Blood Gas Analysis and Critical Care Medicine

JOHN W. SEVERINGHAUS, POUL ASTRUP, and JOHN F. MURRAY

Departments of Anesthesiology and Medicine, and the Cardiovascular Research Institute, University of California San Francisco, and the San Francisco General Hospital Medical Center, San Francisco, California; and Department of Clinical Chemistry, Rigshospitalet, Copenhagen, Denmark

	INTRODUCTION

TOP INTRODUCTION CONCLUSION REFERENCES

Critical care medicine is one of the newest and most rapidly growing medical specialties. Surprisingly new, in fact, because critical care medicine is, basically, applying physiologic principles to the care of seriously ill patients, something physicians have been trying to do for centuries. Modern critical care medicine is distinguished from its predecessors by incredible products of technology, advances in biochemistry, and astonishing know-how. We now have at our disposal sophisticated monitoring devices that provide moment to moment information about key circulatory and respiratory physiologic variables, how they are deranged by disease, and how they respond to intervention. We also have available an astonishing variety of high-tech instruments and powerful medications that we use to remedy ailing physiology, ventilators for breathing, machines to rid the body of excess fluid and impurities, vasopressor drugs to shore up flagging blood pressure, and even instruments to supplement a failing heart. Another distinguishing feature of critical care medicine is that it is practiced in specialized facilities, intensive care units, within acute care hospitals; these focal points for costly instrumentation are also headquarters for the expertly trained and knowledgeable physicians, nurses, and other professionals who care for desperately ill patients.

This paper retraces the history of the development of knowledge about blood gas transport, including the discovery of oxygen and carbon dioxide, the evolution of techniques to measure respiratory gases in the blood, and finally, how all this came together in Blegdamshospital, Copenhagen, on August 25, 1952, when an ingenious anesthetist, Bjorn Ibsen, came out of the operating room and started the modern critical care movement. We conclude with some comments about the remarkable changes that have occurred during the 45 years between then and now, and we make a few speculations about what the future might have in store.

	BLOOD GAS TRANSPORT

According to Hippocrates (460-377 BC), good health resided in a proper balance among the four humors: blood, phlegm, black bile, and yellow bile, a balance that depended on the generation of life-giving heat within the left ventricle. Aristotle (384-323 BC) concluded that arteries carried air, but Erasistratus of Cos (about 330-250 BC) taught that "pneuma," created within the left ventricle from lung air, was the substance pumped through arteries to the tissues. Galen (130-199 AD) believed that the heart sucked blood-cooling air from the lungs into the left ventricle where the vital heat was generated, that pneuma was transported in arteries to the tissues, hence to veins via anastomoses, and that after arriving back in the heart, blood passed through minute pores in the septum from the right into the left ventricle for replenishment. These ideas went unchallenged by physicians until the 16th century.

Michael Servetus (1511-53) studied and practiced medicine, but his principal interest became theology (1). In Christianismi Restitutio (1553), Servetus contradicted Galen, concluding that the communication between the right and left sides of the heart was "not through the middle wall of the heart . . . but by a very ingenious arrangement the subtle blood is urged forward by a long course through the lungs," the first postulate of the existence of pulmonary capillaries. Severtus sent his book to John Calvin, who considered it heresy, had him arrested, jailed, and burned at the stake within the year of publication.

It remained for William Harvey (1578-1657), a brilliant anatomist and physician, to describe the circuit of blood flow around the body, including its circulation through the lungs. In his monumental De Motu Cordis (1628), Harvey flatly stated that blood was pumped from the right ventricle through the pulmonary circulation to the left ventricle, passing through "the invisible porosities of the lungs and the minute connections of the lung vessels." These theoretic pulmonary porosities became anatomic reality when first seen by the celebrated Italian microscopist Marcello Malpigi (1628-94) (2). Thus, the anatomy of the circulation was concisely described, but the nature of the vital ingredient by which breathing fed the inner life-giving flame remained elusive. It took over 100 years to find it.

Discovery of Carbon Dioxide

Joseph Black (1728-99), who became Professor of Chemistry in Edinburgh, showed while he was a medical student that large quantities of a gas, which he called "fixed air" (carbon dioxide), were generated by heating or acidifying chalk. He was the first to prove that the same gas was present in exhaled air (3).

