based on Kidney Disease and Quality of Life-Physical Function Domain (KDQoL-PFD) total score (p < 0.0001)
- no patients in veverimer group discontinued treatment due to adverse events
- Reference – Lancet 2019 Aug 3;394(10196):396, editorial can be found in Lancet 2019 Aug 3;394(10196):363
veverimer may increase serum bicarbonate level at 12 weeks in adults ≤ 85 years old with non-dialysis-dependent chronic kidney disease and metabolic acidosis (level 3 [lacking direct] evidence)
- based on nonclinical outcome in randomized trial with limited follow-up
- 217 adults ≤ 85 years old (mean age 62 years, 52% ≥ 65 years old, 62% male, 97% White) with non-dialysis-dependent chronic kidney disease (estimated glomerular filtration rate [GFR] 20-40 mL/minute/1.73 m²) and metabolic acidosis (serum bicarbonate 12-20 mmol/L) were randomized to veverimer vs. placebo for 12 weeks
  - veverimer started at 6 g orally once daily and titrated from week 4 to achieve target serum bicarbonate of 22-29 mmol/L
  - patients fasted for ≥ 4 hours (water only) before serum bicarbonate measurement
- response defined as ≥ 4 mmol/L increase from baseline in serum bicarbonate or serum bicarbonate in normal range (22-29 mmol/L)
- 96% included in analysis of outcome of response, 100% included in safety analysis
- comparing veverimer vs. placebo at 12 weeks
  - response in 59% vs. 22% (p < 0.0001, NNT 3)
    - ≥ 4 mmol/L increase in serum bicarbonate at week 12 in 56% vs. 21% (p < 0.0001, NNT 3)
    - serum bicarbonate in normal range at week 12 in 50% vs. 17% (p < 0.0001, NNT 3)
  - mean change from baseline in serum bicarbonate 4.5 mmol/L vs. 1.7 mmol/L (p < 0.0001)
  - any adverse event in 54% vs. 46% (no p values reported)
    - gastrointestinal disorder in 17% vs. 9%
    - treatment-related gastrointestinal adverse event in 13% vs. 5%
    - metabolism and nutrition disorder in 13% vs. 9%
    - paraesthesia in 1 patient vs. 1 patient
- consistent results for outcome of response in subgroup analyses by age (< 65 years, ≥ 65 years), sex, serum bicarbonate level at baseline (≤ 18 mEq/L, > 18 mEq/L), and estimated GFR at screening (< 30 mL/minute/1.73 m², ≥ 30 mL/minute/1.73 m²)
- prespecified exploratory analysis of physical function
  - assessment using KDQoL Short Form-36 Physical Functioning subscale (range for minimal clinically important differences reported for KDQoL subscales 3-5 points)
  - veverimer associated with improved physical function (adjusted mean difference 5.2 points, 95% CI 1.1-9.2 points), significant, but CI includes differences that may not be clinically important
- 2 deaths (from unstable angina or pneumonia) in placebo group
- Reference – Lancet 2019 Apr 6;393(10179):1417, editorial can be found in Lancet 2019 Apr 6;393(10179):1387, commentary can be found in Aktuelle Urol 2019 Sep;50(5):475 [German]

