Electrolytes and Blood Gases

Mitchell G. Scott, Ph.D., Vicky A. LeGrys, Ph.D., Dr.A., M.T.(A.S.C.P.), C.L.S.(N.C.A.) and Joshua L. Hood, M.D., Ph.D.

Maintenance of water homeostasis is paramount to life for all organisms. In humans, the maintenance of water homeostasis in various body fluid compartments is primarily a function of the four major electrolytes, Na⁺, K⁺, Cl⁻, and . These electrolytes also have a role in acid-base balance and heart and muscle function, and serve as cofactors for enzymes. Virtually no metabolic process is independent or unaffected by electrolytes. Abnormal electrolyte concentrations may be the cause or the consequence of a variety of medical disorders. Because of their physiologic and clinical interrelationships, this chapter discusses determination of (1) electrolytes, (2) osmolality, (3) sweat testing, (4) blood gases and pH, and (5) blood oxygenation.

Electrolytes

Electrolytes may be classified as anions, negatively charged ions that move toward an anode, or cations, positively charged ions that move toward a cathode. Important physiologic electrolytes include Na⁺, K⁺, Ca²⁺, Mg²⁺, Cl⁻, , , , and , and some organic anions, such as lactate. Although amino acids and proteins in solution also carry an electrical charge, they are usually considered separately from electrolytes. Hydrogen ion (H⁺) concentration is routinely measured as pH, but its concentration is so low relative to other ions (10⁻⁹ vs. 10⁻³ mol/L) that for clinical purposes it is not categorized as an electrolyte. The major electrolytes (Na⁺, K⁺, Cl⁻, ) occur primarily as free ions, whereas significant amounts (>40%) of Ca²⁺, Mg²⁺, and trace elements are bound by proteins, mainly albumin. Determination of body fluid concentrations of the four major electrolytes (Na⁺, K⁺, Cl⁻, and ) is commonly referred to as an electrolyte profile.

A typical reference interval for serum Na⁺ is 136 to 145 mmol/L.⁵¹ The central 95% of Na values from more than 16,000 subjects in the National Health and Nutrition Examination Survey III (NHANES III) was 136 to 146 mmol/L.⁵⁴ The interval for premature newborns at 48 hours is 128 to 148 mmol/L, and the value for umbilical cord blood from full-term newborns is ≈127 mmol/L.

Urinary sodium excretion varies with dietary intake, but for an adult male on an average diet containing 7 to 14 g of NaCl per day, an interval of 120 to 240 mmol/d is typical.⁸⁴ A large diurnal variation in Na⁺ excretion has been noted, with the rate of Na⁺ excretion during the night being only 20% of the peak rate during the day. The Na⁺ concentration of cerebrospinal fluid is 136 to 150 mmol/L.⁹⁴ Mean fecal Na⁺ excretion is less than 10 mmol/d.¹⁶

Potassium

Potassium is the major intracellular cation. In tissue cells, its average concentration is 150 mmol/L, and in erythrocytes, the concentration is 105 mmol/L. High intracellular concentrations are maintained by the Na⁺, K⁺ adenosine triphosphate (ATP)ase pump, which is fueled by oxidative energy and continually transports K⁺ into the cell against a concentration gradient. This pump is a critical factor in maintaining and adjusting the ionic gradients on which nerve impulse transmission and contractility of muscle depend. Diffusion of K⁺ out of the cell into the ECF and plasma occurs whenever pump activity is decreased because of (1) depletion of metabolic substrates such as glucose for ATP production; (2) competition for ATP between the pump and other energy-consuming activities of the cell; or (3) slowing of cellular metabolism (as occurs with refrigeration). The importance of these considerations on sample integrity for analysis of K⁺ is discussed later.

The body requirement for K⁺ is satisfied by an average dietary intake of 2.4 to 4.4 g/d (60 to 120 mmol/d). Potassium absorbed from the gastrointestinal tract is rapidly distributed, with a small amount taken up by cells, and most excreted by the kidneys. Potassium filtered through the glomeruli is almost completely reabsorbed in the proximal tubules and is then secreted in the distal tubules in exchange for Na⁺ under the influence of aldosterone. Aldosterone enhances K⁺ secretion and Na⁺ reabsorption in the distal tubules by an Na⁺-K⁺ exchange mechanism. The kidneys respond almost immediately to K⁺ loading with an increase in K⁺ output, so that urine collected during or after a period of high K⁺ intake may have K⁺ concentrations as high as 100 mmol/L. In contrast, the tubular response to conserve K⁺ is very slow in the initial stages of depletion. Unlike the prompt response to conserve Na⁺ in deficit states, it can take up to 1 week for the tubules to reduce K⁺ excretion to 5 to 10 mmol/d from the typical 50 to 100 mmol/d.