Discovery of Oxygen

Robert Boyle (1627-91) established the fact that the long-sought, life-sustaining substance was contained within air itself (4). His assistant, Robert Hooke (1635-1703), demonstrated in 1667 that a dog whose exposed lungs had multiple pleural punctures could be kept alive by providing a constant flow of air through the trachea without any movement of the lungs. Hooke showed, as had Richard Lower (1631-91), that arterialization of blood in the lungs occurred through the introduction of fresh air. No one noted that something was taken out of the air and something else was added.

The English Unitarian "dissenting" minister and amateur chemist, Joseph Priestley (1733-1804), who lived next door to a brewery, got interested in the waste gas product of fermentation and started investigating gases. He discovered that the gas given off by heating mercuric oxide caused a much brighter flame than plain air. In 1774, he showed that this gas was essential not only to combustion, but also to respiration and to the greening of plants. Priestley was the first to demonstrate that ordinary air, in which a candle would no longer burn and a mouse no longer live, might regain its former vital properties if green plants were kept within the sealed chamber. He eventually managed to isolate 10 new gases, including nitrous oxide and carbon monoxide, invented carbonated beverages, gum rubber erasers, and refrigeration. In 1791 his Birmingham home was burned and his laboratory trashed by a royalist-sectarian mob incensed by his support of the French revolution. He emigrated with his family to Pennsylvania in 1794. Priestley was one of the great social and political minds of the Enlightenment. He had a significant influence on his good friend Thomas Jefferson, and had his portrait painted (Figure 1) by the most famous American painter of the time, Gilbert Stuart.

View larger version (142K): [in this window] [in a new window]   Figure 1.   Gilbert Stuart's portrait of Joseph Priestley, dissenting Unitarian minister, political activist, and self-taught chemist, who discovered the gas we call oxygen without understanding its role in respiration or combustion.

The Swedish pharmacist, Carl Wilhelm Scheele (1742-86), also discovered the gas we call oxygen about 1772, but delayed publishing his findings until 1777. Neither he nor Priestley understood that their gas combined with fuel in burning or respiration, because they believed in phlogiston as the fiery substance that came out of combustible materials during burning.

Antoine Lavoisier (1743-94), France's greatest chemist (Figure 2), reported to the French Academy on April 14, 1774 that metals like phosphorus and sulfur gained weight when burned by combining with a constituent of air. Later that year, Lavoisier was visited in Paris by Priestley, who described generating his new gas in which a candle could burn with a much brighter flame than usual. Lavoisier then realized that it was Priestley's gas in ordinary air that had combined with his phosphorus and sulfur, and that combined with all fuel when burning takes place. Air, he realized, contained two distinct constituents: one that was respirable, which he called "air éminemment respirable," and another that was nonrespirable (5). In 1777 he realized that Black's "fixed air" must be a compound of coal, and that it was produced both by respiration and by combustion. Together with the mathematician Pierre Simon de Laplace (1749-1827), Lavoisier concluded that the generation of heat in a coal fire was in principle of the same nature as that taking place in the body. Both processes required Priestley's new gas, which Lavoisier now called oxygen, and led to the production of carbon dioxide and water, ultimately yielding the same quantity of heat per unit of oxygen consumed. Lavoisier's tremendous achievements immediately revolutionized chemistry and had a profound influence on medicine and physiology.

View larger version (150K): [in this window] [in a new window]   Figure 2.   Portrait of Antoine Lavoisier, who described the true chemistry of respiration and combustion after Priestley told him about the new gas he had detected in air, which Lavoisier named oxygen. Probably France's greatest chemist, but sent to the guillotine in 1794. (©Photothèque des Musées de la Ville de Paris/Cliché Habouzit.)

Gas Exchange in Lungs and Blood

The first to document the presence of both oxygen and carbon dioxide in blood was (Sir) Humphrey Davy (1778-1829), who published the results of his extraction process in 1799. Thirty-eight years later, in Berlin, Heinrich Gustav Magnus (1802- 70), using quantitative techniques, found more oxygen and less carbon dioxide in arterial blood than in venous blood, and he concluded that carbon dioxide must be formed in or added to the blood during its circulation. He upset the standard idea that heat production occurred in the lungs by showing that blood gas exchange took place within the lungs, whereas the oxidation and generation of body heat occurred elsewhere in the body (6, 7).