Management of Underlying Causes of Metabolic Acidosis

Management of Lactic Acidosis

For detailed management options, see Lactic Acidosis for additional information.
Initial management of lactic acidosis should assure adequate tissue perfusion.
- If there is shock, consider volume resuscitation, vasopressors, or inotropes (Mayo Clin Proc 2013 Oct;88(10):1127 full-text).
- If there is hypoxemia, consider oxygen supplementation, including intubation and mechanical ventilation (Cleve Clin J Med 2015 Sep;82(9):615).
- If there is severe anemia with demonstrated impaired oxygen delivery, considerBlood transfusion (Cleve Clin J Med 2015 Sep;82(9):615).
- If there is local ischemia, consider surgery to restore circulation or remove damaged tissue (Mayo Clin Proc 2013 Oct;88(10):1127 full-text).
Management of the underlying cause of lactic acidosis can include the following therapies:
- Administration of appropriate antibiotics for sepsis (N Engl J Med 2014 Dec 11;371(24):2309 full-text)
- Left-ventricular assist device for advanced heart failure (N Engl J Med 2014 Dec 11;371(24):2309 full-text)
- Coronary intervention for acute myocardial infarction (N Engl J Med 2014 Dec 11;371(24):2309 full-text)
- Surgery for trauma or tissue ischemia (N Engl J Med 2014 Dec 11;371(24):2309 full-text)
- Removal of toxins by dialysis (N Engl J Med 2014 Dec 11;371(24):2309 full-text)
- Correction of diabetic ketoacidosis (Mayo Clin Proc 2013 Oct;88(10):1127 full-text)
- Thiamine replacement (Mayo Clin Proc 2013 Oct;88(10):1127 full-text)
use of sodium bicarbonate
- International Surviving Sepsis Campaign recommends against using sodium bicarbonate for the purpose of improving hemodynamics or reducing vasopressor requirements in patients with hypoperfusion-induced lactic acidemia with pH ≥ 7.15 (SCCM Weak recommendation, Moderate-quality evidence) (Crit Care Med 2017 Mar;45(3):486)
- EVIDENCE SYNOPSIS: Sodium bicarbonate does not appear to decrease mortality or improve hemodynamic variables in patients with lactic acidosis.
  - sodium bicarbonate does not appear to decrease mortality in patients with acidemia and might increase mortality in patients with sepsis who develop metabolic acidosis (including lactic acidosis) (level 2 [mid-level] evidence)
    - based on systematic review with inadequate assessment of trial quality
    - systematic review of 8 studies (4 randomized trials, 3 cohort studies, and 1 case-control) evaluating indications and benefits of sodium bicarbonate therapy in patients with sepsis who develop metabolic acidosis, including lactic acidosis
    - review did not assess allocation concealment, dropouts, or if intention-to-treat analyses performed
    - 3 studies showed no significant difference in mortality with sodium bicarbonate
      - no significant differences in 28-day mortality in 1 trial with 94 patients who were randomized to resuscitation with normal saline vs. sodium chloride 3.5% solution vs. sodium bicarbonate 5% solution
      - no significant difference in 28-day mortality in 1 case-control study comparing 36 patients with septic shock treated with sodium bicarbonate compared to 36 patients with an infection and similar predicted mortality who were not given sodium bicarbonate
      - no significant association between sodium bicarbonate administration and mortality in 1 prospective cohort study of 200 patients with severe acidemia
    - sodium bicarbonate associated with increased mortality in 1 retrospective cohort study with 103 patients with lactic acidosis
    - Reference – Crit Care Res Pract 2015;2015:605830 full-text
  - sodium bicarbonate therapy does not appear to be associated with improvements in hemodynamic variables compared to sodium chloride in patients with lactic acidosis (level 2 [mid-level] evidence)
    - based on nonclinical outcomes from 1 randomized trial with unclear blinding and 2 small randomized crossover studies
    - 94 patients with severe sepsis and hypotension were randomized to 1 of 3 types of fluid resuscitation within 15 minutes of initial treatment and followed for 28 days
      - study did not mention blinding of assessors or care providers
      - fluid resuscitation was given with 5 mL/kg of normal saline, 5 mL/kg of 3.5% sodium chloride, or 5 mL/kg of 5% sodium bicarbonate
      - no significant difference among groups regarding cardiac output, systolic blood pressure, respiratory rate after 120 minutes and after 8 hours
      - 28-day mortality was similar in all 3 groups (about 16%)
    - Reference – BMC Infect Dis 2008 Apr 17;8:50 full-text
    - 10 patients with metabolic acidosis, increased arterial plasma lactate (> 2.45 mmol/L) and without significant renal failure randomized to sequential therapy with sodium bicarbonate and sodium chloride in crossover trial
      - treatment order determined by random assignment
      - no significant difference in hemodynamic responses or tissue oxygenation between groups
      - increased arterial/venous pH, serum bicarbonate, and partial pressure of arterial/venous carbon dioxide with bicarbonate therapy
      - Reference – Crit Care Med 1991 Nov;19(11):1352
    - 14 patients with metabolic acidosis and increased arterial lactate randomized to sodium bicarbonate and equimolar sodium chloride in crossover trial
      - treatment order determined by random assignment
      - mean arterial pressure was unchanged and hemodynamic responses were equivalent
      - treatment group had increased arterial pH and serum bicarbonate partial pressure of carbon dioxide and decreased plasma ionized calcium compared with control group
      - Reference – Ann Intern Med 1990 Apr 1;112(7):492
    - no other randomized trials found in systematic review of studies evaluating benefit of sodium bicarbonate for treatment of acidosis in sepsis (Crit Care Res Pract 2015;2015:605830 full-text)

Management of Ketoacidosis

Diabetic Ketoacidosis (DKA)

