After completing this chapter, the reader should be able to

  • Discuss the chemistry associated with acid-base balance

  • Describe the components of an arterial blood gas analysis and their contribution to acid-base physiology

  • Compare and contrast the physiologic approach to the Stewart approach for acid-base balance

  • Describe the methods used by the human body to maintain acid-base balance

  • Evaluate a patient’s acid-base status and identify common causes given the clinical presentation, laboratory, and arterial blood gas data

  • List the four simple acid-base disorders, their accompanying laboratory test results, and possible causes

  • Describe how the anion gap can be used to determine the primary cause of metabolic acidosis

Acid–base homeostasis is a fundamental component for the maintenance of normal metabolic function. Acid–base disorders, however, are extremely common in the intensive care unit, and rapid, careful assessment is required to prevent unwanted morbidity and mortality. This chapter provides a review of acid–base homeostasis, laboratory tests used to assess acid–base status, and a step-wise approach to classify acid–base disorders and their potential causes.


An acid is a substance that can donate a proton (eg, HCl → H+ + Cl) whereas a base is a substance that can accept a proton (eg, H+ + NH3 → NH4+). Every acid has a corresponding base. Every base has a corresponding acid. Some common acid–base pairs are carbonic acid/bicarbonate, ammonium/ammonia, monobasic/dibasic phosphate, and lactic acid/lactate.

The terms acidemia and alkalemia are used to describe an abnormal blood pH. Specifically, acidemia denotes a low blood pH whereas alkalemia denotes a high blood pH. The terms acidosis and alkalosis, on the other hand, refer to the process by which either acid or alkali accumulates. It is therefore possible to have an acidosis but not an acidemia. For this to occur (ie, acidosis without acidemia), a corresponding alkalosis must also be present.

The acidity of a body fluid is determined by the concentration of hydrogen ion (H+). Normal H+ concentration is approximately 40 nanoequivalents/liter. Because this is expressed in such small amounts (a nanoequivalent is one one-millionth of a milliequivalent), acid–base status is measured in pH units using a logarithmic scale. Normal blood pH is 7.4 with a range of 7.35 to 7.45. The range of pH values considered compatible with life is 6.8 to 7.8, which corresponds to a hydrogen ion concentration of only 16 to 160 nEq/L.1 In general, the body tolerates acidemia much better than alkalemia. This is due to the fact that as pH decreases, a larger change in H+ is required for a given change in pH.2 In alkalemic states, small changes in H+ can markedly affect pH. Furthermore, with alkalemia, the oxyhemoglobin dissociation curve shifts to the left and hemoglobin is less willing to release oxygen to the tissues (Figure 13-1).3,4

FIGURE 13-1.
FIGURE 13-1.

The oxyhemoglobin dissociation curve.


Assessment of acid–base status is primarily determined using an arterial blood gas (ABG) measurement. ABG evaluation is typically performed in critically ill patients in whom complex disorders and treatment modalities can influence acid–base status (eg, mechanical ventilation). Outside the intensive care unit, many acid–base disorders are predicable (eg, respiratory acidosis in a patient with chronic obstructive pulmonary disease) and rarely require interpretation via an ABG. Arterial blood reflects how well the blood is being oxygenated by the lungs whereas venous blood reflects oxygen consumption by the tissues. It is important that arterial blood is used for these assessments because substantial differences may exist between the two, particularly in the setting of critical illness.5 Note that the “normal ranges” listed next are approximate and slight variability may be noted across different references or local laboratory standards.

Arterial pH

Normal range: 7.35 to 7.45

The pH of arterial blood is the first value to consider when using an ABG to assess acid–base status. The pH is inversely related to hydrogen concentration. Generally speaking, pH values <7.35 represent acidemia and pH values >7.45 represent alkalemia, with 7.4 being the threshold for categorization during ABG assessment.

It is difficult to identify the pH value that will dictate the urgency whereby treatment must be initiated. Consequences of abnormal pH include arterial vasodilation, diminished myocardial contractility, impaired hepatic and renal perfusion, decreased oxygen-hemoglobin binding, and coma for acidemia; cerebral vasoconstriction, reduced contractility, increased oxygen-hemoglobin binding, decreased oxygen delivery, and coma are encountered with alkalemia. These deleterious effects become more prominent when pH is <7.2 or >7.55.6

Arterial Partial Pressure of Carbon Dioxide

Normal range: 35 to 45 mm Hg

Evaluation of arterial partial pressure of carbon dioxide (PaCO2) provides information about the adequacy of lung function in excreting carbon dioxide. The amount of carbon dioxide dissolved in the blood is directly proportional to the concentration of carbonic acid (PaCO2 × 0.03 = H2CO3). Elevations in PaCO2 therefore contribute to acidosis. Changes to ventilatory status that alter carbon dioxide concentrations affect carbonic acid concentrations. Specifically, hypoventilation leads to a higher PaCO2 whereas hyperventilation results in a lower PaCO2. Regulation of ventilation is a major mechanism for respiratory compensation in the setting of primary metabolic disorders.

Arterial Partial Pressure of Oxygen

Normal range: 80 to 100 mm Hg

Evaluation of the arterial partial pressure of oxygen (PaO2) provides information about the level of oxygenation of arterial blood. The PaO2 is important because it reflects not only the functional capabilities of the lungs but also the rate at which oxygen can enter the tissues. Although there is no set cutoff for defining hypoxemia because it is typically relative to metabolic requirements, most would define clinically significant hypoxia at <60 mm Hg. Factors that influence PaO2 are the amount of ventilation, the fraction of inspired oxygen (FiO2), the functional capacities of the lung, and the oxyhemoglobin dissociation curve.

The oxyhemoglobin dissociation curve describes the relationship between PaO2 and oxygen saturation (Figure 13-1). Oxygen saturation is the percentage of hemoglobin binding sites in the bloodstream occupied by oxygen. During states of acidemia, this curve shifts to the right and hemoglobin has a decreased affinity for oxygen (resulting in increased unloading of oxygen). PaO2 must therefore be higher to maintain a particular level of oxygen saturation. During alkalemia, on the other hand, this curve shifts to the left and hemoglobin’s affinity for oxygen is increased, resulting in decreased oxygen delivery to tissues. Other factors that can influence the oxyhemoglobin dissociation curve are temperature and the amount of 2,3-diphosphoglycerate in red blood cells.

