Abnormalities in acid-base balance should be anticipated in all critically ill children due to their underlying disease or intensive care unit therapy and should be monitored closely.
In severe circulatory failure, such as in cardiac arrest or septic shock, differences between arterial and venous pH, partial pressure of carbon dioxide, and lactate commonly occur. In the presence of tissue hypoperfusion, hypercapnia and acidemia are best detected in central venous blood.
Many available tools, equations, and corrections are available to assist in the interpretation of acid-base disorder. Tools such as Henderson-Hasselbalch, bicarbonate, and standard base excess are the first line for assessing metabolic and respiratory acid-base disturbances. For more complex situations, apparent strong ion difference, urinary anion gap, and corrected anion gap may provide additional information.
Acidosis is the most common derangement of acid-base homeostasis seen in the intensive care unit. Metabolic acidosis can represent a more severe disease presentation and portends a worse outcome.
The most common causes of metabolic acidosis in children with critical illness are sepsis-induced lactic acidosis, diabetic ketoacidosis, and renal insufficiency. The most common causes of respiratory acidosis are severe status asthmaticus and acute respiratory distress syndrome.
Initial management of acid-base derangements in children with critical illness begins with stabilization of the patient, irrespective of the underlying cause of the acid-base disorder. Depending on the severity of the derangement, the underlying cause may need to be aggressively identified and reversed.
Identification of the cause of the underlying disorder often comes after stabilization and treatment. However, identification is essential in preventing worsening of the abnormality, treatment of the underlying disturbance, and may provide information to help determine prognosis.
Acid-base disorders are common in critical illness. Interpretation of these disorders requires an understanding of not only acids and bases but also knowledge of gas exchange and ventilation, dynamics of water and electrolyte movement at the cellular level, and renal-related mechanisms of hydrogen ion, electrolyte, and water excretion. In addition, to fully integrate the physiology with clinically relevant pathophysiology, one must also understand the possible dysfunctions, both structural and metabolic, that contribute to acid or base disturbances in the extracellular or intracellular spaces, such as sepsis, acute kidney injury, severe metabolic dysregulation (such as diabetic ketoacidosis [DKA] or inborn errors of metabolism), or toxic effects of certain compounds and medications.
Most acid-base disturbances do not directly cause harm and are self-limited after management of the underlying clinical condition. Despite this, they are commonly considered a disease process, with significant time spent treating abnormal numbers in the intensive care unit (ICU). A majority of physicians experience difficulty in interpreting various acid-base disturbances. , Given this, it is not a surprise to find easily available smartphone applications, podcasts, blood-gas calculators and interpreters, mnemonics, videos, tutorials, and primers for the interpretation of blood gases and acid-base status, with a variable level of quality and accuracy. Contrary to their original purpose, some of these tools may provide easy interpretation and solutions to acid-base disorders without explaining the physiology. , , As disorders of acid-base homeostasis are common in critical illness, a thorough and practical understanding of the physiology and pathophysiology behind these disorders is crucial in the ICU.
Overview of acid-base physiology
Defining acids and bases
Understanding of acid-base physiology was transformed at the end of the 19th century. Prior to this, Svante Arrhenius had outlined the theories of dissociation and ionization, and Ostwald described the ability of water to dissociate into hydrogen (H + ) and hydroxide or hydroxyl (OH − ) ions. Acid was defined by Arrhenius as any substance that increased the concentration of hydrogen ions when dissolved in water, and a base decreased the concentration of H + ions. As water can itself be a source of hydrogen ions, it is not mandatory for a given substance to have hydrogen atoms. Later, work by Faraday and others showed that acid-base status in biological systems was partly determined by the concentration of electrolytes, most notably sodium and chloride, which had been previously described as a base-forming cation and acid-forming anion, respectively.
By the 1920s, Brønsted and Lowry, working independently, described an acid as a proton donor substance that contains hydrogen atoms and is able to release them. , Thus, an acid (HA) may dissociate to donate a proton to the solution, forming the conjugate base (or anion) in a reversible manner according to the law of mass action:
In Eq. 72.1 , K a (or Keq ) is the equilibrium or dissociation constant, specific for every substance and influenced by the solution in which the reaction is taking place. Based on this principle, a base is defined as a substance with the ability for binding free hydrogen ions. , Examples of such bases include bicarbonate, proteins, and phosphates. The Brønsted-Lowry definition also led to the concept of conjugate acid-base pairs.