General acute management of DKA:
- Consider fluid resuscitation, commonly with 0.9% saline solution.
- Consider electrolyte replacement with close monitoring of the level, if necessary, including sodium, potassium, magnesium, and phosphate.
- Consider insulin therapy to reverse acidosis, with monitoring and maintenance of the glucose level.
- If there is severe acidemia (pH < 6.9), consider sodium bicarbonate. Otherwise, bicarbonate is not recommended.
- See the following topics for details:
  - Diabetic Ketoacidosis (DKA) in Adults
  - Diabetic Ketoacidosis (DKA) in Children and Adolescents
EVIDENCE SYNOPSIS: Sodium bicarbonate does not appear to improve glycemic control in adults with DKA and pH ≥ 6.85. Even in patients with severe acidemia (pH < 7), bicarbonate does not appear to shorten time to acidosis resolution or time to hospital discharge.
- bicarbonate treatment does not appear to improve glycemic control in adults with DKA and pH ≥ 6.85 and might be associated with increased risk of cerebral edema and prolonged hospitalization in children (level 2 [mid-level] evidence); inconsistent evidence on its efficacy on improvement of metabolic acidosis
  - based on systematic review limited by clinical heterogeneity
  - systematic review of 44 studies comparing bicarbonate administration vs. no bicarbonate in patients with DKA
    - 14 case reports/series, 8 acid-base studies, 9 case-control studies, 3 randomized trials in adults, 6 case reports of cerebral edema in DKA, 4 studies on risk of cerebral edema in DKA
    - no studies included patients with initial pH < 6.85
  - there was significant clinical heterogeneity among studies in terms of pH threshold and bicarbonate administration (concentration, amount, timing)
  - no meta-analysis performed
  - improvement of metabolic acidosis
    - significant increase in pH at 2 hours with bicarbonate treatment in 2 randomized trials with 52 adults
    - no improvement in pH comparing bicarbonate to no bicarbonate in 1 randomized trial with 21 adults and 4 retrospective cohort studies
  - no significant difference in rate of glucose decline or insulin requirement comparing bicarbonate to no bicarbonate in
    - 3 randomized trials, 3 retrospective cohort studies , and 7 case reports/series with > 289 adults
    - 1 case series and 2 cohort studies with > 106 children
  - risk of cerebral edema
    - compared to controls, bicarbonate therapy associated with increased risk of cerebral edema (p < 0.05) in 2 case-control studies with 447 children
    - no significant difference in risk of cerebral edema in 1 case-control study with 63 children
  - compared to controls, bicarbonate therapy associated with nonsignificant increase in duration of hospitalization (adjusted p = 0.07) in 1 retrospective cohort study with 106 children
  - Reference – Ann Intensive Care 2011 Jul 6;1(1):23 full-text
- IV sodium bicarbonate does not appear to shorten time to resolution of acidemia or time to hospital discharge in patients with severe diabetic ketoacidosis of pH < 7 (level 2 [mid-level] evidence)
  - based on retrospective cohort study
  - 86 patients with diabetic ketoacidosis and initial pH < 7 who were treated with IV sodium bicarbonate (51%) vs. no bicarbonate (49%) were evaluated
  - median dose of IV sodium bicarbonate was 100 mEq
  - comparing IV sodium bicarbonate vs. no bicarbonate
    - time to return of pH to > 7.2 in 8 hours vs. 8 hours (not significant)
    - time to hospital discharge in 68 hours vs. 61 hours (not significant)
  - Reference – Ann Pharmacother 2013 Jul-Aug;47(7-8):970
Management of DKA in adults:
- Fluids and electrolytes:
  - IV fluids are first-line treatment.
    - Use 0.9% sodium chloride (normal saline) IV for initial fluid replacement.
    - Subsequent IV fluid choice is based on calculated corrected serum sodium.
      - If corrected serum sodium is high or normal, give 0.45% sodium chloride at 250-500 mL/hour (4-14 mL/kg/hour), depending on hydration state.
      - If corrected serum sodium is low, give 0.9% sodium chloride (normal saline) at 250-500 mL/hour (4-14 mL/kg/hour), depending on hydration state.
  - Bicarbonate therapy in patients with DKA is controversial due to a lack of benefit in randomized trials; however, no prospective randomized studies concerning the use of bicarbonate in DKA with pH values < 6.9 have been reported.
    - Bicarbonate therapy is recommended by the American Diabetes Association in patients with pH < 6.9.
    - Suggested dosing and administration: 100 mEq of sodium bicarbonate in 400 mL of sterile water with 20 mEq of potassium chloride at a rate of 200 mL per hour for 2 hours (repeat every 2 hours until pH ≥ 6.9).
  - Replace potassium after establishing adequate kidney function (urine output 50 mL/hour).
  - Consider treatment with magnesium if serum magnesium level is < 1.2 mg/dL (0.5 mmol/L) or if symptoms of hypomagnesemia develop (such as paresthesia, tremor, muscle spasm, seizures, or cardiac arrhythmia).
  - Phosphate supplementation is not routinely recommended but should be considered in patients with any of the following:
    - Serum phosphate level < 1 mg/dL (0.3 mmol/L)
    - Cardiac dysfunction, anemia, or respiratory distress, to avoid possible skeletal and cardiac weakness and respiratory depression
- Insulin administration:
  - Indications for IV insulin administration include:
    - Severe DKA
    - Anasarca
    - Hypotension
    - Associated severe critical illness
  - Do not initiate IV insulin until after beginning fluid resuscitation and hypokalemia correction. Insulin can begin promptly after starting fluid resuscitation.
  - Consider subcutaneous insulin instead of IV insulin (using rapid-acting insulin analog) in patients who are alert, tolerate oral fluids (no nausea or vomiting), have a pH > 7, and have bicarbonate ≥ 10 mEq/L (10 mmol/L).
  - Consider basal insulin in patients with newly diagnosed type 1 diabetes and in patients who were previously on basal insulin regimens prior to development of DKA.
- Identify and address the underlying cause of DKA.
Management of DKA in children and adolescents:
- Management of fluids and electrolytes:
  - Begin fluid resuscitation with 0.9% sodium chloride (normal saline) or Ringer lactate 10-20 mL/kg IV over 20-30 minutes. Repeat bolus as needed for shock.
  - After the initial bolus, estimate the degree of dehydration and aim to replace the fluid deficit over 24-48 hours.
    - Avoid infusion rates > 1.5-2 times the maintenance requirement because excessive fluid may increase the risk of cerebral edema.
    - Use 0.9% sodium chloride or Ringer lactate IV for at least 4-6 hours.
    - After 4-6 hours, use a solution with 0.45%-0.9% sodium chloride or Ringer lactate for subsequent fluid administration.
  - Add potassium to the IV fluid (if hyperkalemic, defer until urine output established).
  - Consider magnesium supplementation if magnesium < 1.2 mg/dL (0.5 mmol/L, 1 mEq/L).
  - Consider phosphate repletion in the form of potassium phosphate if serum phosphate level < 1 mg/dL (0.32 mmol/L).
- Management of insulin and dextrose:
  - Begin insulin therapy after 1 hour of fluid bolus.
    - Administer 0.05-0.1 units/kg/hour IV as continuous infusion until diabetic ketoacidosis (DKA) resolves as determined by pH > 7.3, bicarbonate > 18 mEq/L (18 mmol/L), and/or anion gap < 12 mEq/L (12 mmol/L).
    - A dose of 0.05 units/kg/hour may be used in milder cases, and in young children with significant sensitivity to insulin as long as metabolic acidosis is resolving.
  - Add dextrose to the IV fluids (begin with 5% dextrose) when plasma glucose decreases to 250-300 mg/dL (14-17 mmol/L).
  - Consider adding dextrose earlier if plasma glucose falls > 100 mg/dL/hour (5.5 mmol/L/hour) after initial period of volume expansion.
  - Transition to subcutaneous insulin when acidosis has resolved, blood glucose is < 200 mg/dL (11.1 mmol/L), and patient is tolerating oral fluids.

Alcoholic Ketoacidosis

Therapy involves thiamine supplementation and isotonic volume expansion.
- Thiamine supplementation should be administered prior to any dextrose.
  - The optimal dose of thiamine is unknown.
  - Doses of thiamine ranging from 200 to 500 mg via intramuscular injection 3 times daily have been used.
  - Reference – Emerg Med Clin North Am 2014 May;32(2):293
- Isotonic volume expansion is usually performed by using dextrose-containing fluids, such as 5% dextrose in normal saline (Emerg Med Clin North Am 2014 May;32(2):293).
Consider treatment of malnutrition and electrolyte deficiencies (Oxf Med Case Reports 2016 Mar;2016(3):31 full-text).
Bicarbonate and insulin should not be used (Emerg Med Clin North Am 2014 May;32(2):293).