Although PaO2 and oxygen saturation are both measurements of oxygenation, it is important to not confuse one with the other. For example, a PaO2 of 80 mm Hg would typically not be considered abnormal because in most patients this is reflective of an oxygen saturation >93%. An oxygen saturation of 80%, on the other hand, would be considered critical and require immediate intervention. Oxygen saturation is expressed as a percentage and cannot exceed 100%. Conversely, PaO2 is expressed in millimeters of mercury, and values that exceed 100 can exist (especially when supplemental oxygen is administered). The ratio of PaO2 to FiO2, which is referred to as PF ratio, is frequently used to evaluate respiratory status in patients receiving supplemental oxygen. A PF ratio in a healthy individual would be expected to be approximately 500 and values less than 300, 200, and 100 are encountered in mild, moderate, and severe acute respiratory distress syndrome, respectively.7 Although PaO2 assessment is crucial for determining pulmonary status, it does not directly impact acid–base balance.

Arterial Bicarbonate

Normal range: 22 to 26 mEq/L

The concentration of HCO3 reported from ABG is not a direct measurement but calculated using pH and PaCO2 via the Henderson-Hasselbalch equation. It is important to compare this value (ie, bicarbonate reading from the ABG) with the total CO2 content (commonly referred to as serum bicarbonate) on a basic metabolic panel from venous sampling. Under normal circumstances, bicarbonate from the ABG is approximately 1.5 to 3 mEq/L less than the total CO2 content from a plasma electrolyte panel (higher end of this range for venous samples).8 Results should be interpreted with caution if this correlation does not exist.


Venous Total Carbon Dioxide Content (Serum Bicarbonate)

Normal range: 22 to 30 mEq/L

The total carbon dioxide content refers to the total of all carbon dioxide present in the blood. This consists of bicarbonate, dissolved CO2, carbonic acid, and carbamino compounds.8 Because roughly 95% of this is made up of bicarbonate, the term serum bicarbonate is often used interchangeably with total carbon dioxide content. Although the name total carbon dioxide content implies that it is a measure of acid (similar to PaCO2), it is important to recognize that this test represents bicarbonate. Increases in total carbon dioxide content contribute to alkalosis. Clinicians should refer to this as serum bicarbonate and not carbon dioxide to avoid confusion with pCO2 obtained on an ABG test.

Anion Gap

Normal range: 3 to 16 mEq/L

The anion gap (AG) is a calculated value that is an estimate of the relative abundance of unmeasured anions. It is commonly used to determine the possible causes of metabolic acidosis. Anion gap is based on the principles of electrochemical balance; that is, the concentration of negatively charged anions must equal the concentration of positively charged cations. Anion gap is calculated using the following formula:
AG = Na+(Cl+HCO3)

Whereas the normal value for AG can vary, values that exceed 16 mEq/L are generally indicative of anion accumulation (eg, lactate, pyruvate, acetoacetate). This is often due to lactic acidosis, ketoacidosis, toxic ingestions, or end-stage renal failure.

The AG is largely influenced by plasma albumin; therefore, an adjustment is required when hypoalbuminemia exists. For every decrement of 1 g/dL in serum albumin concentration, the calculated AG should be increased by 2.5.9
Corrected AG=AG+2.5(normal albumin per local laboratory standardmeasured albumin)

Anion gap interpretation may also be affected by serum phosphate, magnesium, calcium, and even some β-lactam antibiotics.10 The influence of these confounding variables, coupled with the heterogeneity of critical illness, has led to questions regarding its diagnostic value.11-13 Nevertheless, clinicians should consider the AG one method to assist with potential causes of metabolic acidosis and not the sole factor for decision-making at the bedside.

Arterial Oxygen Saturation

Normal range: 93% to 100%

Arterial oxygen saturation is a measure of the fraction of hemoglobin molecules that are saturated with oxygen. Its relationship with PaO2 is described using the oxyhemoglobin dissociation curve. Under normal circumstances (eg, pH, temperature, etc.), a PaO2 of approximately 60 mm Hg corresponds to an arterial oxygen saturation (SaO2) of 90%. Arterial oxygen saturation is the major determinant of arterial oxygen content (CaO2) (normal range is roughly 17 to 20 mL/dL; Equation 3) and oxygen delivery (DO2) (normal range is roughly 950 to 1,150 mL/min or 550 to 650 mL/min/m2 when indexed to body surface area; Equation 4).
CaO2(mL/dL)=1.34×Hgb (g/dL)×SaO2+[PaO2(mmHg)×0.003]
DO2(mL/min)=CaO2 (mL/dL)×CO (L/min)×10

where CO is cardiac output.

Serum Lactate

Normal range: 0.6 to 2 mmol/L

Lactate is a byproduct of anaerobic metabolism and often used as a surrogate for inadequate tissue perfusion. In fact, the Third International Consensus Definition for Sepsis and Septic Shock (Sepsis-3) lists lactate levels >2 mmol/L as part of the diagnostic criteria for septic shock (coupled with vasopressors to maintain MAP ≥65 mm Hg).14 Levels >4 mmol/L are generally associated with increased mortality and used as a threshold to initiate aggressive fluid resuscitation.15,16 Serial lactate levels are often used to assess adequacy of resuscitation and substantial decreases are associated with improved survival.17,18 Lactate levels may first increase at the onset of resuscitation as it is flushed from the tissues to central circulation before being metabolized. Trends must therefore be assessed.

Venous Oxygen Saturation

Normal range: 65% to 75%

The oxygen saturation in mixed venous blood describes the balance between systemic oxygen delivery and oxygen uptake. It can also be used as a target for resuscitation. Venous oxygen saturation (SvO2) measurement is taken from a pulmonary artery catheter and sometimes referred to as mixed venous oxygenation because blood from the pulmonary artery is considered to be a mix of venous blood from all tissue beds. Decreases in SvO2 indicate systemic oxygen delivery is impaired, which is often encountered in low-flow states (eg, heart failure, low cardiac index) or anemia.19 In contrast, SvO2 may be normal or high in patients with distributive shock.19,20 This often represents an inability of the tissues to extract the oxygen that has been delivered.