Magnus was unable to measure the solubility of oxygen and carbon dioxide in blood because he did not understand that chemical binding was occurring. The discovery of the high affinity of oxygen for hemoglobin at low partial pressures was made in his thesis research by Lothar Meyer (1830-95) (8). Meyer dedicated his dissertation to Professor Carl Ludwig (1816-94), which stimulated Ludwig to investigate blood gas exchange himself. Ludwig eventually concluded that the respiratory gases were actively secreted by the lungs (9), whereas Eduard Pflüger (1828-1910) claimed all exchange could be explained solely by diffusion (10). Their heated debates in the 1870s were sufficiently inconclusive to lead to further studies in Copenhagen by Christian Bohr (1855-1911), a former pupil of Ludwig's. Using improved methods for measuring PO₂ in blood, Bohr was convinced he had shown active pumping of oxygen (11). After spending time working with Bohr, John Scott Haldane (1869-1936) joined the secretionists, and remained convinced, despite later proof to the contrary, until his death. The controversy ended with a brilliant series of seven papers in a single issue of the Scandinavian Archives of Physiology in 1910 by August Krogh (1874-1949), with the help of his wife Marie (1874-1943). With apologies for disproving the secretion theory held by his mentor, Christian Bohr, Krogh proved to everyone except the stubborn Haldane that the mechanism of gas exchange in the lungs was uniquely explained by the physical forces of diffusion (12).

Hemoglobin and Oxyhemoglobin Dissociation

Vincenzo Menghini (1704-59), at the University of Bologna, was the first to show that erythrocytes contained considerable quantities of iron whereas plasma did not. In Stockholm, Jöns Jacob Berzelius (1779-1848) was able to split the red material in blood into a protein called "globin" and a colored component containing iron oxide (13). Johannes Mulder (1802-80), Professor of Chemistry in Utrecht, determined the chemical composition of the pigmented portion, which he named "hematin," and showed that it took up oxygen. In 1862, this red pigment was renamed "hemoglobin" by Felix Hoppe-Seyler (1825-95) after he was able to crystallize it and describe its spectrum (14). He demonstrated that the crystalline form differed from one animal species to another. Using his own newly constructed gas pump, he found that oxygen formed a loose, dissociable compound with hemoglobin, which he called "oxyhemoglobin."

Carl Gustav von Hüfner (1840-1908), who succeeded Hoppe-Seyler as Professor of Physiological Chemistry in Tübingen, reported experimental evidence that 1.34 ml of oxygen combined with 1 g of crystalline hemoglobin; this was precisely the same as his theoretic value based on the iron content that he had also determined (15). The agreement of the two numbers led to much skepticism, but it was later essentially confirmed: the current theoretic value being 1.39 ml/g.

By drawing blood samples from animals exposed to different barometric pressures and determining the oxygen content of the blood, the Frenchman Paul Bert (1833-86) produced the first in vivo relationships between oxygen pressure and oxygen content (16). More detailed descriptions were provided by Bohr, who showed the effect of carbon dioxide on the position of the oxyhemoglobin dissociation curve, known as the "Bohr effect," which he reported in 1904 together with Karl Albert Hasselbalch (1874-1962) and August Krogh (17). The dissociation of oxyhemoglobin was affected by the pH, ionic strength, and temperature of the solution.

In 1910, Archibald Vivian Hill (1886-1977) proposed a simple equation for the dissociation curve, with slope n of about 2.7, S/(1  S) = kPⁿ, where S is saturation and P is PO₂, mm Hg. It fit poorly at low saturation. Hill's equation was modified by John Severinghaus by using two terms, one with n = 3 and one, n = 1: S/(1  S) = k(P³ + 150P). For the standard human dissociation curve at pH = 7.40, T = 37° C, k = 1/23,400. This provided a remarkably accurate standard dissociation curve with maximum error of ± 0.5% saturation from 0 to 100% (18). Its accuracy may relate to the kinetics by which the last three oxygen molecules combine essentially simultaneously, because the second oxygen causes a shape and affinity change.

