The concentrations of hydrogen and bicarbonate ions in plasma must be precisely regulated to optimize enzyme activity, oxygen transport, and rates of chemical reactions within cells. Each day approximately 15,000 mmol of carbon dioxide (which can generate carbonic acid as it combines with water) and 50 to 100 mEq of nonvolatile acid (mostly sulfuric acid) are produced and must be eliminated safely. The body is able to maintain this intricate acid-base balance by utilizing buffers, pulmonary excretion of carbon dioxide, and renal elimination of acid. This chapter will define concepts important for understanding acids and bases, discuss clinical measurements of blood gases and their interpretation, and present a diagnostic approach to common acid-base disturbances.
Acids and Bases
Bronsted and Lowry defined an acid as a molecule that can act as a proton (H + ) donor and a base as a molecule that can act as a proton acceptor. In physiologic solutions, a strong acid is a substance that readily and irreversibly gives up an H + , and a strong base avidly binds H + . In contrast, biologic molecules are either weak acids or bases, which reversibly donate H + or reversibly bind H + .
Acidemia and Acidosis
A blood pH less than 7.35 is called acidemia and a pH greater than 7.45 is called alkalemia , regardless of the mechanism. The underlying process that lowers the pH is called an acidosis , and the process that raises the pH is known as an alkalosis . A patient can have a mixed disorder with both an acidosis and an alkalosis concurrently, but can only be either acidemic or alkalemic. The last two terms are mutually exclusive.
Base excess (BE) is usually defined as the amount of strong acid (hydrochloric acid for BE greater than zero) or strong base (sodium hydroxide for BE less than zero) required to return 1 L of whole blood exposed in vitro to a P co 2 of 40 mm Hg to a pH of 7.4. Instead of an actual titration, the blood gas machine calculates the BE with algorithms utilizing plasma pH, blood P co 2 , and hemoglobin concentration. The number is supposed to refer to the nonrespiratory or metabolic component of an acid-base disturbance. A BE less than zero (also called a base deficit ) suggests the presence of a metabolic acidosis, and a value greater than zero suggests the presence of a metabolic alkalosis. In vitro, the number has been accurate, but in the living organism, because ions do cross beyond vascular and cellular boundaries, a primary acute change in Pa co 2 sometimes can cause the BE to move in the opposite direction, despite an unchanged metabolic acid-base status. In clinical practice, the BE is often used as a surrogate measure for lactic acidosis, which is one measurement to help determine adequacy of intravascular volume resuscitation.
Regulation of the Hydrogen Ion Concentration
At 37° C, the normal hydrogen ion concentration in arterial blood and extracellular fluid is 35 to 45 nmol/L, which is equivalent to an arterial pH of 7.45 to 7.35, respectively. The normal plasma bicarbonate ion concentration is 24 ± 2 mEq/L. The intracellular hydrogen ion concentration is approximately 160 nmol/L, which is equivalent to a pH of 6.8.
Physiologic changes to acid-base disturbances are corrected by three systems—buffers, ventilation, and renal response. The buffer systems provide an immediate chemical response. The ventilatory response occurs in minutes whenever possible, and, lastly, the renal response can slowly provide nearly complete restoration of the pH, but it can take days.
A buffer is defined as a substance within a solution that can prevent extreme changes in pH. A buffer system is composed of a base molecule and its weak conjugate acid. The base molecules of the buffer system bind excess hydrogen ions, and the weak acid protonates excess base molecules. The dissociation ionization constant (pKa) indicates the strength of an acid and is derived from the classic Henderson-Hasselbalch equation ( Fig. 21.1 ). The pKa is the pH at which an acid is 50% protonated and 50% deprotonated. Hydrochloric acid, a strong acid, has a pKa of −7, whereas carbonic acid, a weak acid, has a pKa of 6. The most important buffer systems in blood, in order of importance, are the (1) bicarbonate buffer system (H 2 CO 3 /HCO 3 − ), (2) hemoglobin buffer system (HbH/Hb), (3) other protein buffer systems (PrH/Pr − ), (4) phosphate buffer system (H 2 PO 4 − /HPO 4 2− ), and (5) ammonia buffer system (NH 3 /NH 4 + ).