A surrogate of SvO2 is central venous oxygen saturation (ScvO2). ScvO2 is measured in the superior vena cava and reflects the oxygen saturation of venous blood from only the upper half of the body.19 It must be obtained from a central venous catheter. In general, ScvO2 is typically 5% to 7% higher than SvO2.21 Target thresholds for ScvO2 therefore are >70%.22 High ScvO2 values, however (ie, >90%), have been associated with poor outcomes.23


The Physiologic Approach (Henderson-Hasselbalch)

The physiologic approach is largely composed of the carbonic acid–bicarbonate buffering system and described by the Henderson-Hasselbalch equation.
pH= 6.1 + log [HCO3/(0.03×pCO2)]

In this equation, both HCO3 and pCO2 are independent variables, with HCO3 representing the base and pCO2 representing the acid. Disturbances that primarily affect pCO2 concentrations are called respiratory whereas those that affect HCO3 are called metabolic.

The pH is determined not by the absolute values of either but by the ratio of HCO3 to pCO2. That being said, both values may be largely abnormal (indicating an acid–base disorder is present) but pH may be in the normal range. In fact, a common pitfall is assuming an acid–base disorder can exist only when pH is abnormal.

The Stewart Approach

A second theory originated from the principles proposed by Peter Stewart and is based on the concept of strong ion difference (SID) and the laws of electrical neutrality.24

where A equals dissociated weak acids. This theory states that bicarbonate and hydrogen ions are the dependent variables and represent the effects rather than the causes of acid–base imbalances. The three independent variables that ultimately control blood pH are SID, pCO2, and the total weak acid concentration.

Strong ions are the ones that are completely dissociated (eg, Na+, K+, Ca2+, Mg2+, Cl, lactate) as opposed to weak ions, which can exist in both charged and uncharged forms (eg, albumin, phosphate, HCO3).25 The SID is the difference between the sum of all the strong cations and strong anions.12 Because not all strong ions can be measured, the apparent SID (aSID) can be calculated as follows:

In a healthy individual, the normal SID is approximately +40 to +42.25,26 As such, this difference must be counterbalanced with an equal opposing charge obtained from pCO2 and weak acids (eg, albumin and phosphate). This is referred to as the effective SID (eSID). The difference between the aSID and eSID is called the strong ion gap; in healthy individuals, it is equal to zero. When the aSID and eSID are not equal, as in a critically ill patient, the imbalance must be matched by a change in the concentration of another charged entity (HCO3, CO32−, OH, H+). This makes hydrogen and bicarbonate the dependent variables in this model.

In summary, the Stewart approach can be useful to explain pathophysiologic principles but is not associated with improved outcomes compared with the physiologic approach (ie, Henderson-Hasselbalch). As a result, the physiologic approach is more commonly used in bedside practice.


The metabolism of carbohydrates and fat results in the production of approximately 15,000 mmol of CO2 per day. In addition, digestion of proteins and tissue metabolism results in the production of nonvolatile acids. For normal cellular function to occur, hydrogen ion concentration must be maintained within a narrow therapeutic range. In fact, the normal variance of hydrogen ion in extracellular fluid is <10 nEq/L.2 The three mechanisms the body uses to maintain this tight therapeutic range are buffers, respiratory regulation, and renal regulation.


Buffers represent the first line of defense when an acid–base imbalance exists. A buffer is a substance that can absorb or donate hydrogen ions when in the presence of a strong acid or base and minimize resultant changes in pH. The principal extracellular buffer system in the body is the bicarbonate/carbonic acid system.

This system plays a central role because both HCO3 and CO2 can be regulated independently. Reactions in this system flow both ways depending on the concentration of each component. In this model, carbonic acid (H2CO3) and bicarbonate (HCO3) exist in equilibrium with hydrogen ions. In the presence of carbonic anhydrase, carbonic acid is converted to CO2. Carbon dioxide concentrations are regulated through ventilation (ie, the respiratory component) whereas bicarbonate concentrations are regulated through the kidney (ie, the metabolic component).

Other buffer systems that are present are the phosphate buffer system and intracellular and extracellular proteins. Both of these function more so as intracellular buffers.

Respiratory Regulation

The second line of defense against acid–base disturbances is the respiratory system. Within minutes of detecting an imbalance, chemoreceptors located in the medulla of the brain can modify ventilation to either retain or eliminate CO2. Specifically, an increase in ventilation decreases CO2 while a decrease in ventilation increases CO2. A new steady-state PaCO2 is typically reached within hours.1

Renal Regulation

The kidneys maintain acid–base homeostasis by regulating the concentration of bicarbonate in the blood. Approximately 4,300 mEq of bicarbonate is filtered to the kidney each day, all of which must be reabsorbed to maintain normal acid–base balance.27 Approximately 90% of this reabsorption takes place in the proximal tubule and is catalyzed by carbonic anhydrase. The remaining 10% is reabsorbed in the more distal segments. Filtered bicarbonate combines with hydrogen ions secreted by the renal tubule cell to form carbonic acid. The enzyme carbonic anhydrase, located in the brush border of the renal tubule, catalyzes conversion of carbonic acid to carbon dioxide. The uncharged CO2 readily crosses the cell membrane and passively diffuses into the renal tubule cell. Inside the cell, CO2 is converted to carbonic acid in the presence of intracellular carbonic anhydrase. Carbonic acid dissociates into hydrogen ion (which is later secreted into the tubular lumen) and bicarbonate (which is reabsorbed into capillary blood). Drugs that inhibit carbonic anhydrase (eg, acetazolamide) interfere with this process by blocking the reabsorption of bicarbonate, hence creating a metabolic acidosis.

The second mechanism used by the kidney is excretion of the daily load of nonvolatile acids that are produced by the body (approximately 50 to 100 mEq/day).27 This is accomplished by hydrogen ions combining with urinary buffers such as phosphates or with ammonia to form ammonium.

In general, renal compensation begins approximately 6 to 12 hours after an acid–base derangement, but full compensation takes roughly 3 to 5 days.28


Acid–base disorders can be categorized based on the pH derangement (ie, acidosis and alkalosis) and the primary disorder leading to that derangement (ie, respiratory and metabolic). There are four classifications for acid–base disorders: metabolic acidosis, metabolic alkalosis, respiratory acidosis, and respiratory alkalosis. The term respiratory is used when the primary problem is related to abnormal PaCO2 concentrations, whereas metabolic is used when the primary disorder is related to abnormalities in serum bicarbonate. For homeostasis to be maintained, there must be a compensatory response performed by the opposing system (respiratory for primary metabolic disorders and metabolic for primary respiratory disorders). This compensatory response, however, will neither completely correct nor overcorrect the primary disorder.