Hemoglobin has probably been studied more than any other protein. Yet it was not until after World War II that the Nobel Laureates Linus Pauling (1901-94), California Institute of Technology, and Max Perutz, University of Cambridge, working independently, defined the chemical structure of the hemoglobin molecule, explained the binding and release of oxygen, and documented the accompanying molecular conformational changes. Genetic disorders of hemoglobin were found to afflict millions of people and to be the most common single gene disorder of mankind. They are usually caused by the formation of a hemoglobin variant with a single amino acid substitution in either the alpha- or beta-globin chains. Some of the aberrant hemoglobin molecules interfere with oxygen transport, some impair red blood cell survival.

	ACID-BASE BALANCE AND CARBON DIOXIDE

While fermentation and respiration were recognized as producing carbon dioxide in Black's time, the mid-18th century, and acids and bases were identified far earlier, the connection between them was slow to appear. The alkalinity of blood was discovered in Paris by Hilaire Marin Rouelle (1718-79), using titration and color indicators. In 1831, William B. O'Shaughnessy (1809-89), an Irish physician working in London and later in India, showed that cholera reduced the "free alkali" of the blood. Later, Henry Bence Jones (1813-78), a physician at St. George's Hospital in London, recognized the relationship between blood alkalinity and stomach acid secretion. The relationship between the carbon dioxide content of blood and its alkalinity was established in his 1877 thesis by Friedrich Walter (b. 1850), which made it possible to study acidosis and alkalosis by extracting from and quantifying the carbon dioxide in blood. Perhaps this association of carbon dioxide with alkali was responsible for the long delay in understanding its role as carbonic acid when dissolved in water.

In 1907 the remarkable ability of blood to neutralize large amounts of acid led Lawrence J. Henderson (1878-1942), then an instructor in biochemistry at Harvard University, to investigate the relationship of bicarbonate to dissolved carbon dioxide gas, and how they acted as buffers of fixed acids (19). It was his insight that helped chemists and physiologists to realize that when acids are added to blood, the hydrogen ions react with blood bicarbonate, generating carbon dioxide gas, which is then excreted by the lungs, almost eliminating the increased acid. Henderson (20) rewrote the laws of mass action for weak acids and their salts (Figure 3). In the case of carbon dioxide and bicarbonate he defined the dissociation constant k as the hydrogen ion concentration at which half of the carbonic acid is dissociated: k = [H+][HCO₃]/[H₂CO₃]. Assuming that all dissolved carbon dioxide was carbonic acid, the denominator became SPCO₂, where S is the solubility of carbon dioxide in mM/mm Hg. Following this lead, in 1917 Hasselbalch adapted Henderson's mass law for carbonic acid to the logarithmic form known as the Henderson-Hasselbalch equation (21), a staple of contemporary clinical acid-base analysis: pH = pK' + log[HCO₃/SPCO₂].

View larger version (124K): [in this window] [in a new window]   Figure 3.   Photograph of Lawrence J. Henderson, biochemist at Harvard University, who was first to understand and express quantitatively the buffering effect of carbon dioxide and bicarbonate interacting with hydrogen ions in blood.

Hydrogen Ions

Physical chemistry began as a discipline about 1884, the year Jacobus van't Hoff (1852-1911), a student of the thermodynamic theories of Josiah W. Gibbs (1839-1903) and Henri Louis Le Chatelier (1850-1936), realized that the osmotic pressure generated by molecules (or later, ions) in solution was exactly the same as they would exert at the same concentration in a gas, thus linking solution theory to the long established laws describing the behavior of gases. New understanding of electrolyte solutions was provided by Svante Arrhenius (1859-1927), who used conductivity in his thesis research to infer the existence of ionization of salts as their concentration was reduced (22). The Arrhenius discovery stimulated Wilhelm Ostwald (1853-1932) to make the first electrometric measurement of hydrogen ion concentration by the potential on a platinum electrode in solutions saturated with hydrogen gas (23). He discovered that this potential was a logarithmic function of the strength of the acid. Ostwald's student, Hermann Nernst (1864-1941), discovered the energetic equivalence of Faraday's constant F to PV/n of the gas laws, thereby mathematically linking electrometric ion activity to the behavior of gases (24). After Nernst moved to Göttingen, his assistant Heinrich Danneel (1867-1942) discovered the reaction of oxygen with a negatively charged metal (cathode), the basis of oxygen polarography, later developed by Jaroslav Heyrovsky (1890-1967) in Prague (25). Nobel prizes in chemistry were awarded to van't Hoff (1901), Arrhenius (1903), Ostwald (1909), Nernst (1920), and Heyrovsky (1959).