Starvation Ketoacidosis

Management of starvation ketoacidosis may involve correction with IV fluids and supplementation.
- The administration of IV fluids, dextrose, thiamine, and folic acid has been reported to resolve starvation ketoacidosis in a 41-year-old pregnant woman in a case report (Case Rep Crit Care 2014;2014:906283 full-text).
- The administration of dextrose-containing IV fluids has been reported to quickly resolve starvation ketoacidosis in a 1-year-old boy in a case report (Am J Kidney Dis 2003 Nov;42(5):E16).

Management of Chronic Kidney Disease (CKD), Acute Kidney Injuries, and Uremia

For detailed management options of CKD, see the following topics for additional information:
- Chronic Kidney Disease (CKD) in Adults
- Chronic Kidney Disease (CKD) in Children
- Hemodialysis for End-stage Renal Disease
- Peritoneal Dialysis for End-Stage Kidney Disease
If severe, there are life-threatening acidosis or renal failure, consider dialysis. ⁽¹⁾
sodium bicarbonate may decrease mortality and risk of organ failure and reduce need for renal replacement therapy in patients admitted to intensive care unit (ICU) with severe acidemia and acute kidney injury
- based on prespecified subgroup analysis of BICAR-ICU trial without blinding
- 182 patients with acute kidney injury (Acute Kidney Injury Network [AKIN] stage 2 or 3) who were admitted to ICU within 48 hours of severe acidemia and randomized to receive sodium bicarbonate infusion or no sodium bicarbonate were evaluated
- severe acidemia defined as pH ≤ 7.2, partial pressure of arterial carbon dioxide (PaCO₂) ≤ 45 mm Hg, and sodium bicarbonate ≤ 20 mmol/L
- all patients had total Sequential Organ Failure Assessment score ≥ 4 points or arterial lactate ≥ 2 mmol/L
- primary outcome was composite of all-cause death at 28 days and ≥ 1 organ failure at 7 days
- comparing sodium bicarbonate vs. no sodium bicarbonate in subgroup of patients with acute kidney injury
  - primary outcome in 70% vs. 82% (p = 0.0462, NNT 9)
  - 28-day mortality 46% vs. 63% (p = 0.0166, NNT 6)
  - ≥ 1 organ failure at day 7 in 66% vs. 82% (p = 0.0142, NNT 7)
  - renal replacement therapy during ICU stay 51% vs. 73% (p = 0.002, NNT 5)
  - median renal replacement therapy-free days during ICU stay 10 days vs. 1 day (p = 0.004)
  - dependence on dialysis at ICU discharge in 20% vs. 48% (p = 0.046, NNT 4)
  - median vasopressor-free days 18 days vs. 1 day (p = 0.022)
- Reference – Lancet 2018 Jul 7;392(10141):31, editorial can be found in Lancet 2018 Jul 7;392(10141):3

Management of Methanol or Ethylene Glycol Poisoning

American Academy of Clinical Toxicology suggests correction of metabolic acidosis with bicarbonate if pH is < 7.3 (J Toxicol Clin Toxicol 1999;37(5):537).
Administration of antidote to block metabolism to toxic metabolites:
- IV ethanol:
  - Alcohol dehydrogenase has higher affinity for ethanol than for ethylene glycol (J Toxicol Clin Toxicol 1999;37(5):537).
  - Serum ethanol concentration of 100 mg/dL (22 mmol/L liter) should provide competitive inhibition of the enzyme, but the amount of ethanol to prevent formation of toxic metabolites depends on the amount of methanol present (N Engl J Med 2018 Jan 18;378(3):270, J Toxicol Clin Toxicol 2002;40(4):415).
- Fomepizole (4-methylpyrazole):
  - Fomepizole is a strong inhibitor of alcohol dehydrogenase. It has an affinity for alcohol dehydrogenase which is 8,000 times that of ethanol.
  - FDA has approved for fomepizole for methanol and ethylene glycol poisoning and it is the preferred antidote for methanol poisoning.
  - Reference – N Engl J Med 2018 Jan 18;378(3):270
- CLINICIANS’ PRACTICE POINT: There are no comparative clinical efficacy data to support the superiority of fomepizole over ethanol in patients with methanol or ethylene glycol poisoning, but there are significant disadvantages associated with ethanol use (more frequent monitoring, higher risk of hepatotoxicity and hypoglycemia) and most clinical guidelines support the use of fomepizole as the primary treatment. However, fomepizole is expensive and may not be available in some countries whereas ethanol is readily available.
Hemodialysis and hemodiafiltration:
- Hemodialysis and hemodiafiltration have been used successfully to treat ethylene glycol and methanol toxicity (Clin Toxicol (Phila) 2009 Jul;47(6):525).
- Ethanol and fomepizole are typically used immediately upon patient presentation, but hemodialysis should be considered if the patient presents with renal failure, or severe metabolic acidosis, or is unresponsive to antidotes (FDA DailyMed 2009 Jul 27).
- Fomepizole is often continued during hemodialysis (Clin Toxicol (Phila) 2009 Jul;47(6):525).
insufficient evidence to compare fomepizole and ethanol as antidotes for ethylene glycol or methanol poisoning
- based on systematic review of case series and case reports
- systematic review of 145 case series and case reports evaluating fomepizole or ethanol use in treatment of methanol or ethylene glycol poisoning in 897 patients
- no randomized trials or head-to-head studies identified in literature
- 293 had ethylene glycol poisoning, 602 had methanol poisoning, and 2 were poisoned with both
- treatment across studies
  - 80.3% treated with ethanol
  - 16.5% treated with fomepizole
  - 3.7% treated with both
- mortality
  - 18.1% for ethylene glycol poisoning treated with ethanol
  - 4.1% for ethylene glycol poisoning treated with fomepizole
  - 21.8% for methanol poisoning treated with ethanol
  - 17.1% for methanol poisoning treated with fomepizole
- few adverse events were reported with either antidote
- Reference – Emerg Med Int 2013;2013:638057 full-text
See also:
- Ethylene Glycol Toxicity – Emergency Management
- Methanol Poisoning