A simple disorder is considered to have a single disturbance with the expected degree of compensation. Mixed disorders, on the other hand, consist of a combination of disturbances that occurs simultaneously. While more than one metabolic disorder can coexist (eg, metabolic acidosis and metabolic alkalosis), there can only be one respiratory disorder at the same time.

Alternatively, clinicians should think of the following three groups when reviewing causes of acid–base disorders and resultant treatments: iatrogenic (eg, hyperchloremic metabolic acidosis from saline resuscitation), fixed feature of a preexisting disease process (eg, renal failure), or a liable feature of an evolving disease process (eg, lactic acidosis from shock).29 While the conditions listed in each example may all be classified as metabolic acidosis, the treatment plans are inherently different.


Step 1: pH Assessment

The first step in interpreting an ABG is to evaluate the pH and determine if an acidemia or alkalemia exists. A blood pH <7.4 is acidemia and >7.4 is alkalemia. As stated earlier, a normal pH does not mean that an acid base disorder does not exist.

Step 2: Determine Primary Acid–Base Disorder

The second step is to determine if the primary acid–base disorder is respiratory or metabolic. To do this the clinician should (1) pose the question, “If the primary problem were respiratory, would the PaCO2 be high or low?”; (2) use the principles of acid–base physiology to answer the question. Given PaCO2 is considered to be an acid, primary respiratory acidosis occurs when PaCO2 concentrations are high (>40 mm Hg), while primary respiratory alkalosis occurs when PaCO2 concentrations are low; and (3) confirm the actual PaCO2 value from the ABG.

The same approach can then be performed on the metabolic side by evaluating bicarbonate: (1) Pose the question, “If the primary problem were metabolic, would the HCO3 be high or low?” (2) Use the principles of acid–base physiology to answer the question. Given HCO3 is considered to be a base, primary metabolic acidosis occurs when HCO3 concentrations are low (<24 mEq/L), while primary metabolic alkalosis occurs when HCO3 concentrations are high. (3) Confirm the actual HCO3 value from the ABG. If the primary problem appears to be both respiratory and metabolic, then a mixed disorder exists.

Step 3: Compensatory Response

The third step is to evaluate the degree of compensation for the primary acid–based disturbance. Respiratory compensation can occur quickly after the detection of a primary metabolic problem (through adjustment in ventilatory rate), but metabolic compensation for a primary respiratory disorder occurs more slowly. Therefore, when primary respiratory disorders are detected, they are further classified as being either acute or chronic. From there, the expected change in either bicarbonate (for primary respiratory disorders) or PaCO2 (for primary metabolic disorders) can be calculated using the appropriate compensation formula (Table 13-1). If the calculated expected value differs substantially from the actual value, then a secondary disorder is present.

TABLE 13-1.

Summary of Primary Acid–Base Disorders and Their Compensatory Response






Metabolic acidosis

↓ HCO3

↓ PaCO2

Expected PaCO2 = (1.5 × HCO3) + (8 ± 2)

Metabolic alkalosis

↑ HCO3

↑ PaCO2

Expected PaCO2 = (0.7 × HCO3) + (21 ± 2)

Respiratory acidosis

↑ PaCO2

↑ HCO3

Acute: ΔpH = 0.008 × Δ PaCO2

ΔHCO3 = Δ PaCO2/10

Chronic: ΔpH = 0.003 × Δ PaCO2

ΔHCO3 = 3.5 (ΔPaCO2)/10

Respiratory alkalosis

↓ PaCO2

↓ HCO3

Acute: ΔpH = 0.008 × Δ PaCO2

ΔHCO3 = ΔPaCO2/5

Chronic: ΔpH = 0.003 × Δ PaCO2

∆HCO3 = ΔPaCO2/2

Step 4: Conditional Assessments for Metabolic Derangements

If a metabolic acidosis exists, the fourth step is to calculate the anion gap. The anion gap can be used to identify the cause of a metabolic acidosis. While the threshold for “high” will vary, values >16 are generally considered “positive.” The following disorders have been associated with an anion gap acidosis: lactic acidosis, ketoacidosis, toxic ingestions, and end-stage renal failure. A classic mnemonic that is used to distinguish causes of anion gap acidosis is MUDPILES: methanol, uremia, diabetic ketoacidosis, paraldehyde or propylene glycol, isoniazid or iron, lactic acid, ethylene glycol, and salicylates. A more recent mnemonic is GOLD MARK: glycols (ethylene and propylene), oxoproline (associated with chromic acetaminophen ingestion), l-lactate (most common measured form of lactic acid), d-lactate (typically seen in patients with short gut syndromes), methanol, aspirin, renal failure, and ketoacidosis.30

If a metabolic alkalosis exists, then the clinician should determine if the disorder is chloride responsive or chloride resistant. This is performed by assessing the urinary Cl. Urinary Cl concentrations <25 mEq/L are suggestive of chloride-responsive alkalosis while urinary Cl concentrations >40 mEq/L are chloride-resistant.1 Chloride-responsive alkalosis typically responds to intravenous normal saline (0.9% sodium chloride) administration. Common causes of chloride-responsive alkalosis include vomiting, gastrointestinal losses, gastrointestinal drainage, and diuretics. Chloride-resistant alkalosis, on the other hand, does not respond to normal saline administration and is often reflective of mineralocorticoid excess or severe hypokalemia. Other causes include Bartter syndrome, Cushing syndrome, Gitelman syndrome, severe hypercalcemia, and severe magnesium deficiency.


Metabolic Acidosis

Metabolic acidosis is caused by the net retention of nonvolatile acids or loss of bicarbonate. On ABG, both the pH and the serum bicarbonate would be low and the PaCO2 would decrease in an attempt to compensate. This would be manifested by an increase in respirations.