Factors that regulate distal tubular secretion of K⁺ include intake of Na⁺ and K⁺, mineralocorticoid concentration, and acid-base balance. Because renal conservation mechanisms are slow to respond, K⁺ depletion can be an early consequence of restricted K⁺ intake or loss of K⁺ by extrarenal routes such as diarrhea. A diminished glomerular filtration rate is typical of renal failure, and the consequent decrease in distal tubular flow rate is an important factor in the retention of K⁺ seen in chronic renal failure. Renal tubular acidosis (RTA) and metabolic and respiratory acidoses and alkaloses also affect renal regulation of K⁺ excretion. These topics are discussed in Chapters 48 and 49.

Specimens

Comments made earlier on specimens for Na⁺ analysis are generally applicable to those for K⁺ analysis. However, some additional points must be made. Potassium concentrations in plasma and whole blood are 0.1 to 0.7 mmol/L lower than those in serum, and most reference intervals for serum K⁺ are 0.2 to 0.5 mmol/L higher than those for plasma K⁺. The extent of this difference depends on the platelet count, because the additional K⁺ in serum is primarily a result of platelet rupture during coagulation.42,72 This variability in the amount of additional K⁺ in serum makes plasma the specimen of choice and emphasizes the necessity of noting on reports whether serum or plasma was assayed and using the appropriate reference interval.

Specimens for determining K⁺ concentrations in serum or plasma must be collected by methods that minimize hemolysis, because release of K⁺ from as few as 0.5% of erythrocytes can increase K⁺ values by 0.5 mmol/L. An increase in K⁺ of 0.6% has been estimated for every 10 mg/dL of plasma hemoglobin caused by hemolysis.¹³ Thus slight hemolysis (Hb ≈50 mg/dL) can be expected to raise K⁺ values by ≈3%, marked hemolysis (Hb ≈200 mg/dL) by 12%, and gross hemolysis (Hb > 500 mg/dL) by as much as 30%. Several correction factors for estimating K⁺ in hemolyzed samples have been examined, but routine use of these is not recommended.⁶⁵ Therefore it is imperative that any visible hemolysis be noted with reported K⁺ values with a comment that results are falsely elevated. If K⁺ concentrations are determined by ISE on whole blood specimens using a blood gas instrument or a point-of-care device, increases in K⁺ concentrations caused by hemolysis will often be overlooked. Whenever hemolysis is suspected, a portion of the specimen should be centrifuged and visually inspected.

Clinically significant preanalytical errors can occur for K⁺ determinations if blood samples are not processed expediently.⁴¹ As mentioned earlier, maintenance of the intracellular-extracellular K⁺ gradient depends on the activity of the energy-dependent Na⁺-K⁺-ATPase. If a whole blood specimen is maintained at 4 °C versus 25 °C before separation, glycolysis is inhibited and the energy-dependent Na⁺-K⁺-ATPase cannot maintain the Na⁺/K⁺ gradient. An increase in plasma K⁺ will occur as a result of K⁺ leakage from erythrocytes and other cells. The increase of K⁺ in serum is on the order of 0.2 mmol/L by 1.5 hours at 25 °C, whereas at 4 °C, the increase is considerably greater and has been reported to be as much as 2 mmol/L after 4 hours at 4 °C.⁷⁴

The opposite effect, namely, a falsely decreased K⁺ value, can be observed if an unseparated sample is stored at 37 °C, because glycolysis occurs and K⁺ shifts intracellularly. Even at room temperature, severe leukocytosis can initially cause falsely decreased K⁺ concentrations. The extent of this decrease depends on leukocyte count, temperature, and glucose concentrations but has been reported to be as much as 0.7 mmol/L at 37 °C.⁵⁵ This effect is, however, biphasic. Initially, plasma K⁺ decreases as a result of glycolysis, but after the glucose substrate is exhausted, K⁺ will leak from cells. When the leukocyte count is greater than 100,000/µL and hypokalemia is present (as in acute myeloid leukemia), glycolysis at room temperature may cause the K⁺ deficit to seem greater than actuality.² In addition to causing pseudohypokalemia,² samples from leukemic patients with very high white blood cell counts (>300 × 10⁹ cells/L) can result in a pseudohyperkalemia due to WBC rupture.¹ Together, with the effect of glycolysis leading to increased cellular uptake of K⁺ followed by K⁺ leakage as a result of exhaustion of glucose, or inhibition of glycolysis by refrigeration, the recommendation for reliable K⁺ determinations is to collect blood with heparin and to maintain it near 25 °C, then to separate the plasma within minutes by high-speed centrifugation without cooling. In practical terms, separation within 1 hour when samples are maintained at room temperature is unlikely to introduce great error.