Management of Salicylate Poisoning

Salicylate poisoning does not have a specific antidote and management is based on supportive care and elimination of salicylate.
Consider early consultation with toxicologist and nephrologist for hemodialysis in patients with signs of severe toxicity.
Patients with known ingestion of < 125 mg/kg and no symptoms can be discharged with follow-up.
Perform urinary alkalinization for all symptomatic patients with acute or chronic salicylate poisoning who have intact renal function.
- Urinary alkalinization can be performed alone in combination with hemodialysis, but should not be used as an alternative to hemodialysis if hemodialysis is also indicated.
- Example regimen includes dextrose (D5W) 1 L combined with 3 ampules (50 mL each) of 7.5% or 8.4 % sodium bicarbonate (a total of 132-150 mEq).
- Infusion rate should induce a urine output of 2-3 mL/kg/hour (infusion rate of 1.5 to 2 times the maintenance dose for IV fluids is usually sufficient).
- Potassium chloride 30-40 mEq can be added to treat hypokalemia in patients without acute kidney injury or oliguria.
Perform aggressive IV fluid resuscitation in patients with hypotension and decreased extracellular fluid volume.
- Goal of fluid therapy is to establish euvolemia without forcing diuresis, which is associated with increased risk of pulmonary edema.
- Lactated Ringer’s solution or isotonic saline is administered at 10-20 mL/kg/hour IV for the first 2 hours with adjustment to urine output of 1-1.5 mL/kg/hour.
- Monitor patients frequently with fluid adjustments as needed if signs of pulmonary or cerebral edema develop.
Endotracheal intubation is associated with a short period of apnea, which can cause a rapid drop in pH and increased accumulation of salicylate in the central nervous system; therefore, routine use of endotracheal intubation is not suggested for patients with salicylate poisoning due to potential for worsening toxicity.
- Reserve endotracheal intubation for patients with salicylate poisoning who present with any of the following:
  - hypoventilation as indicated by clinical evaluation and blood gas analysis
  - deteriorating mental status and unprotected airway
  - acute lung injury
- Endotracheal intubation may be performed if sedation is required during dialysis IV catheter access placement.
- Procedural considerations to reduce risk for worsening toxicity by improving plasma pH in patients having endotracheal intubation:
  - Sodium bicarbonate 2 mmol/kg IV can be given during induction of intubation to maintain blood pH 7.45-7.5 while sedated.
  - While intubated, increased minute volume and low pCO₂should be maintained.
Activated charcoal can inhibit gastrointestinal absorption of salicylate, and may be particularly effective in patients who present soon after salicylate ingestion and in those with rising serum salicylate levels or other signs of incomplete absorption.
- Activated charcoal is most effective if given within 2 hours of ingestion, but may also be given to alert and cooperative patients who present ≥ 2 hours after ingestion due to salicylate’s association with delayed gastric emptying and prolonged retention in the gastrointestinal tract .
- Suggested dose for activated charcoal:
  - adults 1-2 g/kg to max dose 100 g
  - children 1 g/kg to max dose 50 g
- Sorbitol is often given with first dose, but repeat use is not suggested due to potential exacerbation of electrolyte disturbances and abdominal cramping.
- Multiple doses of activated charcoal can be given every 4 hours as needed until charcoal appears in the stool and clinical symptoms resolve.
Whole bowel irrigation is not used routinely in the management of salicylate poisoning, but can be used in select patients who do not respond to activated charcoal alone. Indications include suspicion or confirmation of bezoar formation, or ingestion of enteric-coated or sustained-release formulations.
Hemodialysis is an effective way to clear free salicylate in serum and correct acidosis, volume disturbances, and electrolyte abnormalities.
- Extracorporeal dialysis is recommended in patients with severe salicylate poisoning, including patients with any of the following:
  - Salicylate level > 100 mg/dL (7.2 mmol/L)
  - Salicylate level > 90 mg/dL (6.5 mmol/L) in patient with impaired kidney function
  - Altered mental status, regardless of salicylate level
  - New hypoxemia requiring supplemental oxygen, regardless of salicylate level
- Consider extracorporeal dialysis in patients without a response to standard therapy (such as urinary alkalinization and fluid resuscitation) if any of the following are present:
  - Salicylate level > 90 mg/dL (6.5 mmol/L)
  - Salicylate level > 80 mg/dL (5.8 mmol/L) in patient with impaired kidney function
  - pH ≤ 7.2
- In patients with chronic salicylate poisoning, the threshold for initiating hemodialysis may be lower (such as plasma salicylate level > 50 mg/dL [3.6 mmol/L]).
- Urinary alkalinization is not necessary during hemodialysis but may be initiated during preparation for hemodialysis and restarted after hemodialysis if patient is symptomatic or salicylate levels are pending.

Management of Propylene Glycol Poisoning

Avoid or discontinue medications that use propylene glycol as the delivery vehicle.⁽¹⁾
Consider hemodialysis if the patient is worsening clinically.⁽¹⁾
There is no consensus regarding the use of fomepizole (N Engl J Med 2018 Jan 18;378(3):270).