Traditionally, metabolic acidosis is classified by the anion gap, which can be used to identify the underlying cause. Metabolic acidosis with an increased anion gap commonly results from increased endogenous organic acid production whereas nonanion gap acidosis is often related to extensive loss of bicarbonate. Common causes of anion gap and nonanion gap acidosis are listed in Table 13-2. One cause of anion gap acidosis is lactic acidosis, which is further classified as being Type A (ie, hypoxic) or Type B (ie, nonhypoxic). Causes of Type A lactic acidosis include septic shock, mesenteric ischemia, hypoxemia, hypovolemic shock, carbon monoxide poisoning, and cyanide toxicity. Type B lactic acidosis can be caused by seizures, intoxication (eg, salicylate, ethylene glycol, propylene glycol), and medications (Table 13-2). A complete review of the patient’s medication list should therefore be performed to rule out potential drug-induced causes.

TABLE 13-2.

Causes of Metabolic Acidosis



Lactic acidosis

 Type A

  Septic shock

  Hypovolemic shock

  Mesenteric ischemia


  Severe anemia

  Carbon monoxide poisoning


 Type B


   Nonnucleoside reverse-transcriptase inhibitors







   IV lorazepam (vehicle)

   Sodium nitroprusside (cyanide)


  Diabetes mellitus



Gastrointestinal losses


 Ureteral diversions


Medications and Iatrogenic Causes

 Intravenous sodium chloride (excessive doses)

 Parenteral nutrition

 Carbonic anhydrous inhibitors



Renal tubular acidosis



 Ethylene glycol

 Propylene glycol


 Toluene ingestion






 Metabolic errors

Renal failure

Nonanion gap acidosis occurs when the decrease in bicarbonate ions corresponds with an increase in chloride ions to maintain electrical neutrality. A common cause of nonanion gap acidosis is hyperchloremic metabolic acidosis due to resuscitation with large volumes of normal saline. Nonanion gap acidosis is also encountered with renal tubular acidosis, excessive gastrointestinal loses (eg, diarrhea, fistula drainage, ureteral diversion), or iatrogenic causes (eg, parenteral nutrition, carbonic anhydrase inhibitors) (Minicases 1 and 3).

Metabolic Alkalosis

Metabolic alkalosis is caused by a net gain of bicarbonate or loss of hydrogen ion from the extracellular fluid. It is characterized on ABG by an increase in both pH and bicarbonate values. Although some respiratory compensation occurs as a result of hypoventilation and CO2 retention, the degree is relatively minor.

Patient Involved in a Motor Vehicle Collision

John W. is a 21-year-old man, previously healthy, who presents to the emergency department after a motor vehicle collision in which he sustained multiple rib fractures, a pelvic fracture, and a liver laceration. On physical exam, he is not following commands and is perseverating. His blood pressure on presentation is 81/42 mm Hg and his heart rate is 120 beats/min. He is aggressively resuscitated with normal saline and blood products (packed red blood cells, fresh frozen plasma, and platelets) and subsequently taken to the operating room for hemorrhagic shock. In the operating room, surgical hemostasis is achieved. The patient is extubated and admitted to the ICU on normal saline at a rate of 150 mL/hr. Laboratory values and ABG postop are as follows:

  • pH, 7.32

  • PaCO2, 36 mm Hg

  • PaO2, 90 mm Hg

  • HCO3, 20 mEq/L

  • Sodium, 149 mEq/L

  • Potassium, 3.1 mEq/L

  • Cl, 119 mEq/L

  • Total carbon dioxide, 22 mEq/L

  • Blood urea nitrogen (BUN) 22 mg/dL

  • Serum creatinine (SCr), 0.9 mg/dL

  • White blood cell count, 17,000 cells/mm3

  • Albumin, 3 g/dL

QUESTION: What acid–base disorder does this patient present with? What is the most likely cause?

DISCUSSION: The pH of 7.32 indicates acidemia. Because both the HCO3 and the PaCO2 are low, the primary disorder is metabolic acidosis. This is consistent with the patient’s admitting presentation of shock. The degree of respiratory compensation can be calculated using the formula PaCO2 = 1.5 × HCO3 + 8 ± 2. In this example, the actual PaCO2 (36 mm Hg) is within the range for expected PaCO2 (36 to 40 mm Hg), indicating this is a simple disorder. The next step is to calculate the anion gap, which is 8 (corrected for hypoalbuminemia, 10.5), consistent with a nonanion gap metabolic acidosis. Common causes of nonanion gap metabolic acidosis are hyperchloremic acidosis due to resuscitation with normal saline, renal tubular acidosis, and excessive gastrointestinal losses. Because this patient received large amounts of normal saline secondary to hypotension and shock, this is the most likely cause. Treatment should consist of reevaluating the rate of IV fluid administration and changing to a balanced crystalloid solution such as lactated ringers.

Patient with Acute Respiratory Depression

Pat R. is an 80-year-old woman who is admitted to the hospital after falling and breaking her hip. On hospital day 2, she undergoes surgery to repair her hip. Postoperatively, her stay has been relatively unremarkable with the exception of pain control. Upon evaluation, her pain scores have been no lower than 8 out of 10 for which she has been receiving IV hydromorphone. On postoperative day 3, a “rapid assessment” code is called because her oxygen saturation falls to 81% and her respiratory rate is 4 breaths/min. The nurse states she has been receiving 2 mg of hydromorphone every 2 hours around the clock and her last dose was approximately 15 minutes ago. Her vital signs are heart rate 100 beats/min, blood pressure 110/70 mm Hg, temperature 37.6°C. Laboratory results, and ABG are as follows:

  • Na, 136 mEq/L

  • K, 5.1 mEq/L

  • Chloride, 98 mEq/L

  • Total carbon dioxide, 28 mEq/L

  • BUN, 25 mg/dL

  • SCr, 0.8 mg/dL

  • ABG, pH 7.26

  • PaCO2, 58 mm Hg

  • PaO2, 55 mm Hg

  • HCO3, 26 mEq/L

QUESTION: What acid–base disorder does this patient present with? What is the most likely cause?

DISCUSSION: The pH of 7.26 indicates acidemia. Because the PaCO2 and HCO3 are both high, the primary cause is a respiratory acidosis. This is consistent with her clinical presentation of respiratory distress and inability to eliminate carbon dioxide. With respiratory acidosis, metabolic compensation is delayed and changes in HCO3 are often minimal and rarely >31 mEq/L in the acute setting. In this example, the expected degree of compensation is appropriate for an acute disorder. The most likely cause of respiratory acidosis in this case is respiratory depression secondary to narcotic overuse. Treatment should consist of supplemental oxygen and reversal of hydromorphone with naloxone.