Finally, skeletal muscle activity causes K⁺ efflux from muscle cells into plasma and can cause a marked elevation in plasma K⁺ values. A common example occurs when an upper arm tourniquet is not released before beginning to draw blood after a patient clenches his fist repeatedly. The plasma K⁺ values can artificially increase as much as 2 mmol/L because of the muscle activity.³⁰

Reference Intervals ⁸⁷

Reported reference intervals for the serum of adults vary from 3.5 to 5.1 mmol/L and from 3.7 to 5.9 for newborns. For plasma, a frequently cited interval is 3.4 to 4.8 mmol/L for adults. The central 95% of K⁺ values from more than 16,000 subjects in the NHANES III were from 3.4 to 4.7 mmol/L.⁵⁴ Cerebrospinal fluid concentrations are ≈70% those of plasma.⁹⁴ Urinary excretion of K⁺ varies with dietary intake, but a typical range observed in an average diet is 42 to 86 mmol/d for males and 33 to 70 mmol/d for females.⁸⁴ Gastric juice contains K⁺ at ≈10 mmol/L. Fecal excretion has been reported as 18.2 ± 2.5 mmol/d, but in severe diarrhea, gastrointestinal loss may be as much as 60 mmol/d.¹⁶

Methods for the Determination of Sodium and Potassium

Although AAS, FES, or spectrophotometric methods have been used for Na⁺ and K⁺ analyses in the past, most laboratories now use ISE methods. For example, in 2011, of the laboratories reporting proficiency data for Na⁺ and K⁺ to the College of American Pathologists (CAP), approximately 90 % were using ISE methods.¹⁵

Flame Emission Spectrophotometry

Although at one time the most common method for Na⁺ and K⁺ analyses, FES is no longer a common laboratory method. Advances in electrochemistry combined with the large number of maintenance and safety procedures required for FES have essentially led to the demise of this method for electrolyte analysis.

Principle.: Samples are diluted in a diluent containing known amounts of lithium (or cesium, if lithium itself is being measured) and are aspirated into a propane-air flame. Sodium, potassium, lithium, and cesium ions, when excited, emit spectra with sharp, bright lines at 589, 768, 671, and 852 nm, respectively. Light emitted from the thermally excited ions is directed through separate interference filters to corresponding photodetectors. The Li⁺ or Cs⁺ emission signal is used as an internal standard against which the Na⁺ and K⁺ signals are compared. Details about reagents, procedures, and laboratory safety issues for flame photometry can be found in the third edition of this text.⁸⁹

Ion-Selective Electrodes ⁶³

Analyzers fitted with ISEs usually contain Na⁺ electrodes with glass membranes and K⁺ electrodes with liquid ion-exchange membranes that incorporate valinomycin. Simply stated, potentiometry is the determination of change in electromotive force (E, potential) in a circuit between a measurement electrode (the ISE) and a reference electrode, as the selected ion interacts with the membrane of the ISE. In instrument applications, the measuring system is calibrated by the introduction of calibrator solutions containing defined amounts of Na⁺ and K⁺. The potentials of the calibrators are determined, and the ΔE/Δ log concentration responses are stored in microprocessor memory as a comparison for calculating unknown concentration when E of the unknown is measured. It is important to note that the response of potentiometric electrodes to analytes is a complex process that depends on the composition and thermodynamic and kinetic properties of the sensor membrane, bathing solution, and interface zone between membrane and analyte and between membrane and bathing solution.¹¹ For simplicity, we classically describe E as the sum of the boundary potential (EPB1) at the (sample/ion-sensitive film boundary) and EPB2 at the (membrane/internal contact) boundary, and by the diffusion potential (ED) inside the membrane itself. A constant, C, is added to account for potential at the internal sensor/contact interface.¹¹ Thus, we have the equation, E = EPB1 + ED + EPB2 + C. For simplicity, we typically assume that C and ED are zero. Thus, E = EPB1 + EPB2. This approximation is appropriate for most modern day electrodes, but the successful development of future solid state ion-selective electrodes (SS-ISEs) containing conducting polymers will require a newer approach to the complexity of E determinations. SS-ISEs are based on the covalent binding of ion recognition sites (metal complexing ligands) to a partially oxidized conductive polymer membrane, such as polypyrrole, where pyrrole nitrogens act as electron donors.⁶⁷ Thus, ion-selective integration sites are directly coupled with the ion electron transducer (conductive polymer) on the electrode, eliminating internal filling solutions or gels.¹⁰ Development of SS-ISEs will allow for less electrode maintenance and electrode durability and miniaturization.¹¹ This affords the possibly of constructing microscale or nanoscale sensors.