Management of Cyanide Poisoning

Prehospital emergency response for first responders:
- Protect first responders from exposure with both respiratory and skin protection.
  - Consider using positive-pressure self-contained breathing apparatus (SCBA) for respiratory protection in situations with unsafe levels of hydrogen cyanide.
  - Chemical-protective clothing can shield against both hydrogen cyanide vapor and liquid.
- If proper protective equipment is unavailable or responders are untrained, contact local or regional HAZMAT team or other response organization to provide assistance.
- Remove person from source of exposure.
- Perform cardiopulmonary support and/or resuscitation. Do not attempt resuscitation without barrier protection to prevent cyanide exposure in rescue worker.
- Perform initial decontamination (preferably in prehospital setting).
  - Patients only exposed to gas who have no eye irritation do not need decontamination.
  - Remove contaminated clothing and double bag along with any other contaminated belongings.
  - Irrigate exposed skin and hair with water for at least 20 minutes.
  - Irrigate exposed or irritated eyes for 5 minutes with water or saline.
Administer cyanide antidotes.
- Hydroxocobalamin (Cyanokit) is the preferred antidote in North America and Europe based on practicality and safety.
  - Hydroxocobalamin is appropriate in patients with suspected concurrent carbon monoxide exposure.
  - Dosing in adults is 5 g IV over 15 minutes and dosing in children is 70 mg/kg IV.
  - Dose may be repeated over 15 minutes to 2 hours if severe poisoning.
- If hydroxocobalamin is not available, other antidotes include cyanide antidote kit, dicobalt edetate (Kelocyanor), and 4-dimethylaminophenol (4-DMAP).
Provide supportive care.
- Give 100% oxygen for all patients with cyanide poisoning.
- In patients with recent cyanide ingestion, consider activated charcoal 1 g/kg if patient is alert, asymptomatic, and has gag reflex.
  - Benefit of activated charcoal will vary by time since ingestion, and may only be useful soon after ingestion (such as within 1 hour).
  - Do not induce emesis.
- Consider sodium bicarbonate 1 mEq/kg IV in patients with shock, seizures, or arrhythmias.

Management of Carbon Monoxide Poisoning

Carbon monoxide (CO) poisoning may cause lactic acidosis due to tissue hypoxia.⁽¹⁾
Treat all patients with suspected CO exposure with oxygen inhalation (after drawing blood for carboxyhemoglobin levels).
- Use nonrebreather mask to deliver 100% oxygen until patient is asymptomatic and carboxyhemoglobin levels are ≤ 3%-4% for nonsmokers or ≤ 10% for smokers.
- Consider hyperbaric oxygen for patients with CO poisoning.
  - The definitive criteria or evidence for hyperbaric oxygen therapy are lacking.
  - Experts often suggest hyperbaric oxygen therapy in the following clinical scenarios:
    - Loss of consciousness
    - Persistent abnormal neurologic findings, such as mental status alteration
    - Pregnancy
    - High carboxyhemoglobin levels (> 25%)
    - Prolonged exposure (> 24 hours)
    - Cardiovascular dysfunction
    - Methylene chloride as the source of CO poisoning

Complications

Complications of severe metabolic acidosis:
- Cardiovascular complications include the following:
  - Weakened cardiac output
  - Conduction defects
  - Arterial vasodilation
  - Venous vasoconstriction
- Neuromuscular complications include the following:
  - Respiratory depression
  - Decreased sensorium
- Metabolic complications include the following:
  - Protein degradation
  - Bone demineralization
  - Parathyroid hormone stimulation
  - Catecholamine stimulation
  - Aldosterone stimulation
  - Insulin resistance
  - Formation of free radicals
- Gastrointestinal complications include the following:
  - Nausea and vomiting
  - Gut barrier dysfunction
- Complications in electrolyte abnormalities include the following:
  - Hyperkalemia
  - Hypercalcemia
  - Hyperuricemia
- Decreased oxygen-hemoglobin binding is also a complication.
- Reference – Nephrol Dial Transplant 2015 Jul;30(7):1104