Metabolic alkalosis is delineated into two types: chloride responsive and chloride resistant (Table 13-3). A common drug-related cause of metabolic alkalosis is diuretic therapy. Diuretics cause a wasting of Cl in association with sodium (Na)+ and potassium (K)+ without a proportional increase in bicarbonate excretion. In addition, volume depletion leads to hyperaldosteronism, H+ secretion, and bicarbonate resorption (Minicase 4).

TABLE 13-3.

Causes of Metabolic Alkalosis



Gastrointestinal losses


 Gastric drainage

 Nasogastric suctioning

 Chloride wasting diarrhea



Mineralocorticoid excess

Exogenous steroids

Increased renin/aldosterone states

Cushing syndrome

Liddle syndrome

Bartter syndrome

Respiratory Acidosis

Respiratory acidosis is usually a direct result of hypoventilation; therefore, any situation associated with decreased respiratory rate can lead to its occurrence. It is, therefore, important to identify and treat the underlying cause (Table 13-4). Laboratory findings consistent with respiratory acidosis are decreased pH and increased PaCO2. Because renal compensation is slower to respond, the increase in bicarbonate is only modest at first. With respiratory acidosis, the biggest threat to life is not from acidemia but from hypoxia. In patients breathing room air, PaCO2 cannot exceed 80 mm Hg or else life-threatening hypoxia may result10 (Minicase 2).

TABLE 13-4.

Causes of Respiratory Acidosis






 Head injury

 Status epilepticus

Perfusion abnormalities

 Pulmonary embolism

 Cardiac arrest

Airway abnormalities



 Chronic obstructive pulmonary disease


 Pulmonary edema

 Acute respiratory distress syndrome


 Brainstem or cervical cord injury

 Guillain-Barré syndrome

 Myastenia gravis

Parenteral nutrition

Respiratory Alkalosis

Respiratory alkalosis is characterized by excessive elimination of CO2 through hyperventilation. Laboratory derangements include increased pH and decreased PaCO2. Renal compensation is characterized by inhibition of bicarbonate reabsorption, which is complete within several days. It is one of the most frequently encountered acid–base disorders and is associated with several pathologic conditions (Table 13-5).10

TABLE 13-5.

Causes of Respiratory Alkalosis

CNS-Respiratory Stimulation






 Progesterone derivatives


 Cerebrovascular accidents



 Congestive heart failure

 High altitude

 Pulmonary edema

 Pulmonary embolism


Acid–base disorders are highly prevalent in critically ill patients. Acid–base status is assessed measuring an arterial blood gas and evaluation of pH, PaCO2, and HCO3. Carbon dioxide is the most abundant acid that is controlled through respiratory regulation, whereas bicarbonate is the most abundant base and concentrations are maintained by the kidney. There are four primary acid–base disturbances that are categorized by the type (ie, acidosis versus alkalosis) and origin (metabolic versus respiratory). When evaluating for the presences of acid–base disorders, a systematic approach should be taken using data obtained from the ABG, serum electrolytes, and clinical presentation. Medications are frequently implicated in acid–base disorders; thus, careful assessment of the medication profile is necessary. These skills can be used to assist clinicians with the identification, prevention, and treatment of acid–base disorders.

Patient with Severe Abdominal Pain

Jill M. is a 67-year-old woman who presents to the hospital with reports of severe abdominal pain and distention that has been present for the last 2 days. She is tachycardic, tachypneic, and normotensive. Her temperature is 38.4°C. A computed tomography scan reveals a large amount of intraabdominal free air and she is taken to the operating room for an exploratory laparotomy. In the operating room, a colon perforation is noted with a significant amount of fecal contamination and pus. The colon is repaired and the wound is left open, with a wound-vac in place. She returns to the ICU, mechanically ventilated on maintenance IV fluids (normal saline at 100 mL/hr) and antibiotics. Postoperatively, her blood pressure remains low and is currently 88/56 mm Hg. A bolus of 0.9% sodium chloride is administered and a norepinephrine infusion is initiated. Pertinent laboratory values are as follows:

  • Sodium, 146 mEq/L

  • Potassium, 3.8 mEq/L

  • Chloride, 112 mEq/L

  • Total carbon dioxide, 16 mEq/L

  • BUN, 38 mg/dL

  • SCr, 1.3 mg/dL

  • White blood cells, 21,000 cells/mm3

  • Albumin, 2.5 g/dL

  • Lactate, 5.2 mmol/L

  • ABG, pH 7.28

  • PaCO2, 38 mm Hg

  • PaO2, 72 mm Hg

  • HCO3, 14 mEq/L

QUESTION: What acid–base disorder does this patient present with? What is the most likely cause?

DISCUSSION: The pH is 7.28, which indicates acidemia. Because the HCO3 is low and the PaCO2 is low, the primary disturbance is metabolic acidosis. The expected degree of respiratory compensation, which is calculated using the formula expected PaCO2 = 1.5 × HCO3 + 8 ± 2, is less than the actual PaCO2; therefore, a secondary respiratory acidosis exists. Because the primary problem is a metabolic acidosis, the next step is to calculate the anion gap. The anion gap is 18 (corrected for albumin, 22); therefore, this is an anion gap, metabolic acidosis. Common causes of anion gap acidosis are lactic acidosis, ketosis, renal failure, and poisonings. Because this patient has an elevated lactate level and presents with an infection and hypotension requiring vasopressors, the most likely cause is lactic acidosis secondary to septic shock. The treatment plan should consist of fluid resuscitation and vasopressor support (as needed) to restore tissue perfusion along with broad-spectrum antibiotics. Low-dose corticosteroid therapy could be considered if hypotension persists despite fluids and vasopressor therapy.