In the historical context of body fluid composition and acid-base balance, these two definitions work synergistically to explain the physiology of complex biological solutions. The Arrhenius definitions recognize the principal role of sodium, chloride, and other strong ions, while Brønsted-Lowry’s definition highlights the role of the conjugate acid-base pairs, such as the bicarbonate/carbonic acid system. This definition gradually became the most popular among acid-base physiologists and the Arrhenius definitions became outdated, as many thought that this approach to acid-base balance did not include the central and direct role of hydrogen ions.
Role of water
As water supplies an inexhaustible reservoir of hydrogen (and hydroxyl) ions, acid-base balance must account for the properties of water. The hydrogen ion, H + , immediately protonates another water molecule to form H 3 O + , also known as the hydronium ion (or hydroxonium ion):
Therefore, acid-base abnormalities can also be regarded as an alteration of water dissociation. For example, the presence of electrolytes, carbon dioxide (CO 2 ), and other weak acids in plasma produces powerful electrochemical forces that influence the disassociation of water. , Human plasma is an aqueous solution that must comply with certain physiologic principles, including electrical neutrality, mass action (which explains the dissociation equilibrium of partially dissociated substances), and the constancy of the ionic product for water (which explains that ionization of plasma water varies according to the plasma ionic composition). , ,
Principles of electroneutrality
In aqueous solutions, electrical neutrality must always be maintained—the sum of all positively charged ions (cations) must be equal to the sum of all negatively charged ions (anions). To maintain electroneutrality, the kidneys produce urine with varying concentration of ions depending on body requirements. , , Variations in the concentration of plasma ions (either anions or cations, organic acids or electrolytes) may predictably lead to a change in the plasma concentration of hydrogen ions by driving changes and adjustments in water ionization. , In physiologic conditions, only a small percentage of water molecules disassociate into components ( Eq. 72.2 ), with electrical charge and temperature as the main determinants for dissociation. According to the electroneutrality principle, the plasma concentration and activity of hydrogen ions depend on the plasma level of other ions, and variations in the ionic composition lead to a change in plasma pH. , ,
Normally, the extracellular concentration of hydrogen ions is extremely low, in the nanoequivalent (nEq) or nanomole (nmol) range. The precision with which H + is regulated highlights its critical impact on cellular functions. The usual arterial blood [H + ] is about 40 nEq/L (or nmol/L), or 3 million times less than the serum sodium concentration. This value is usually expressed as pH units, which is the negative log 10 of the hydrogen ion concentration. Although the use of pH instead of concentration of H + has been repeatedly challenged, the concept has survived, as it provides a measurement of much more than H + concentration. , A normal pH of 7.4 is equivalent to an H + concentration in the blood of 40 nEq/L, at 37°C. The relationship between pH and serum [H + ] is nonlinear, but it is close to linear over the normal pH range of 7.35 to 7.45 (corresponding to 45–35 nEq/L of [H + ]). , The logarithmic scale of pH can be confusing and can obscure the magnitude of deviations from normal. For example, a change in pH from 7.4 to 7.2, which may appear to be a decrease of only 0.2, is a 60% increase in [H + ].
The term acidosis is used to describe the underlying process that leads to increase in [H + ], whether or not there is a change in blood pH. Alkalosis , in contrast, is the process that tends to produce a decrease in [H + ], with or without changes in blood pH. Acidemia and alkalemia are the corresponding terms for situations with an associated change in blood pH. ,
Acids, bases, and buffers
Hydrogen ion concentration and activity have a strong influence on the function of almost all enzymatic systems of the body. , , Tight regulation of the H + is mandatory for the body because of the high reactivity of H + ions. Given this, H + ions undergo variable ionization with changes in pH. , , Cellular processes, enzymatic action, and transmembrane transport processes are all highly pH sensitive. The body has a relatively large capacity to tolerate dramatic shifts in pH. For example, from a pH of 7.4 (40 nEq/L or nmol/L) to a severe acidosis of pH 6.7 (equivalent to 175 nEq/L or nmol/L), the [H + ] increases over fourfold, a change that could not be tolerated without intact buffering systems. Each acid or base has a characteristic tendency to release or to accept hydrogen ions in an aqueous solution according to the reversible reaction depicted in Eq. 72.3 .