Use of pH for Hydrogen Ion Activity

The credit (or blame) for introducing the term pH, the negative log of hydrogen ion (H+) concentration, goes to S. P. L. Sørensen (1868-1939), who apparently tired of writing seven zeros in a paper on enzyme activity and wanted a simpler designation (26). Although the use of pH instead of nanomoles of H+ has been repeatedly challenged, pH has survived in large part because the behavior of a substance in a chemical system is proportional to its energy (chemical potential), and this, in turn, is a logarithmic function of the activity of the substance. A pH electrode responds to the chemical potential of H+, and thus the instrument provides a precise and readily obtained measurement of the chemical behavior of H+ in the system, exactly what the chemist, physiologist, and clinician need to know. The pH of blood, and of neutral water, changes linearly with temperature, whereas H+ concentration is a log function of temperature.

pH Electrode

In 1906, Max Cremer (1865-1935) discovered an electrical potential proportional to the acid concentration difference across thin glass membranes. By 1909, Fritz Haber (1868- 1934) and Zygamunt Klemensiewicz (1886-?) had constructed and studied glass H+ electrodes. A modified Ostwald platinum electrode was used by Hasselbalch in 1912 to measure blood pH at body temperature (27); to avoid the loss of carbon dioxide during hydrogen gas equilibration, he equilibrated a small bubble of hydrogen with successive samples of blood until PCO₂ in the bubble was equal to that in the blood sample. This method allowed Hasselbalch to advance the understanding and definition of clinical acid-base disturbances. The first blood glass pH electrodespecifically designed to keep carbon dioxide in solutionwas constructed by Phyllis T. Kerridge (1902-40) in London in 1925. Seven years later, D. A. McInnes and D. Belcher replaced the cup with capillary tubing and added a clever three-way glass stop-cock for making a fresh liquid junction with saturated potassium chloride, thus creating the first truly precise blood pH electrode. The pH of blood is a strong function of the temperature of measurement, falling 0.015 unit per degree Celsius rise; to mitigate this effect, a thermostated blood pH apparatus was invented in 1931, but did not become commercially available until the mid-1950s. Accurate temperature correction factors for blood pH were first published by T. B. Rosenthal in 1948.

	BLOOD GAS ANALYSIS

Until the introduction of electrochemical methods of analysis in the mid-1900s, measurement of blood oxygen and carbon dioxide contents depended on vacuum extraction, usually in combination with acidification to liberate the contained carbon dioxide, and chemical alteration of oxyhemoglobin to liberate the oxygen. The freed gases were quantified volumetrically until Donald D. van Slyke (1883-1971) developed a more accurate manometric method, which became the gold standard of blood gas analysis for more than a quarter of a century (28).

PCO₂ Analysis by Equilibration

Blood carbon dioxide content is almost all bicarbonate. Since the formulation of the Henderson-Hasselbalch equation, it had been possible to calculate PCO₂, but only after first measuring the blood pH with a glass electrode and the carbon dioxide content with the Van Slyke technique. This cumbersome and time-consuming methodology was used on a grand scaleclinicallyfor the first time in Copenhagen during the polio epidemic of the early 1950s to document the need for, and consequences of, breathing support (described more fully below). These exigencies led Poul Astrup to come up with a novel technique based on the principle that in the clinically relevant range, there was a linear relationship between the pH and log PCO₂ of blood. Astrup (29) designed an apparatus in which one could first measure the pH of a blood sample, and then bubble gas of known PCO₂ through the sample, and measure the pH again. He did this at two different PCO₂ gas values, plotted the measured pH against log PCO₂, drew a line between the points, and located the initial pH on this line, to identify the original PCO₂. The deviation of this line from a normal position was used to define the acid-base imbalance of the patient. Terms such as standard pH and standard bicarbonate were used at first, but later the term base excess, or its in vivo equivalent, standard base excess (SBE), came into widespread use. This system was promptly marketed by Radiometer A/S as the Astrup Apparatus, and shortly thereafter modified by Astrup's associates, especially Ole Siggaard Andersen, to use very small blood samples (30).