Patient with Fever and Shortness of Breath

Beth D. is a 65-year-old woman who is admitted to the hospital with fever and shortness of breath. A pulse oximeter reveals an oxygen saturation of 89%, her chest radiograph demonstrates a right lower lobe infiltrate, and her blood pressure is 88/49 mm Hg. IV fluids are administered and she is subsequently intubated. She is admitted to the ICU with the diagnosis of sepsis secondary to pneumonia. Throughout her ICU stay, she has required a significant amount of IV fluids to maintain a mean arterial pressure (MAP) goal of 65 mm Hg with intermittent use of vasopressors. Currently, her blood pressure remains stable, but the cumulative fluid balance for her hospital stay is +10.6 L and the team would like to wean her off the ventilator; a furosemide infusion is administered. Over the next 48 hours, her daily fluid balance has been –5 L (day 1) and –3.8 L (day 2). Laboratory results and ABG reveal the following: pH 7.53, PaCO2 48 mm Hg, PaO2 65 mm Hg, and HCO3 35 mEq/L. The remainder of her laboratory results are as follows:

  • Sodium, 141 mEq/L,

  • Potassium, 4.2 mEq/L

  • Chloride, 99 mEq/L

  • Total carbon dioxide, 36 mEq/L

  • BUN, 48 mg/dL

  • SCr, 1.2 mg/dL

  • Urine chloride, 7 mEq/L

QUESTION: What acid–base disorder does this patient present with? What is the most likely cause?

DISCUSSION: The pH is 7.53, which indicates alkalemia. Because the HCO3 is high and the PaCO2 is high, the primary problem is metabolic. The degree of compensation by the respiratory side is appropriate as per the formula expected PaCO2 = (0.7 × HCO3) + 21 ± 2. Because the primary problem is metabolic alkalosis, the next step is to determine if it is chloride responsive or resistant. This is done by assessing the urinary chloride. Because the urinary chloride is 7 mEq/L, this would be classified as chloride responsive. There are several potential causes of metabolic alkalosis in this patient. Diuretic administration can cause metabolic alkalosis through aldosterone secretion and increased chloride excretion. Loss of gastrointestinal fluid through nasogastric suction leads to a loss of hydrogen and chloride ions. In addition, parenteral nutrition can be an iatrogenic cause of alkalosis if excessive amounts of acetate are provided in the formula. The most likely cause in this patient is excessive diuresis secondary to the furosemide infusion. Treatment would first consist of discontinuing the diuresis. IV fluid administration can be considered but this may interfere with ventilator weaning. Instead, acetazolamide would be preferred to lower the serum bicarbonate.


1. Can an acid–base disorder exist if a patient presents with a normal pH?

ANSWER: Yes, an acid–base disorder can still exist even when the pH is within the normal range. pH is determined by the ratio of base to acid as opposed to the individual concentration of one (ie, acid or base) independently. Therefore, an acidosis can be present without an acidemia if a coexisting alkalosis is present. Conversely, an alkalosis can be present without an alkalemia if a coexisting acidosis is present. In these settings, evaluation of the PaCO2 and HCO3 reveals abnormal values (hence a pathophysiologic process), but each offsets the other and the net effect on pH is negligible. Careful evaluation of the patient’s history, clinical presentation, physical exam, and laboratory values is necessary to identify the acid–base disorders that are present.

2. What are the steps to follow to assess an ABG?

ANSWER: The first step is to determine if an acidemia or alkalemia is present. This is done by evaluating the pH. Once the correct categorization is made, the second step is to determine if the primary cause is metabolic or respiratory. This is performed by assessing the PaCO2 and the HCO3 on the ABG. Once the primary cause is determined, the opposing side should compensate. The third step is to assess if the degree of compensation is appropriate. If it is not, then a secondary disorder is present. Finally, if the primary problem is metabolic acidosis, or metabolic alkalosis, then the anion gap or urinary chloride, respectively, should be assessed to assist with identification of the possible cause.

3. How can the anion gap be used to assess acid–base disorders?

ANSWER: The anion gap is an estimate of the relative abundance of unmeasured anions. It can suggest the possible causes of metabolic acidosis, particularly if the disorder is secondary to an accumulation of nonvolatile acids or a net loss of bicarbonate. When the anion gap is normal, the acidosis is usually caused by a loss of bicarbonate ions; common causes include diarrhea, early renal insufficiency, and infusion of large amounts of isotonic saline. When the anion gap is elevated, causes may include lactic acidosis, ketoacidosis, end-stage renal failure, or certain toxic ingestions. Correction of the anion gap for hypoalbuminemia can improve the accuracy of this approach. The anion gap represents one factor that can help determine the etiologic cause of metabolic acidosis and should not be interpreted as absolute, especially in a complex intensive care unit (ICU) patient.


  • 1.

    Berend K, de Vries AP, Gans RO. Physiological approach to assessment of acid-base disturbances. N Engl J Med. 2014;371(15):1434-1445.PubMed

    • Search Google Scholar
    • Export Citation
  • 2.

    Marino PL. Marino’s The ICU Book. 4th ed. Philadelphia, PA: Wolters Kluwer Health/Lippincott Williams & Wilkins; 2014.

  • 3.

    Adrogué HJ, Madias NE. Management of life-threatening acid-base disorders. Second of two parts. N Engl J Med. 1998;338(2):107-111.PubMed

    • Search Google Scholar
    • Export Citation
  • 4.

    Narins RG, Emmett M. Simple and mixed acid-base disorders: a practical approach. Medicine (Baltimore). 1980;59(3):161-187.PubMed

  • 5.

    Weil MH, Rackow EC, Trevino R, et al.Difference in acid-base state between venous and arterial blood during cardiopulmonary resuscitation. N Engl J Med. 1986;315(3):153-156.PubMed

    • Search Google Scholar
    • Export Citation
  • 6.

    Faridi AB, Weisberg LS. Acid-Base, Electrolyte and Metabolic Abnormalities. Parrillo JE, and Dellinger RP, eds. Critical Care Medicine: Principles of Diagnosis and Management in the Adult. 3rd ed. Philadelphia, PA: Mosby Elsevier; 2008:1203-1243.

    • Search Google Scholar
    • Export Citation
  • 7.

    Ranieri VM, Rubenfeld GD, Thompson BT, et al.Acute respiratory distress syndrome: the Berlin Definition. JAMA. 2012;307(23):2526-2533.PubMed

    • Search Google Scholar
    • Export Citation
  • 8.

    Wilson RF. Critical Care Manual: Applied Physiology and Principles of Therapy. 2nd ed. Philadelphia, PA: F.A. Davis Company; 1992.

  • 9.

    Figge J, Jabor A, Kazda A, et al.Anion gap and hypoalbuminemia. Crit Care Med. 1998;26(11):1807-1810.PubMed

  • 10.