Prognosis

anion gap metabolic acidosis associated with higher mortality than hyperchloremic metabolic acidosis in critically ill patients
- based on retrospective cohort study
- data extracted from 851 patients in intensive care unit (ICU) at the University of Pittsburgh between January 1, 2001 and June 30, 2002
- inclusion criteria
  - patients suspected of having lactic acidosis (clinical/medical history grounds)
  - arterial blood gas and simultaneous serum electrolytes/lactate drawn
- 548 (64%) had metabolic acidosis (standard base excess < 2 mEq/L)
  - mortality 45% in patients with metabolic acidosis vs. 26% in patients without (p < 0.001)
  - mortality much higher when metabolic acidosis associated with lactate/other strong anion vs. hyperchloremic acidosis
- Reference – Crit Care 2006 Feb;10(1):R22 full-text
in patients with severe metabolic or mixed acidosis, Simplified Acute Physiology Score > 65 at admission, lactic acidosis at admission, and < 0.1 change in plasma pH after 24 hours in ICU each associated with increased mortality
- based on prospective cohort study
- 200 patients with severe acidemia (pH < 7.2) being admitted to the ICU were followed
- 155 patients with severe metabolic or mixed acidosis were evaluated
- common causes of acidosis included shock, cardiac arrest, trauma, and acute renal failure
- in addition to laboratory analysis, patients were evaluated by Simplified Acute Physiology Score (SAPS) II score
  - score based on 17 variables, including 12 physiology variables, age, type of admission, and 3 underlying disease variables
  - score range 0 to > 60 with higher score indicating increased likelihood of mortality
  - calculator can be found at SAPS II score
- 57% of patients died during the ICU stay
- factors associated with increased mortality in patients with severe metabolic or mixed acidosis included
  - SAPS II score > 65 at admission (odds ratio [OR] 7.53, 95% CI 2.3-24.7)
  - lactic acidosis at admission (OR 1.2, 95% CI 1.04-1.39)
  - < 0.1 change in plasma pH after 24 hours in the ICU (OR 0.18, 95% CI 0.05-0.56)
- Reference – Crit Care 2011;15(5):R238 full-text
Survival in patients with metformin-associated lactic acidosis:
- 85.8% overall survival in patients with metformin-induced lactic acidosis or metformin-associated lactic acidosis
  - based on systematic review of observational studies
  - systematic review of 83 observational studies evaluating patients with metformin poisoning who were treated with continuous renal replacement therapy
  - metformin-associated lactic acidosis was defined as acute kidney injury in patients with chronic metformin use and metformin-induced lactic acidosis was defined as lactic acidosis that occurs after acute ingestion of metformin; these types of are sometimes difficult to distinguish in the literature
  - studies including polysubstance overdoses and some rare extracorporeal clearance modalities were excluded
  - pooled overall survival was 85.8%; pooled overall survival was 75% for metformin-induced lactic acidosis and 87.4% for metformin-associated lactic acidosis (no p value reported)
  - Reference – Clin Toxicol (Phila) 2022 Nov;60(11):1266
- 36.2% overall mortality (or about 64% survival) in patients with metformin-associated lactic acidosis; both initial pH level and initial lactate level appear to be poor predictors of mortality (level 2 [mid-level] evidence)
  - based on systematic review of observational studies
  - systematic review of 44 observational studies evaluating metformin-associated lactic acidosis in 170 patients (median age 68 years)
  - median initial pH was 7.02 and median initial lactate was 14.45 mmol/L
  - overall mortality was 36.2% (95% CI 29.6%-43.94%) (equivalent to about 64% overall survival)
  - for prediction of mortality in setting of metformin-associated lactic acidosis, both initial pH (c-statistic 0.43) and initial lactate (c-statistic 0.593) each had poor performance
  - Reference – J Med Toxicol 2020 Apr;16(2):222 full-text
EVIDENCE SYNOPSIS: Elevated serum lactate and base deficit at admission are associated with increased mortality in trauma patients.
- serum lactate > 2 mmol/L and base deficit > -2 mmol/L at admission each associated with increased mortality in ethanol-negative trauma patients
  - based on retrospective cohort study
  - 2,482 trauma patients (mean age 44.9 years) had serum ethanol, lactate, and base deficit measured upon admission to trauma center
    - 45.4% had elevated lactate (> 2 mmol/L)
    - 44% had elevated base deficit (> -2 mmol/L)
  - all patients had negative urine drug screen
  - in ethanol-negative patients, strongest predictors for mortality included
    - elevated lactate at admission (odds ratio [OR] 2.6, 95% CI 1.6-4.2)
    - elevated base deficit at admission (OR 1.9, 95% CI 1.2-3.1)
  - in ethanol-positive patients, Injury Severity Score (ISS) was only predictor of mortality (OR 1.1, 95% CI 1.07-1.14)
  - Reference – Am J Emerg Med 2015 May;33(5):607 full-text
- elevated lactate and base deficit at admission each associated with increased mortality in elderly blunt trauma patients
  - based on retrospective cohort study
  - 1,747 patients ≥ 55 years old with blunt trauma and admission systolic blood pressure (SBP) ≥ 90 mm Hg at baseline were evaluated
    - 364 had arterial lactate drawn at admission
    - 324 had base deficit drawn at admission
    - 264 of the above patients had both arterial lactate and base deficit drawn at admission
  - lactate categories defined as
    - < 2.5 mmol/L (normal)
    - ≥ 2.5 mmol/L
    - ≥ 4 mmol/L
  - base deficit (BD) categories defined as
    - > -2 mEq/L (normal)
    - ≤ -2 mEq/L
    - ≤ -4 mEq/L
    - ≤ -6 mEq/L
  - factors associated with increased mortality
    - base lactate ≥ 2.5 mmol/L
      - odds ratio (OR) 3.7 (95% CI 1.6-8.2) for patients with SBP ≥ 90 mm Hg
      - OR 4.3 (95% CI 1.8-10.5) for patients with SBP ≥ 110 mm Hg
    - base deficit ≤ -4 mEq/L
      - OR 5.2 (95% CI 2.5-11.2) for patients with SBP ≥ 90 mm Hg
      - OR 4.1 (95% CI 1.8-9.3) for patients with SBP ≥ 110 mm Hg
  - Reference – Am Surg 2011 Oct;77(10):1337
- higher BD on admission associated with increased mortality and risk for complications in trauma patients admitted to intensive care unit
  - based on retrospective cohort study
  - 2,954 trauma patients (median age 28 years) with arterial blood gas values taken within 1 hour of admission to trauma service between July 1992 and August 1995 evaluated
  - BD categories defined as
    - 2 to -2 mEq/L (normal)
    - -3 to -5 mEq/L (mild)
    - -6 to -9 mEq/L (moderate)
    - ≤ -10 mEq/L (severe)
  - survival by BD category
    - 93% among patients with normal BD
    - 89% among patients with mild BD
    - 77% among patients with moderate BD
    - 51% among patients with severe BD
  - no significant difference between survival predicted by stratified BD level and observed survival
  - larger BD associated with
    - increased length of both ICU (p < 0.015) and hospital stay (p < 0.05)
    - incidence of acute respiratory distress syndrome (ARDS) (p < 0.01)
    - renal failure (p = 0.015)
    - coagulopathy (p < 0.001)
    - multiorgan system failure (p = 0.002)
  - Reference – J Trauma 1996 Nov;41(5):769, commentary can be found in J Trauma 1997 Mar;42(3):571

Prevention

For prevention of diabetic ketoacidosis, see also the following topics:
- Diabetic Ketoacidosis (DKA) in Children and Adolescents
- Diabetic Ketoacidosis (DKA) in Adults
- Lactic Acidosis

Guidelines and Resources

Guidelines

International Guidelines

Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guideline on evaluation and management of chronic kidney disease can be found in Kidney Int 2024 Apr;105(4S):S117.
Surviving Sepsis Campaign (SSC) international guideline on management of sepsis and septic shock can be found in Intensive Care Med 2017 Mar;43(3):304.
International Society for Pediatric and Adolescent Diabetes (ISPAD) guideline on diabetic ketoacidosis and hyperglycemic hyperosmolar state can be found in Pediatr Diabetes 2018 Oct;19 Suppl 27:155.