Bicarbonate Buffer System
Carbon dioxide, generated through aerobic metabolism, slowly combines with water to form carbonic acid, which spontaneously and rapidly deprotonates to form bicarbonate ( Fig. 21.2 ). In this system, the base molecule is bicarbonate, and its weak conjugate acid is carbonic acid. Less than 1% of the dissolved carbon dioxide undergoes this reaction because it is so slow. However, the enzyme carbonic anhydrase, present in the endothelium, erythrocytes, and kidneys, catalyzes this reaction to accelerate the formation of carbonic acid and make this the most important buffering system in the human body when combined with renal control of bicarbonate and pulmonary control of carbon dioxide.
Hemoglobin Buffer System
The hemoglobin protein is the second most important buffering system because of multiple histidine residues. Histidine is an effective buffer from pH 5.7 to 7.7 (pKa 6.8) because it contains multiple protonatable sites on the imidazole side chains. Buffering by hemoglobin depends on the bicarbonate system to facilitate the movement of carbon dioxide intracellularly. Carbon dioxide freely diffuses into erythrocytes, where carbonic anhydrase resides. There, carbon dioxide combines with water to form carbonic acid, which rapidly deprotonates. The generated protons are bound by hemoglobin. The bicarbonate anions are exchanged electroneutrally back into plasma with extracellular chloride (chloride or Hamburger shift) ( Fig. 21.3 ). At the lungs, the reverse process occurs. Chloride ions move out of the red blood cells as bicarbonate enters for conversion back to carbon dioxide. The carbon dioxide is released back into plasma and is eliminated by the lungs. This process allows a large fraction of extrapulmonary carbon dioxide to be transported back to the lungs as plasma bicarbonate.
Oxygenated and deoxygenated hemoglobin have different affinities for hydrogen ions and carbon dioxide. Deoxyhemoglobin takes up more hydrogen ions, which shifts the carbon dioxide/bicarbonate equilibrium to produce more bicarbonate and facilitates removal of carbon dioxide from peripheral tissues for release into the lungs. Oxyhemoglobin favors the release of hydrogen ions and shifts the equilibrium to more carbon dioxide formation. At physiologic pH, a small amount of carbon dioxide is also carried as carbaminohemoglobin. Deoxyhemoglobin has a greater affinity (3.5 times) for carbon dioxide, so venous blood carries more carbon dioxide than arterial blood. These two mechanisms combine to account for the difference in carbon dioxide content of arterial versus venous plasma (25.6 mmol/L vs. 27.7 mmol/L, respectively) (Haldane effect).
Central chemoreceptors lie on the anterolateral surface of the medulla and respond to changes in cerebrospinal fluid pH. Carbon dioxide diffuses across the blood-brain barrier to elevate cerebrospinal fluid (CSF) hydrogen ion concentration, which activates the chemoreceptors and increases alveolar ventilation. The relationship between Pa co 2 and minute ventilation is almost linear except at very high arterial Pa co 2 , when carbon dioxide narcosis develops, and at very low arterial Pa co 2 , when the apneic threshold is reached. There is a very wide variation in individual Pa co 2 /ventilation response curves, but minute ventilation generally increases 1 to 4 L/min for every 1 mm Hg increase in Pa co 2 . During general anesthesia, spontaneous ventilation will cease when the Pa co 2 decreases to less than the apneic threshold, whereas in the awake patient, cortical influences prevent apnea, so the apneic threshold is not ordinarily observed.
Peripheral chemoreceptors are located at the bifurcation of the common carotid arteries and surrounding the aortic arch. The carotid bodies are the principal peripheral chemoreceptors and are sensitive to changes in Pa o 2 , Pa co 2 , pH, and arterial perfusion pressure. They communicate with the central respiratory centers via the glossopharyngeal nerves. Unlike the central chemoreceptors, which are more sensitive to hydrogen ions, the carotid bodies are most sensitive to Pa o 2 . Bilateral carotid endarterectomies abolish the peripheral chemoreceptor response, and these patients have almost no hypoxic ventilatory drive (also see Chapter 25 ).