Frequent calibration, initiated by the user or by microprocessor-controlled uptake of sample from a reservoir of the calibrator, is typical of most current ISE systems. Some instruments are designed to measure Na⁺ and K⁺ in whole blood, particularly point-of-care testing (POCT) devices and many blood gas analyzers.

Two types of ISE methods are in use and must be distinguished. With indirect ISE methods, the sample is introduced into the measurement chamber after mixing with a rather large volume of diluent. Indirect ISE methods are the methods used most commonly on today’s automated, high-throughput clinical chemistry systems. Indirect methods were developed early in the history of ISE technology, when dilution was necessary to present a small sample in a volume large enough to adequately cover a large electrode and to minimize the concentration of protein at the electrode surface. With direct ISE methods, the sample is presented to the electrodes without dilution. This approach became possible with the miniaturization of electrodes. Direct ISEs are most common in blood gas analyzers and point-of-care devices where whole blood is directly presented to the electrodes. Single-use, thin-film ISEs for Na⁺, K⁺, and Cl⁻ are unique applications of a direct ISE method used by Ortho Vitros analyzers (Ortho Diagnostics, Raritan NJ).²⁵

Errors observed in the use of ISEs fall into three categories. First are errors caused by lack of selectivity. For instance, many Cl⁻ electrodes lack selectivity against other halide ions. Most chloride selective membranes use a quaternary ammonium chloride ion exchanger.⁷³ Any ion that has a hydration energy equivalent to or higher than chloride can interfere with chloride selectivity. Examples include (1) iodide, (2) bromide, (3) thiocyanate, (4) bicarbonate, (5) salicylate, and (6) heparin. By far, bicarbonate is the most common interference, and its ionic activity is taken into account when the Nikolsky-Eisenman equation is used,⁷³ when the electrical potential difference of the chloride electrode is calculated. Second are errors introduced by repeated protein coating of the ion-sensitive membranes, or by contamination of the membrane or salt bridge by ions that compete or react with the selected ion and thus alter the electrode response. Such errors in ISE measurements necessitate periodic changes of the membrane as part of routine maintenance. Finally, the electrolyte exclusion effect, which applies only to indirect methods and is caused by the solvent-displacing effect of lipid and protein in the sample, results in falsely decreased values.⁴

Spectrophotometric Methods

Spectrophotometric methods fall into three categories: those based on enzyme activation and those that detect the spectral shift produced when Na⁺ or K⁺ binds to a macrocyclic chromophore. These approaches have been applied to smaller instruments. However, the high cost of reagents for these methods and the fact that few problems exist with ISE methods have resulted in small “niche” use of these methods, primarily with smaller instruments used in physicians’ offices or clinics.⁸³

Kinetic spectrophotometric assays for Na⁺ are based on activation of the enzyme β-galactosidase by Na⁺ to hydrolyze o-nitrophenyl-β-D-galactopyranoside.⁷⁵ The rate of production of o-nitrophenol (the chromophore) is measured at 420 nm.

K⁺-specific enzyme activation assays are illustrated by methods using tryptophanase,⁵⁰ one of several K⁺-enhanced enzymes.

Macrocyclic ionophores are molecules whose atoms are organized to form a cavity into which metal ions fit and bind with high affinity. Different macrocyclics can be made with cavities tailored to fit the ionic radii of different elements. When chromogenic properties are imparted to these ionophores, spectral shifts occur when the cation is bound. The specificity of many of these ionophores appears sufficient for clinical purposes.⁵²

Electrolyte Exclusion Effect ⁴

The electrolyte exclusion effect describes the exclusion of electrolytes from the fraction of the total plasma volume that is occupied by solids. The volume of total solids (primarily protein and lipid) in an aliquot of plasma is approximately 7%, so that ≈93% of plasma volume is actually water. The main electrolytes (Na⁺, K⁺, Cl⁻, ) are confined to the water phase. When a fixed volume of total plasma (e.g., 10 µL) is pipetted for dilution before flame photometry or indirect ISE analysis, only 9.3 µL of plasma water that contains the electrolytes is added to the diluent. Thus a concentration of Na⁺ determined by flame photometry or indirect ISE to be 140 mmol/L is the concentration in the total plasma volume, not in the plasma water volume. In fact, if the plasma contains 93% water, the concentration of Na⁺ in plasma water is [140 × (100/93)], or 150 mmol/L. This negative “error” in plasma electrolyte analysis has been recognized for many years.³ Even though it is the electrolyte concentration in plasma water that is physiologic (the Na⁺ concentration of normal saline is indeed 150 mmol/L), it was assumed that the volume fraction of water in plasma is sufficiently constant that this difference could be ignored. In fact, all electrolyte reference intervals are based on this assumption and actually reflect concentrations in total plasma volume and not in water volume. Indeed, virtually all concentrations measured in the clinical chemistry laboratory are related to the total sample volume rather than to the water volume. This electrolyte exclusion effect becomes problematic when pathophysiologic conditions are present that alter the plasma water volume, such as hyperlipidemia or hyperproteinemia. In these settings, falsely low electrolyte values are obtained whenever samples are diluted before analysis, as in flame photometry or with indirect ISE methods⁴ (Figure 28-1).