The larger the value of the equilibrium constant (K a ), the greater the tendency of the acid to dissociate and the stronger the acid. , The Henderson-Hasselbalch equation is the logarithmic transformation of the equilibrium constant equation, in which pH is the −log of [H + ] and pK a is the −log of K a :
The pK a estimates the relative strength of an acid or base. Strong acids have a low pK a (high K a ), and weak acids have high pK a (low K a ). An ideal buffer is a weak acid in equilibrium with its corresponding weak base, particularly when the pK a is close to physiologic pH. , ,
Buffers are a key defense mechanism against changes in systemic pH. There are two main physiologic buffering systems: the bicarbonate/carbonic acid (HCO 3 − /H 2 CO 3 ) system, which acts in both the extracellular space and inside the erythrocytes; and the nonbicarbonate buffers, which include hemoglobin and oxyhemoglobin, phosphates, and plasma proteins. These are naturally occurring weak acids and bases, which act by converting strong acids or bases to weak acids or bases, while minimizing the change in pH. , Buffers are located in both the extracellular and intracellular fluids and in the bone. The ability of each buffer system to protect against changes in pH is proportional to its pK a and its concentration.
The bicarbonate/carbonic acid system is physiologically the most important buffering system in the blood. It is an open-buffer system given its interconversion with CO 2 , allowing effective buffering as P co 2 can be regulated by changes in alveolar ventilation:
Any increase in [H + ] (or drop in pH) will shift reaction to the left, increasing ventilation to eliminate the rise in CO 2 . This respiratory response begins within minutes, but it may not reach a steady state for 12 to 24 hours. This respiratory response is particularly useful to cellular tissues, as it has a rapid effect on intracellular and extracellular pH. CO 2 crosses cell membranes easily; thus, changes in P co 2 affect intracellular pH rapidly and in a predictable direction. The bicarbonate/carbonic acid buffering system has immense capacity to buffer, as the lungs can eliminate a vast amount of CO 2 per day. Additionally, the kidneys can regenerate bicarbonate, although their compensatory response is slower than that of the lungs.
The second buffering system involves cations such as ammonium and anions such as nonvolatile metabolic acids. These acids are produced by the normal daily catabolic load or during incomplete catabolism of carbohydrates and fat (lactic acidosis, diabetic ketoacidosis, or, in cases with defective processing or excretion of certain metabolites, as occurs in some inborn errors of metabolism). The free hydrogen ions are neutralized by buffers; however, as these acids are not in equilibrium in the normal plasma, they must be metabolized, mainly in the liver, and then excreted by the kidneys. ,
In addition, bone represents an important site of buffering of acid and base loads. While it is difficult to measure the exact contribution of bone buffering, it has been estimated that as much as 40% of the buffering of an acute acid load takes place in bone.
Tools for interpreting acid-base disorders
Bicarbonate is a circulating anion in the plasma that contributes substantially to the negative charge and therefore plays a role in determining plasma pH. Unlike other plasma ions, bicarbonate concentration is partially facilitated by the ventilatory activity of the lung in addition to the kidney’s handling of bicarbonate. CO 2 formed during cell metabolism diffuses across the plasma membrane into blood within the tissue capillary network and then is transported in a number of ways. Most of it (90%–95%) diffuses into the red blood cells, where it is hydrated to bicarbonate (HCO 3 − ) by carbonic anhydrase II, generating hydrogen ions that bind to oxyhemoglobin. As a result, oxygen is released from oxyhemoglobin and leaves the erythrocyte, diffusing into tissues. Approximately 5% of the CO 2 remains as a gas in the aqueous phase of blood, measured as the partial pressure of CO 2 (P co 2 ). An even smaller proportion is bound to plasma proteins.
The Henderson-Hasselbalch equation quantifies the relationship between CO 2 and HCO 3 − . Using this relationship, various tools were subsequently developed to quantify the metabolic component of acid-base disturbances ( eBox 72.1 ). In an attempt to identify acid-base changes that are independent of CO 2 , Hastings and Singer developed the conceptual framework of a buffer base (BB), which is a summation of weak acids in plasma (HCO 3 − and all weak acid buffers). , , , Other groups proposed a multitude of other P co 2 -independent indicators of acid-base disturbances. These approaches can detect metabolic disturbances, mostly the simple ones; however, if P co 2 varies, the interpretation may be more confusing or complex. , One of these indicators is the standard bicarbonate, which is measured plasma HCO 3 − at a Pa co 2 of 40 mm Hg. Base excess or deficit (BE), developed by Siggard-Andersen, is another indicator as a measure of the deviation of BB from its normal value.
pH: Log 10 H + . Measured directly by electrode. Normal = 7.35–7.45.
P co 2 : Partial pressure of gaseous CO 2 . Measured directly by electrode. Usually expressed in mm Hg. Normal value: 35–45 mm Hg (sea level, arterial blood).