    Al-Jaghbeer M, Kellum JA. Acid-base disturbances in intensive care patients: etiology, pathophysiology and treatment. Nephrol Dial Transplant. 2015;30(7):1104-1111.PubMed

    • Search Google Scholar
    • Export Citation
  • 11.

    Gilfix BM, Bique M, Magder S. A physical chemical approach to the analysis of acid-base balance in the clinical setting. J Crit Care. 1993;8(4):187-197.PubMed

    • Search Google Scholar
    • Export Citation
  • 12.

    Kellum JA. Disorders of acid-base balance. Crit Care Med. 2007;35(11):2630-2636.PubMed

  • 13.

    Salem MM, Mujais SK. Gaps in the anion gap. Arch Intern Med. 1992;152(8):1625-1629.PubMed

  • 14.

    Singer M, Deutschman CS, Seymour CW, et al.The Third International Consensus definitions for sepsis and septic shock (sepsis-3). JAMA. 2016;315(8):801-810.PubMed

    • Search Google Scholar
    • Export Citation
  • 15.

    Aduen J, Bernstein WK, Khastgir T, et al.The use and clinical importance of a substrate-specific electrode for rapid determination of blood lactate concentrations. JAMA. 1994;272(21):1678-1685.PubMed

    • Search Google Scholar
    • Export Citation
  • 16.

    Levy MM, Evans LE, Rhodes A. The Surviving Sepsis Campaign Bundle: 2018 Update. Crit Care Med. 2018;46(6):997-1000.PubMed

  • 17.

    Jansen TC, van Bommel J, Schoonderbeek FJ, et al.Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial. Am J Respir Crit Care Med. 2010;182(6):752-761.PubMed

    • Search Google Scholar
    • Export Citation
  • 18.

    Vincent JL, Dufaye P, Berré J, et al.Serial lactate determinations during circulatory shock. Crit Care Med. 1983;11(6):449-451.PubMed

  • 19.

    Vincent JL, De Backer D. Circulatory shock. N Engl J Med. 2013;369(18):1726-1734.PubMed

  • 20.

    Marik PE. Early management of severe sepsis: concepts and controversies. Chest. 2014;145(6):1407-1418.PubMed

  • 21.

    Reinhart K, Kuhn HJ, Hartog C, et al.Continuous central venous and pulmonary artery oxygen saturation monitoring in the critically ill. Intensive Care Med. 2004;30(8):1572-1578.PubMed

    • Search Google Scholar
    • Export Citation
  • 22.

    Rhodes A, Evans LE, Alhazzani W, et al.Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Crit Care Med. 2017;45(3):486-552.PubMed

    • Search Google Scholar
    • Export Citation
  • 23.

    Pope JV, Jones AE, Gaieski DF, et al.Multicenter study of central venous oxygen saturation (ScvO2) as a predictor of mortality in patients with sepsis. Ann Emerg Med. 2010;55(1):40-46.e1.PubMed

    • Search Google Scholar
    • Export Citation
  • 24.

    Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol. 1983;61(12):1444-1461.PubMed

  • 25.

    Kellum JA. Determinants of blood pH in health and disease. Crit Care. 2000;4(1):6-14.PubMed

  • 26.

    Morgan TJ. The meaning of acid-base abnormalities in the intensive care unit: part III. Effects of fluid administration. Crit Care. 2005;9(2):204-211.PubMed

    • Search Google Scholar
    • Export Citation
  • 27.

    Rose DB, Post TW. Clinical Physiology of Acid-Base and Electrolyte Disorders. 5th ed. New York, NY: McGraw-Hill; 2001.

  • 28.

    Dzierba AL, Abraham P. A practical approach to understanding acid-base abnormalities in critical illness. J Pharm Pract. 2011;24(1):17-26.PubMed

    • Search Google Scholar
    • Export Citation
  • 29.

    Kaplan LJ, Frangos S. Clinical review: acid-base abnormalities in the intensive care unit. Part II. Crit Care. 2005;9(2):198-203.PubMed

  • 30.

    Mehta AN, Emmett JB, Emmett M. GOLD MARK: an anion gap mnemonic for the 21st century. Lancet. 2008;372(9642):892.PubMed

QUICKVIEW | Venous Total Carbon Dioxide Content (Venous Serum Bicarbonate)




Common reference range


22–30 mEq/L

Venous bicarbonate can be 1.5–3 mEq/L higher than arterial measure of bicarbonate

Critical value

<10 mEq/L or >40 mEq/L

Reference range and critical values may vary based on local laboratory standards

Inherent activity


Primary substance responsible for buffering acids



Byproduct of typical cell metabolism


Exchanged via circulation


Renal excretion (reabsorption occurs at proximal tubule)

Causes of abnormal values


Metabolic alkalosis and respiratory acidosis

Change in HCO3 is the primary cause of metabolic acid–base disorders. In contrast, change in HCO3 is the primary method of compensation for respiratory disorders.


Metabolic acidosis and respiratory alkalosis

Signs and symptoms

High level

Related to primary process

Low level

Related to primary process

After event, time to….

Initial elevation

6–12 hr to initiate compensation

Assumes acute insult

Peak values

None (will rise until pH balanced)

Assumes insult not yet removed


3–5 days to complete compensation

Assumes insult removed and nonpermanent damage

Causes of spurious results

Inadvertent venous sampling, delayed time to analysis

QUICKVIEW | Arterial Partial Pressure of Carbon Dioxide




Common reference range


35–45 mm Hg

Critical value

>70 mm Hg (may require mechanical ventilation)

Inherent activity


Primary volatile acid in the body



Generated intracellularly from carbon dioxide and water




Excreted by the lungs during expiration

Causes of abnormal values


Respiratory acidosis and metabolic alkalosis

Change is PaCO2 is the primary cause of respiratory acid–base disorders. In contrast, change in PaCO2 is the primary method of compensation for metabolic disorders.


Respiratory alkalosis and metabolic acidosis

Signs and symptoms

High level

Respiratory failure

Low level

Related to primary process

After event, time to….

Initial elevation

Minutes to hours

Assumes acute insult

Peak values


Assumes insult not yet removed


Hours to days

Assumes insult removed and nonpermanent damage

Causes of spurious results

Inadvertent venous sampling, delayed time to analysis

Higher carbon dioxide content

N/A = not applicable.