Figure 28-1 Predicted influence of water content on sodium measurements for a 100 mmol/L NaCl solution by direct ion-selective electrode (ISE) versus flame emission photometry or indirect ISE. Red areas represent nonaqueous volumes, which could consist of lipids, proteins, or even a slurry of latex or sand particles. (Adapted from Apple FS, Koch DD, Graves S, Ladenson JH. Relationship between direct-potentiometric and flame-photometric measurement of sodium in blood. Clin Chem 1982;28:1931-5.)

Indirect ISE methods dilute the sample in a diluent of fixed high ionic strength so that for Na⁺, the activity coefficient approaches a value of 1. Under these circumstances, the measurement of activity, (a), where a = γ (concentration), and γ is the activity coefficient, is tantamount to measurement of concentration. It is the dilution of total plasma volume and the assumption that plasma water volume is constant that render both indirect ISE and flame photometry methods equally subject to the electrolyte exclusion effect. In certain settings, such as ketoacidosis with severe hyperlipidemia ³⁷ or multiple myeloma with severe hyperproteinemia,⁵⁶ the negative exclusion effect may be so large that laboratory results lead clinicians to believe that electrolyte concentrations are normal or low when, in fact, the concentration in the water phase may be high or normal, respectively. In severe hypoproteinemia, the effect works in reverse, resulting in falsely high (2 to 4%) Na⁺ or K⁺ values. Plasma sodium, potassium, and chloride measurements by an indirect ISE were found to be affected by changes in plasma protein concentration when low plasma protein concentrations lead to an observed “pseudohyper” effect, and high plasma concentrations result in a “pseudohypo” effect.²⁹ The relationship was found to be nonlinear with no ability to calculate an accurate predictive value between changes in plasma protein and electrolyte concentration.

Direct ISE methods still determine the concentration relative to activity but do not require sample dilution. Because there is no dilution, activity is directly proportional to the concentration in the water phase, not the concentration in the total volume. To make results from direct ISEs equivalent to those from flame photometry and indirect ISEs, most direct ISE methods actually operate in what is commonly referred to as the “flame mode.” In this mode, the directly measured concentration in plasma water is multiplied by the average water volume fraction of plasma (0.93). Although the latter may vary widely, as long as the activity of the specific ion is constant, the concentration of the ion in the water phase becomes independent of the relative proportions of water and total solids if the ion is not bound by proteins. Therefore direct ISE methods are free of electrolyte exclusion effects, and the values determined by direct ISE methods—even in the flame mode—are directly proportional to activity in the water phase and define electrolyte concentrations in a more physiologic and physicochemical sense.

Most clinical chemists and physicians have reached the conclusion that direct ISE methods for electrolyte analysis are the methods of choice. They base their conclusion on the fact that great changes in plasma lipid or protein concentration can be expected in relatively common clinical conditions and in therapies such as parenteral alimentation with lipid emulsions. However, it is clear that results from direct methods will continue to be converted to total plasma volume concentrations by use of the “flame mode,” and indeed this is the recommendation of the Clinical Laboratory and Standard Institute (CLSI). This is also a good recommendation in that two thirds of laboratories still use indirect ISE methods.¹⁵ Tables 28-1 and 28-2 summarize methods that are and are not subject to electrolyte exclusion effects, respectively. One approach to improve the physiologic accuracy of electrolyte values from methods subject to the electrolyte exclusion effect consists of centrifugation (100,000 × g) and analysis of the chylomicron-poor infranatant.³ In situations in which both lipid and protein contents are altered, presenting plasma electrolyte values along with concurrent estimates of plasma water have been suggested. One approach to estimate plasma water (f) is Waugh’s empirical equation⁹⁵:

TABLE 28-1

Methods Measuring Concentration in the Whole Sample Volume and Thus Subject to Electrolyte Exclusion Effect

Method	Analytes
Flame photometry	Na⁺, K⁺, Li⁺
Atomic absorption spectrometry	Ca²⁺, Mg²⁺, and others
Amperometry/coulometry	Cl⁻
Indirect potentiometry	Na⁺, K⁺, Ca²⁺, Cl⁻

TABLE 28-2

Methods Measuring Activity, Molality, or Concentration in the Water Phase and Thus Not Subject to Electrolyte Exclusion Effect

ISE, Ion-selective electrode.

where P_s is serum total protein and L_s is serum total lipid, both in grams per liter. Alternatively, the best solution is to use a direct ISE method.