HCO 3 −: Bicarbonate concentration. A calculated parameter, derived from pH and P co 2 values using a nomogram or the Henderson-Hasselbalch equation, or equal to the difference between serum total CO 2 (CO 2TOT ) and the dissolved CO 2 . (P co 2 × 0.03). Normal value: 22–28 mEq/L.
Base excess (BE or BE/D): Also known as base excess/deficit. Defined as the amount of strong base (negative base excess, or “base deficit”) or strong acid (positive base excess) in mmol/L that would be needed to restore a pH of 7.4 to a liter of whole blood equilibrated at P co 2 = 40 mm Hg. It is calculated using a nomogram.
Standard base excess (SBE): An improvement of base excess to allow equilibration across the entire extracellular fluid space with the goal of increasing accuracy at variable P co 2 values. The new equation assumes hemoglobin 5 g/dL.
CO 2TOT or CO 2 : Total CO 2 or CO 2 concentration. A serum chemistry measured value. About 95% exists as HCO 3 − and 4%–5% as dissolved gaseous CO 2 . Remaining forms are negligible. Because the difference between CO 2TOT and HCO 3 − is about 1 mEq/L at physiologic pH, for clinical purposes they are taken almost as equivalents.
Anion gap (AG): A calculated value that, taking advantage of the electroneutrality principle (cations = anions), indicates the presence of unmeasured anions (mostly organic acids). AG = (Na + + K + ) − (CI − + HCO 3 − ). Normal values are 16 mEq/L (if K + is included) or 12 mEq/L (without K + , with variations of ± 2–4 mEq/L. This may be influenced by the way the values of some of the parameters are measured; thus, it is always better to consult each institution’s own expected normal AG.
Corrected anion gap (AG corr ): AG is corrected for the patient’s albumin concentration as follows:
AG CORRECTED = AG OBSERVED × 2.5 [normal albumin (g/dL)] − [observed albumin (g/dL)], considering normal albumin from 3.2–4.5 g/dL. If there is a concern for elevated lactate, lactate correction should also be considered, as follows: AG CORRLACT = Albumin-corrected AG − [lactate − (mmol/L)].
Serum osmol gap: The difference between measured and calculated serum osmolality. Serum osmolality can be calculated as follows: S OSM = (2 × Na + ) + (Glu/18) + (BUN/2.8). The “gap” between the measured and calculated values should be within 6; an elevation is suggestive of an unmeasured solute.
Urinary anion gap (uGap): Urinary equivalent of serum AG. uGap estimates unmeasured urinary ions. It has also been designated by some as the urine strong ion difference (uSID), a functionally correct name, as the most important unmeasured cation (Ur + ) is ammonium (NH 4 + ). In physiologically normal conditions, uSID must equal the plasma SID APP , or 38–42 mEq/L.
The Henderson-Hasselbalch equation expresses the relationship of the bicarbonate/carbonic acid buffering system to pH , , , :
Further refinement and understanding that most H 2 CO 3 exists as dissolved CO 2 is represented in the modified Henderson-Hasselbalch equation , , :
where 0.03 is the solubility coefficient for CO 2 in blood, and P co 2 represents the partial pressure of CO 2 in blood.
The primary utility of the Henderson-Hasselbalch equation is that it allows quantification of the severity of the respiratory component in acid-base derangements. When pH is altered as the result of changes in the increases or decreases of P co 2 , the change is respiratory , as the primary alteration is associated with either hypoventilation or hyperventilation. When pH is modified by any other cause (e.g., changes in serum cations or anions or in nonvolatile acids), the alteration is considered metabolic in origin. , , , Therefore, acid-base disorders can be classified using the Henderson-Hasselbalch equation to determine the primary type of acid being increased or decreased. For example, in respiratory acidosis, the increase in the P co 2 quantifies the derangement even when there are mixed disorders. The relationship between P co 2 and HCO 3 − can provide a clinical guide for discovering the metabolic origin of a derangement in simple acidosis cases, mainly those associated with the presence of organic acids. However, this can be complicated by the fact that P co 2 , H 2 CO 3 , and HCO 3 − are all interlinked ( Eq. 72.7 and 72.8 ), so that when bicarbonate increases, P co 2 increases, and vice versa. , , , , , It is also important to note that the Henderson-Hasselbalch equation does not provide information about acids other than CO 2 /H 2 CO 3 , which means that only the HCO 3 − /CO 2 buffering system is assessed and that the influence of other buffering systems and mineral anions and cations is overlooked. , ,
Base excess and standard base excess
BE is defined as the amount of strong acid (or strong base) in molecules per liter that must be added in vitro to a whole blood sample to return the pH of the sample to 7.4 while the P co 2 is kept constant at 40 mm Hg. , , , , , For example, in a “normal” blood sample (pH is 7.4 and P co 2 of 40 mm Hg), the BE is 0 mmol/L. Positive values signify an excess of metabolic bases, negative values an excess of metabolic acids. , In order to make this tool clinically applicable, a nomogram was developed to allow BE calculation by automated blood gas analyzers. Of the available tools, BE is least influenced by changes in Pa co 2 . This means that BE does not change in acute respiratory acidosis or alkalosis. In metabolic acidosis, BE decreases and becomes negative (−BE or base deficit), whereas it becomes positive in metabolic alkalosis (simply called BE or +BE).