Review Articles

Review of integration of acid-base and electrolyte disorders can be found in N Engl J Med 2014 Nov 6;371(19):1821, commentary can be found in N Engl J Med 2015 Jan 22;372(4):391.
Review of metabolic acidosis in emergency department can be found in Emerg Med Clin North Am 2014 May;32(2):403.
Review of approach to patients with high anion gap metabolic acidosis can be found in Am J Kidney Dis 2021 Oct;78(4):590.
Review of diagnostic approach to an adult patient with metabolic acidosis can be found in Br J Hosp Med (Lond) 2022 Aug 2;83(8):1.
Review of treatment of acute metabolic acidosis can be found in Nat Rev Nephrol 2012 Oct;8(10):589.
Reviews of lactic acidosis can be found in:
- Clin Chem Lab Med 2017 Mar 1;55(3):322
- N Engl J Med 2014 Dec 11;371(24):2309, commentary can be found in N Engl J Med 2015 Mar 12;372(11):1078
Reviews of diabetic ketoacidosis can be found in:
- N Engl J Med 2015 Feb 5;372(6):546
- BMJ 2015 Oct 28;351:h5660, correction can be found in BMJ 2015 Nov 2;351:h5866
Review of starvation ketoacidosis in pregnancy can be found in Eur J Obstet Gynecol Reprod Biol 2013 Mar;167(1):1.
Reviews of toxicologic acid-base disorders can be found in:
- N Engl J Med 2018 Jan 18;378(3):270, correction can be found in N Engl J Med 2019 Jan 10;380(2):202
- Emerg Med Clin North Am 2014 Feb;32(1):149
Reviews of metabolic acidosis in critically ill patients:
- Part 1. Classification and pathophysiology can be found in Anaesthesia 2008 Mar;63(3):294.
- Part 2. Causes and treatment can be found in Anaesthesia 2008 Apr;63(4):396.
Case reports:
- Case report of markedly negative anion gap in woman aged 47 years due to salicylate poisoning (suicide attempt) with a triple acid-base disorder, anion gap metabolic acidosis, metabolic alkalosis, and respiratory alkalosis can be found in Am J Kidney Dis 2016 Jan;67(1):143.
- Case report of alcoholic ketoacidosis in 64-year-old patient with nonspecific symptoms and significant acidemia can be found in Oxf Med Case Reports 2016 Mar;2016(3):31 full-text .
- Case report of encephalopathy and high anion gap metabolic acidosis in patient with buried bumper syndrome can be found in Lancet 2015 Feb 21;385(9969):744.
- Case report of elevated anion gap and elevated osmolal gap in patient with severe metabolic acidosis can be found in Ann Am Thorac Soc 2014 May;11(4):671.
- Case report of elevated anion gap in evaluation of 80-year-old patient with metabolic acidosis in setting of altered mental status and worsening diarrhea can be found in Am J Kidney Dis 2014 Oct;64(4):653.
- Case report of a rare cause of high anion gap metabolic acidosis can be found in Intern Med J 2013 Jan;43(1):100.
- Case report of normal osmolal gap and high anion gap metabolic acidosis in patient with acute kidney injury with oxalate deposition can be found in J Nephropathol 2013 Apr;2(2):139 full-text.

MEDLINE Search

To search MEDLINE for (Anion Gap Metabolic Acidosis) with targeted search (Clinical Queries), click therapy, diagnosis, or prognosis.

Patient Information

Information on metabolic acidosis from National Kidney Foundation

References

General References Used

The references listed below are used in this DynaMed topic primarily to support background information and for guidance where evidence summaries are not felt to be necessary. Most references are incorporated within the text along with the evidence summaries.

Rice M, Ismail B, Pillow MT. Approach to metabolic acidosis in the emergency department. Emerg Med Clin North Am. 2014 May;32(2):403-20.
Berend K, de Vries AP, Gans RO. Physiological approach to assessment of acid-base disturbances. N Engl J Med. 2014 Oct 9;371(15):1434-45, correction can be found in N Engl J Med. 2014 Nov 13;371(20):1948, commentary can be found in N Engl J Med 2015 Jan 8;372(2):195.
Berend K. Review of the Diagnostic Evaluation of Normal Anion Gap Metabolic Acidosis. Kidney Dis (Basel). 2017 Dec;3(4):149-159 full-text.

Anion Gap Metabolic Acidosis

Description

Definitions

Incidence and Prevalence

Pathogenesis

Differential Diagnosis

Causes of Elevated Anion Gap

Overview

Ketoacidosis

Lactic Acidosis

Other Causes

Acidosis Without Elevated Anion Gap

Mimics

Serum Anion Gap (AG) in Conditions Other Than Metabolic Acidosis

History and Physical

Clinical Presentation

History

Medication History

Past Medical History (PMH)

Social History (SH)

Diagnostic Testing

Testing Overview

Initial Testing

Blood Gas Analysis

Blood Tests

Overview

Serum Anion Gap

Delta-delta Gradient and/or Ratio

Serum Lactate and Ketones

Serum Osmolal Gap (OG)

Urine Studies

Management

Management Overview

Management of Metabolic Acidosis

Sodium Bicarbonate for Metabolic Acidosis

Veverimer For Metabolic Acidosis

Management of Underlying Causes of Metabolic Acidosis

Management of Lactic Acidosis

Management of Ketoacidosis

Diabetic Ketoacidosis (DKA)

Alcoholic Ketoacidosis

Starvation Ketoacidosis

Management of Chronic Kidney Disease (CKD), Acute Kidney Injuries, and Uremia

Management of Methanol or Ethylene Glycol Poisoning

Management of Salicylate Poisoning

Management of Propylene Glycol Poisoning

Management of Cyanide Poisoning

Management of Carbon Monoxide Poisoning

Complications

Prognosis

Prevention

Guidelines and Resources

Guidelines

International Guidelines

Review Articles

MEDLINE Search

Patient Information

References

General References Used

Share this:

Like this:

Related Posts

Related Posts:

Recent Post

features

Categories

Weekly Newsletter