Chloride

Chloride is the major extracellular anion. Therefore chloride, similar to Na⁺, is significantly involved in the maintenance of water distribution, osmotic pressure, and anion-cation balance in the ECF. In contrast to its high ECF concentrations (≈103 mmol/L), the concentration of Cl⁻ in the intracellular fluid of erythrocytes is 45 to 54 mmol/L; in the intracellular fluid of most other tissue cells it is only ≈1 mmol/L. In gastric and intestinal secretions, Cl⁻ is the most abundant anion.

Chloride ions are almost completely absorbed from the intestinal tract. They are filtered from plasma at the glomeruli and are passively reabsorbed, along with Na⁺, in the proximal tubules. In the thick ascending limb of the loop of Henle, Cl⁻ is actively reabsorbed by the chloride pump, which promotes passive reabsorption of Na⁺. Loop diuretics such as furosemide and ethacrynic acid inhibit the chloride pump.

Methods for Determination of Chloride in Body Fluids

Chloride is determined by (1) mercurimetric titration, (2) spectrophotometry, (3) coulometric-amperometric titration, or, most commonly today, (4) ISE.

Specimens

Chloride most often is measured in serum or plasma, urine, and sweat. Cl⁻ is stable in serum and plasma. Even gross hemolysis does not significantly alter serum or plasma Cl⁻ concentration because the erythrocyte concentration of Cl⁻ is approximately half of that in plasma. Because very little Cl⁻ is protein bound, change in posture or stasis, or the use of tourniquets, has little effect on its plasma concentration. Measurement of Cl⁻ loss in gastric aspirates or intestinal drainages is an adjunct to parenteral replacement therapy. Fecal Cl⁻ determination may be useful for the diagnosis of congenital hypochloremic alkalosis with hyperchloridorrhea (increased excretion of Cl⁻ in stool). In this condition, the concentration of Cl⁻ in feces may reach 180 mmol/L, with undetectable Cl⁻ in urine.

Mercurimetric Titration

One of the earliest methods for determining Cl⁻ in biological fluids is mentioned for historical purposes. A protein-free filtrate of specimen is titrated with mercuric nitrate solution in the presence of diphenylcarbazone as an indicator. Free Hg²⁺ combines with Cl⁻ to form soluble but essentially nonionized mercuric chloride (HgCl₂). Excess Hg²⁺ reacts with diphenylcarbazone to form a blue-violet color complex ⁹⁰

Spectrophotometric Methods

Spectrophotometric methods based on the reaction of Cl⁻ with mercuric thiocyanate were common on many automated analyzers in the 1970s and 1980s. Chloride ions react with undissociated mercuric thiocyanate to form undissociated mercuric chloride and free thiocyanate ions. In the presence of perchloric acid, the thiocyanate ions react with ferric ion (Fe³⁺) to form the highly colored, reddish complex of ferric thiocyanate [Fe(SCN)₃] with an absorption peak at 480 nm. High concentrations of globulins in the serum interfere in these methods through turbidity.

Mercurimetric automated methods applied to high-volume testing presented the problem of disposal of reagent waste containing a significant amount of toxic mercury. Consequently, the method is no longer in use.

Coulometric-Amperometric Titration

Reactions in coulometric-amperometric determinations of Cl⁻ depend on the generation of Ag⁺ from a silver electrode at a constant rate and on the reaction of Ag⁺ with Cl⁻ in the sample to form insoluble silver chloride (AgCl)²⁶:

After the stoichiometric point is reached, excess Ag⁺ in the mixture triggers shutdown of the Ag⁺ generation system. A timing device records elapsed time between the start and stop of Ag⁺ generation. Because the time interval is proportional to the amount of Cl⁻ in the sample, the concentration of Cl⁻ can be calculated.

Applications of the coulometric-amperometric principle (often called the Cotlove chloridometer technique)²⁶ are the most precise methods for measuring Cl⁻ over the entire range of concentrations found in body fluids. This method is subject to interferences by other halide ions, by CN⁻ and SCN⁻ ions, by sulfhydryl groups, and by heavy metal contamination. Maintenance of the systems is crucial for proper operation. Today, less than 1% of ≈5149 laboratories report Cl⁻ results by using coulometry.¹⁵ However, some laboratories maintain these instruments as backups and for sweat analysis.

Ion-Selective Electrode Methods

Solvent polymeric membranes that incorporate quaternary ammonium salt anion exchangers, such as tri-n-octylpro-pylammonium chloride decanol, are used to construct Cl⁻-selective electrodes in clinical analyzers.⁷³ Although they are by far the most common methods for measuring Cl⁻ in clinical laboratories, these electrodes have been described to suffer from membrane instability and lot-to-lot inconsistency in terms of selectivity to other anions.^73,93 Anions that tend to be problematic include other halides and organic anions, such as SCN⁻, which can be particularly problematic because of their ability to solubilize in the polymeric organic membrane of these electrodes.