However, despite its benefits, there are several limitations of the use of BE. For clinical accuracy and reproducibility, several assumptions were incorporated into a nomogram, including correction factors, adjusted formulas, , , , and an estimate of hemoglobin concentration. The chosen hemoglobin concentration was 50 g/L (5 g/dL) as an estimation of the concentration of hemoglobin as it is diluted throughout the body. , BE calculated this way is called standard base excess (SBE), a parameter reported by modern gas analyzers and used in the classical approach to acid-base balance (see eBox 72.1 ). Additionally, SBE does not provide information about the origin or mechanisms of the metabolic acid-base derangement. Most notably, SBE represents the net effect of all metabolic acid-base abnormalities. In certain situations, coexisting metabolic acidosis and alkaloses may cancel each other; the resultant “normal” SBE will mistakenly suggest that no acid-base derangement exists. Nearly 15% to 18% of critically ill adults with acid-base disorders have normal SBE. , ,
Criticism of the SBE led to the development of alternative tools for assessment of acid-base disturbances, focusing primarily on the role of bicarbonate in determining the appropriateness of the metabolic response. , , , As [H + ] is tightly controlled, when an acid-base derangement develops, several physiologic responses become active over time so that pH changes can be minimized. , Schwartz and Relman concluded that several patterns could be detected and that this approach allowed an improved understanding in how changes in pH and P co 2 modify the concentration of bicarbonate in an acute or chronic respiratory disorder ( eTable 72.1 ). These patterns are sometimes referred to as Winter’s rules or Worthley’s rules depending on the location of clinical practice. , With this approach, primary metabolic disorders can be predicted based on changes in P co 2 in the context of a normal respiratory compensation. These are useful to describe physiologic compensation, for uncovering mixed acid-base disorders, and to differentiate acute from chronic respiratory imbalances (See eTable 72.1 and eBox 72.2 ). ,
Anion gap and corrected anion gap
The anion gap (AG) was first proposed by Emmet and Narins to understand primarily metabolic acidoses. It is founded conceptually on the principle of electroneutrality: in an aqueous solution such as plasma, there can be no net electrical charge. , This means that in order to maintain neutrality, positive-charged cations must equal negative-charged anions , (see eBox 72.1 ):
where [XA − ] equals the unmeasured acid anions (UMAs). Depending on the clinical setting, [K + ] is often omitted and its effect neglected. Therefore, the AG is commonly defined as follows , :
The sum of the difference in charge results in a gap of between 12 and 16 mmol/L or mEq/L. Typically, normal values are 16 mEq/L or mmol/L if [K + ] is included or 12 mEq/L without it, with variations of ±2 to 4 mEq/L. However, this historical reference value is now both lower and narrower due to the use of ion-selective electrode testing, which results in higher chloride values than previous techniques. Several studies have reported “normal” AG values (K + not included) as 6.6 ± 2 mEq/L (range, 2.6–10.6 mEq/L). Because it may be influenced by laboratory measurement techniques, it is necessary to understand each institution’s expected normal AG. , It is affected by any change in concentrations of either unmeasured anions or cations not included in Eq. 72.8A and 72.8B .
Increases in AG are due to the accumulation of unmeasured anions, such as lactate or ketones, which cause a drop in chloride or bicarbonate levels to maintain electroneutrality. This condition is termed anion-gap acidosis or simply gap acidosis. While this approach is good for detecting unmeasured anions as the cause of metabolic acidosis (such as lactic acidosis, ketoacids), it has some limitations. Proteins and phosphate levels are often low in the critical care setting, leading to a decrease in AG. This is particularly true in pediatric patients, in whom hypoalbuminemia may be common. In those with low albumin, the AG can be corrected (AG CORR ) as follows , :
AGCORR=ΔAGOBSERVED+2.5([normal albumin g/dL]-[observed albumin g/dL])