The stimulus from central and peripheral chemoreceptors to either increase or decrease alveolar ventilation diminishes as the pH approaches 7.4 such that complete correction or overcorrection is not possible. The pulmonary response to metabolic alkalosis is usually less than the response to metabolic acidosis. The reason is because progressive hypoventilation results in hypoxemia when breathing room air. Hypoxemia activates oxygen-sensitive chemoreceptors and limits the compensatory decrease in minute ventilation. Because of this, the Pa co 2 usually does not rise above 55 mm Hg in response to metabolic alkalosis for patients not receiving oxygen supplementation.
Renal effects are slower in onset and may not be maximal for up to 5 days. The response occurs via three mechanisms: (1) reabsorption of the filtered HCO 3 − , (2) excretion of titratable acids, and (3) ammonia ( Fig. 21.4 ). Carbon dioxide combines with water in the renal tubular cell. With the help of carbonic anhydrase, the bicarbonate produced enters the bloodstream while the hydrogen ion is exchanged with sodium and is released into the renal tubule. There, H + combines with filtered bicarbonate and dissociates into carbon dioxide and water with help from carbonic anhydrase located in the luminal brush border, and the carbon dioxide diffuses back into the renal tubular cell. The proximal tubule reabsorbs 80% to 90% of the bicarbonate this way, while the distal tubule takes care of the remaining 10% to 20%. Once the bicarbonate is reclaimed, further hydrogen ions can combine with HPO 4 2− to form H 2 PO 4 − , which is eliminated in the urine. The last important urinary buffer is ammonia. Ammonia is formed from deamination of glutamine, an amino acid. The ammonia passively crosses the cell membrane to enter the tubular fluid. In the tubular fluid, it combines with hydrogen ion to form NH 4 + , which is trapped within the tubule and excreted in the urine. All of these steps allow for generation and return of bicarbonate into the bloodstream. The large amount of bicarbonate filtered by the kidneys allows for rapid excretion if necessary for compensation during alkalosis. The kidneys are highly effective in protecting the body against alkalosis except in association with sodium deficiency or mineralocorticoid excess.
Analysis of Arterial Blood Gases
The ability to measure arterial blood gas (ABG) and venous blood gas has revolutionized patient care during anesthesia and in the intensive care unit. Although pulse oximetry and capnography can be monitored continuously, analysis of ABGs has increased our diagnostic ability and the accuracy of our measurements.
Blood Gas and pH Electrodes
The pH electrode is a silver/silver chloride electrode encased in a special pH-sensitive glass that contains a buffer solution with a known pH. The electrode is placed in a blood sample and measures changes in voltage. The potential difference generated across the glass and a reference electrode is proportional to the difference in hydrogen ion concentration. Both electrodes must be kept at 37° C and calibrated with buffer solutions of known pH.
The O 2 electrode is known as the Clark or polarographic electrode. It has a silver/silver chloride reference electrode that is immersed in a potassium chloride solution. Electrons are formed by the oxidation reaction of the silver with the chloride ions of the potassium chloride electrolyte solution. The electrons are then free to combine with O 2 molecules at the platinum cathode. The platinum surface is covered with an oxygen-permeable membrane (polyethylene), on the other side of which is placed the unknown sample. Current flow is increased if oxygen concentration is higher and more electrons are taken up. The current is directly proportional to the P o 2 .
Carbon Dioxide Electrode
The carbon dioxide sensor was first described by Stow in 1957 and then modified by Bradley and Severinghaus. The carbon dioxide electrode is a pH electrode immersed in a sodium bicarbonate solution and is separated from the blood specimen by a Teflon semipermeable membrane. The carbon dioxide in the sample diffuses into the sodium bicarbonate solution producing hydrogen ions and bicarbonate. The measured pH in the bathing solution is altered in direct proportion to the logarithm of the P co 2 .
Arterial blood is most often obtained percutaneously from the radial, brachial, or femoral artery. In certain clinically stable situations, peripheral venous blood may serve as an approximation and save an arterial puncture. Venous pH is only 0.03 to 0.04 less than arterial values. Venous blood cannot be used for estimation of oxygenation because venous P o 2 (Pv o 2 ) is significantly less than Pa o 2 . Also, depending on the site of the venous blood draw, differences in tissue metabolic activity may alter Pv o 2 . The correlation between arterial and venous blood gas measurements varies with the hemodynamic stability of the patient. Periodic correlations of arterial and venous measurements should be performed especially when venous measurements are used for serial monitoring in critically ill patients.