Approximately 82% of ≈5149 laboratories reporting in a 2008 CAP proficiency test survey for Cl⁻ used indirect ISE methods.¹⁵

Reference Intervals 51,91

Reported reference intervals for Cl⁻ in the serum or plasma vary from 98 to 107 mmol/L to 100 to 108 mmol/L. The central 95% of Cl⁻ values from more than 16,000 subjects in NHANES III was 98 to 111 mmol/L.⁵⁴ For neonates, the upper limit of the interval extends to 113 mmol/L. Serum values vary little during the day. Spinal fluid Cl⁻ concentrations are ≈15% higher than those in serum.⁹⁴ Urinary excretion of Cl⁻ varies with dietary intake, but an interval of 110 to 250 mmol/d is typical. Fecal excretion of Cl⁻ (for eight healthy subjects) has been reported as 3.2 ± 0.7 mmol/d (SEM).¹⁶

Bicarbonate (Total Carbon Dioxide)

Total carbon dioxide is used here to describe the quantity that is measured most often in automated clinical chemistry analyzers by acidification of a serum or plasma sample and measurement of carbon dioxide released by the process, or by alkalinization and measurement of total bicarbonate. Under certain conditions of collection and specimen handling, total carbon dioxide values determined in this manner will be almost identical to values for the calculated concentration of total carbon dioxide obtained in blood gas analysis (see later section in this chapter on blood gas methods).

Specimens

The same sample types used for Na⁺ or K⁺ may be assayed. Given a specimen in a vacuum-draw tube, the concentration of total CO₂ is most accurately determined when the assay is done as promptly as possible after collection and centrifugation of the blood in the unopened tube. Ambient air contains far less CO₂ than does plasma, and gaseous dissolved CO₂ will escape from the specimen into the air, with a consequent decrease in the CO₂ value of up to 4 to 5 mmol/L in the course of 1 hour.⁸⁸ In practical terms, the logistics of high-volume processing and automated analysis of specimens almost ensures that most CO₂ measurements are done on specimens that have lost some dissolved gaseous CO₂, simply because preservation of anaerobic conditions is not practical between the time plasma is placed on an instrument and the time it is sampled. Thus the term bicarbonate may be preferable to total CO₂. On the other hand, a sample that is rapidly processed and promptly analyzed has a much smaller error.

Methods for Determination of Serum or Plasma Total Carbon Dioxide

One of the earliest methods for determining total CO₂ was the manometric method for total CO₂ content, using the Natelson microgasometer. This has been supplanted in clinical laboratories by automated methods. This method is described in some detail in an earlier edition of this text.⁹⁰

The first step in automated methods is acidification or alkalinization of the sample. Acidifying the sample converts the various forms of CO₂ in plasma to gaseous CO₂ by dilution with an acid buffer. Alkalinizing the sample converts all CO₂ and carbonic acid to . Methods for total CO₂ measurement with today’s automated instruments may be electrode based or enzymatic. In indirect electrode-based methods, the amount of released gaseous CO₂ after acidification is determined by a PCO₂ electrode in the reaction chamber of the CO₂ module. About 32% of laboratories reporting CAP data used an indirect ISE method in 2008.¹⁵ Direct ISE methods for total CO₂ are no longer common on automated analyzers. Direct methods for total CO₂ had problems with specificity and are no longer in use. For instance, one direct total CO₂ electrode reacted almost equivalently with nitrate.²⁷

In enzymatic methods for CO₂, the specimen is first alkalinized to convert all CO₂ and carbonic acid to . The enzymatic reactions are as follows:

Decreased absorbance of NADH at 340 nm is proportional to the total CO₂ content.

Reference Intervals ⁵¹

The reference interval for total carbon dioxide in adults is 22 to 28 mmol/L but can be instrument dependent. The central 95% of CO₂ values from more than 16,000 subjects in NHANES III was 21 to 35 mmol/L.⁵⁴

Principles of Osmotic Pressure and Osmosis

Osmometry is a technique for measuring the concentration of solute particles that contribute to the osmotic pressure of a solution. Osmotic pressure governs the movement of solvent (water in biological systems) across membranes that separate two solutions. Different membranes vary in pore size and thus in their ability to select molecules of different size and shape. Examples of biologically important selective membranes are those enclosing the glomeruli and capillary vessels that are permeable to water and to essentially all small molecules and ions, but not to large protein molecules. Differences in the concentrations of osmotically active molecules that cannot cross a membrane cause those molecules that can cross the membrane to move to establish an osmotic equilibrium. This movement of solute and permeable ions exerts what is known as osmotic pressure.