A heparinized, bubble-free, fresh blood sample is required for blood gas analysis. In the past, liquid heparin was aspirated into a syringe and then expelled. This small amount of heparin remaining in the syringe was enough to anticoagulate the sample. Excessive amounts of anticoagulant in the sampling syringe could falsely dilute the measured P o 2 , P co 2 , and ionized calcium. Commercially prepared syringes with preweighed lyophilized electrolyte-balanced heparin are used in most hospitals now. Air bubbles should be removed because equilibration of oxygen and carbon dioxide in the blood with the corresponding partial pressures in the air bubble could influence the measured results. A delay in analysis can lead to oxygen consumption and carbon dioxide generation by the metabolically active white blood cells. Usually this error is small and can be reduced by placing the sample on ice. In some leukemia patients with a markedly increased white blood cell count, this error can be large and lead to a falsely low P o 2 even though the patient’s oxygenation is acceptable. This phenomenon is often referred to as leukocyte larceny and has also been described with extreme thrombocytosis (platelet larceny).
Decreases in temperature decrease the partial pressure of a gas in solution, even though the total gas content does not change. Both P co 2 and P o 2 decrease during hypothermia, but serum bicarbonate is unchanged. This leads to an increase in pH if the blood could be measured at the patient’s temperature. A blood gas with a pH of 7.4 and P co 2 of 40 mm Hg at 37° C will have a pH of 7.58 and a P co 2 of 23 mm Hg at 25° C. Unfortunately, all blood gas samples are measured at 37° C, which raises the issue of how to best manage the ABG measurement in hypothermic patients. This has led to two schools of thought: alpha stat and pH stat.
Alpha refers to the protonation state of the imidazole side chain of histidine. The pKa of histidine changes with temperature so that its protonation state is relatively constant regardless of temperature. The term alpha stat developed because as the patient’s pH was allowed to drift with temperature, the protonation state of the histidine residues remained static . This concept arose from the observation that cold-blooded poikilothermic animals functioned well over a wide range of body temperatures, yet they relied on a similar complement of enzymes as warm-blooded homeothermic animals. During cardiopulmonary bypass, an anesthesia provider using alpha stat would manage the patient based on an ABG measured at 37° C and strive to keep that pH at 7.4, but the patient’s true pH would be higher. No extra adjustments would be made for the patient’s hypothermia.
pH stat is different from alpha stat in that it requires keeping a patient’s pH static at 7.4 based on the core temperature (similar to that of a hibernating, homeothermic animal). During cardiopulmonary bypass, an anesthesia provider using pH stat would manage the patient based on an ABG that is corrected for the patient’s temperature. With hypothermia, this usually means adding carbon dioxide so that the patient’s temperature-correct (hypothermic) blood gas has a pH of 7.4. The lower pH and higher P co 2 maintained during pH stat may improve cerebrovascular perfusion during hypothermia; however, there is still debate about which method provides better outcomes.
The same physical properties exist for oxygen and hypothermia as for carbon dioxide. Decreases in temperature decrease the partial pressure of a gas in solution, so temperature correction of P o 2 remains relatively important for assessing oxygenation at the extremes of temperature. To be exact, the change in P o 2 with respect to temperature depends on the degree that hemoglobin is saturated with oxygen, but as a guideline, the P o 2 is decreased approximately 6% for every 1° C that the patient’s body temperature is below 37° C. P o 2 is increased approximately 6% for every 1° C that the body temperature exceeds 37° C.
Differential Diagnosis of Acid-Base Disturbances
Acid-base disturbances are categorized as respiratory or metabolic acidosis (pH less than 7.35) or alkalosis (pH more than 7.45). These disorders are further stratified into acute versus chronic based on their duration, which is gauged clinically by the patient’s compensatory responses. It must be kept in mind that a patient may have a mixed acid-base disorder. The approach to managing acid-base disorders should first involve searching for the causes, rather than an immediate attempt to normalize the pH. Sometimes the treatment may be more detrimental than the original acid-base problem.
Adverse Responses to Acidemia and Alkalemia
Adverse responses can be associated with severe acidemia or alkalemia. Consequences of severe acidosis can occur regardless of whether the acidosis is of respiratory, metabolic, or mixed origin. Acidemia usually leads to decreased myocardial contractility and release of catecholamines. With mild acidosis, the release of catecholamines mitigates the myocardial depression. Permissive hypercapnia, which is used as a protective lung ventilation strategy for acute respiratory distress syndrome (ARDS) patients, has been quite well tolerated. No significant impact on systemic vascular resistance, pulmonary vascular resistance, cardiac output, or systemic oxygen delivery has been seen. With severe acidemia (pH < 7.2), myocardial responsiveness to catecholamines decreases, so myocardial depression and hypotension predominates ( Fig. 21.5 ). Respiratory acidosis may produce more rapid and profound myocardial dysfunction than metabolic acidosis because of the rapid entry of carbon dioxide into the cardiac cell. In the brain, this rapid increase in carbon dioxide can lead to confusion, loss of consciousness, and seizures. This is probably due to an abrupt decrease of intracellular pH, because chronic increases in carbon dioxide as high as 150 mm Hg are typically well tolerated.
Severe alkalemia (pH > 7.6) can lead to decreased cerebral and coronary blood flow as a result of arteriolar vasoconstriction. The consequences of severe alkalosis are also more prominent with respiratory than with metabolic causes because of the rapid movement of carbon dioxide across cell membranes. Acute hyperventilation can produce confusion, myoclonus, depressed consciousness, and seizures.
Respiratory acidosis occurs when alveolar minute ventilation is inadequate relative to carbon dioxide production ( Box 21.1 ). It can occur with a normal or increased minute ventilation if carbon dioxide production is increased from sepsis or overfeeding or if there is decreased carbon dioxide elimination from ARDS or obstructive lung disease. Decreased carbon dioxide elimination from a decreased minute ventilation can occur with volatile or intravenous anesthetics (see Chapter 8 ), neuromuscular blocking drugs (see Chapter 11 ), or neuromuscular disease. Increased rebreathing or absorption, found with exhausted soda lime, an incompetent one-way valve, or laparoscopic surgery can cause respiratory acidosis.
Increased CO 2 production
Decreased CO 2 elimination
Intrinsic pulmonary disease (pneumonia, ARDS, fibrosis, edema)
Upper airway obstruction (laryngospasm, foreign body, OSA)
Lower airway obstruction (asthma, COPD)
Chest wall restriction (obesity, scoliosis, burns)
CNS depression (anesthetics, opioids, CNS lesions)
Decreased skeletal muscle strength (residual effects of neuromuscular blocking drugs, myopathy, neuropathy)
Increased CO 2 rebreathing or absorption
Exhausted soda lime
Incompetent one-way valve in breathing circuit
ARDS, Acute respiratory distress syndrome; CNS, central nervous system; CO 2 , carbon dioxide; COPD, chronic obstructive pulmonary disease; OSA, obstructive sleep apnea.
Compensatory Responses and Treatment
Over the course of hours to days, the kidneys compensate for the respiratory acidosis by increased hydrogen ion secretion and bicarbonate reabsorption. After a few days, the P co 2 will remain increased, but the pH will be near normal, which is the hallmark of a chronic respiratory acidosis. Respiratory acidosis with a pH less than 7.2 indicates the need for tracheal intubation or increased ventilatory support. In patients with chronic respiratory acidosis, the key is to avoid hyperventilation. The alkalosis from excessive ventilation and relative hypocapnia can result in central nervous system (CNS) irritability and cardiac ischemia. Also, the kidneys will now start to lose bicarbonate. The increased bicarbonate has allowed the patient to maintain a normal pH with a relatively smaller alveolar minute ventilation. Losing the bicarbonate will increase the work of breathing when ventilatory support is decreased, making it difficult to wean from the ventilator.
Respiratory alkalosis occurs when alveolar minute ventilation is increased relative to carbon dioxide production. The increased alveolar minute ventilation can be related to a variety of causes ( Box 21.2 ). Pa co 2 is diminished relative to bicarbonate levels, resulting in a pH more than 7.45. The decreased Pa co 2 and increased pH trigger the peripheral and central chemoreceptors to decrease the stimulus to breathe. During prolonged respiratory alkalosis, active transport of bicarbonate ions out of CSF causes the central chemoreceptors to reset to a lower Pa co 2 level.