As an example, consider an aqueous solution of sucrose placed within a sac made up of a membrane permeable only to water, with an open vertical glass tube (a crude manometer) attached to the sac. If the sac is placed into a beaker of distilled water, water will move from the beaker across the membrane into the sucrose solution. The pressure of this solvent movement will cause the sucrose solution to rise up the tube. At equilibrium, the gravitational pressure of the column of solution in the tube equals the osmotic pressure and prevents further net movement of water from the beaker. The height of the rise of the sucrose solution in the manometer tube is a measure of the osmotic pressure of the sucrose solution. This is the pressure that would have to be exerted on the sucrose side of the membrane to prevent the flow of water across the membrane.

Osmosis is the process that constitutes the movement of solvent across a membrane in response to differences in osmotic pressure across the two sides of the membrane. Water migrates across the membrane toward the side containing more concentrated solute.

If the sucrose solution in the aforementioned membrane sac were replaced with a sodium chloride solution of the same molarity, the solution in the manometer would reach equilibrium at a point almost twice as high as that observed with sucrose, because sodium chloride dissociates into two ions per molecule. If ion activity is unrestricted, the sodium chloride solution would have twice as many osmotically active particles (osmoles) for the same molecular concentration as the sucrose solution. In reality, the number of active particles is less than this (0.93 for NaCl), as explained later in this chapter. The total number of individual (solute) particles present in a solution per given mass of solvent, regardless of their molecular nature (i.e., nonelectrolyte, ion, or colloid), determines the total osmotic pressure of the solution. In blood plasma, for example, nonelectrolytes such as glucose and urea and, to a much lesser extent, proteins contribute to the osmotic pressure.

Colligative Properties

In addition to increasing osmotic pressure when the solute is added to the solvent, the vapor pressure of the solution is lowered below that of the pure solvent. As a result of the change in vapor pressure, the boiling point of the solution is raised above and the freezing point of the solution is lowered below that of the pure solvent.

These four properties of solutions—(1) increased osmotic pressure, (2) lowered vapor pressure, (3) increased boiling point, and (4) decreased freezing point—are called colligative properties. All are directly related to the total number of solute particles per mass of solvent. For instance, a 1-molal solution in water boils at a temperature 0.52 °C higher and freezes at a temperature 1.858 °C lower than pure water. The vapor pressure of this solution is 0.3 mm Hg lower than the vapor pressure of pure water, which is 23.8 mm Hg at 25 °C. The osmotic pressure of the same solution is increased from zero to 17,000 mm Hg (22.4 atmospheres). The term osmolality expresses concentrations relative to mass of the solvent (1 osmolal solution is defined to contain 1 Osmol/kg H₂O), whereas the term osmolarity expresses concentrations per volume of solution (1 osmolar solution is defined to contain 1 Osmol/L solution). Osmolality (Osmol/kg H₂O) is a thermodynamically more exact expression because solution concentrations expressed on a weight basis are temperature independent, whereas those based on volume vary with temperature. Although the term osmolarity is often used in the medical literature, osmolality is what the clinical laboratory measures.

An electrolyte in solution dissociates into two (in the case of NaCl) or three (in the case of CaCl₂) particles; therefore the colligative effects of such solutions are multiplied by the number of dissociated ions formed per molecule. However, because of incomplete electrolyte dissociation and associations between solute and solvent molecules, many solutions do not behave in the ideal case, and a 1-molal solution may give an osmotic pressure lower than theoretically expected. The osmotic activity coefficient is a factor used to correct for deviation from the “ideal” behavior of the system:

where Φ = osmotic coefficient, n = number of particles into which each molecule in the solution potentially dissociates, and C = molality in mol/kg H₂O. A table of osmotic coefficients of most solutes of biological interest has been compiled.⁹⁹

Glucose has an osmotic coefficient of 1.00, whereas the Φ for sodium chloride is 0.93 at the concentrations found in serum—thus the derivation of 1.86 × Na⁺ (mmol) in the formula to calculate plasma osmolality (NaCl potentially contributes two osmotically active particles times 0.93 = 1.86). Ethanol has an osmotic coefficient of 0.83.³⁸ The total osmolality or osmotic pressure of a solution is equal to the sum of the osmotic pressures or osmolalities of all solute species present. The electrolytes Na⁺, Cl⁻, and , which are present in relatively high concentrations, make the greatest contributions to serum osmolality. Nonelectrolytes such as glucose and urea, which are present normally at lower molal concentrations, contribute less, and serum proteins contribute less than 0.5% of the total serum osmolality because even the most abundant protein is present at millimolar concentrations.