PCO₂ Electrode

In Columbus, Ohio, Richard Stow, who was also struggling with the care of polio patients, conceived of an electrode for measuring PCO₂ (31). Stow knew that carbon dioxide permeated rubber freely and that it acidified water. He constructed his own glass pH and reference electrode, wrapped it with a thin rubber membrane over a film of distilled water, and showed it responded to changing PCO₂. Stow refused to patent the idea, because he believed his electrode would never be stable enough to be reliable. At the NIH in Bethesda, Severinghaus and A. Freeman Bradley showed that its sensitivity could be doubled and it could be stabilized by adding NaHCO₂ to the electrolyte (32).

PO₂ Electrode

Toward the end of the 19th century, both Pflüger and Krogh had developed methods for equilibrating small gas bubbles with large volumes of blood to permit analysis of the gas tensions in the bubble. In 1942, F. J. W. Roughton (1899-1972) and Per F. Scholander (1905-80) constructed a syringe with a calibrated capillary attached for analysis of carbon monoxide in blood. In 1945 Richard Riley adapted the Roughton-Scholander syringe method for measuring blood PO₂ and PCO₂. The "Riley bubble method" was widely used by respiratory physiologists, particularly to study ventilation-perfusion relationships in the lungs, but it had virtually no clinical utility. That came with the development and perfection by Leland Clark of the PO₂ electrode. In 1952 Clark adapted polarography to measure performance of his blood oxygenator by covering a platinum cathode with cellophane to exclude protein (33). He also tried a polyethylene membrane successfully, but at first rejected it, believing it could not be dependable because the reference electrode was outside the membrane. On October 4, 1954, he suddenly realized he could put a reference anode under the polyethylene along with the cathode, and that day, he constructed the first modern PO₂ electrode.

Clark and Stow independently discovered the technique of using differentially permeable membranes to separate an electrochemical cell from the substance to be analyzed. Clark's PO₂ electrode required stirring and calibration with tonometered blood. In 1957, Severinghaus and Bradley constructed the first blood gas apparatus as a thermostated water bath in which their modification of the Stow PCO₂ electrode and a cuvette with stirring paddle for Clark's electrode were combined with a miniature tonometer, in which a blood sample could be equilibrated with known gas to calibrate the PO₂ electrode (32). The needs for stirring and tonometric calibration were eliminated by miniaturizing the cathode in 1959. Astrup's equilibration method gradually gave way to three-electrode systems for measuring pH, PCO₂, and PO₂, which are extensively used in clinical and research studies of cardiopulmonary physiology, and which are the current gold standard of respiratory monitoring in intensive care units throughout the world.

Transcutaneous Blood Gas Analysis

Beginning in 1972, Dietrich Lübbers (1918-) and several of his students in Marburg demonstrated that when skin was heated to 42-45° C, it was possible to measure transcutaneously a reasonable value for arterial PO₂, especially in newborn babies (34). Shortly afterward, transcutaneous electrodes were developed for measuring PCO₂. The development and evaluation of transcutaneous PO₂ and PCO₂ sensors in the United States was catalyzed in 1974 by the Division of Lung Diseases through the request for contract proposal mechanism (35). Currently, optical (fluorescence) techniques for measuring pH, PCO₂, and PO₂ are competing with electrode methods for both laboratory and clinical (cardiopulmonary bypass apparatus control and intravascular measurement) applications.

Oximetry

The concepts underlying oximetry go back to Hoppe Seyler's connection of oxygen with hemoglobin's red color (above) and the spectroscope invented by Robert W. E. Bunsen (1811-99) and Gustaf R. Kirchoff (1824-87) in Heidelberg. Important contributions were made by an American, Glen Millikan (1906-47) (36), and a German, Kurt Kramer (1906- 85), whose respective research was greatly accelerated during World War II by the need to monitor blood oxygen saturation because pilots of both the Allied and German air forces were blacking out at high altitudes. The concept of using multiple wavelengths to distinguish among pigments of carbon monoxide-, met-, and oxy-hemoglobins was introduced by Robert Shaw of San Francisco in 1964. Finally, in 1972 in Tokyo, Takuo Aoyagi invented the pulse oximeter, which is based on the equation he derived that makes it possible to compute arterial oxygen saturation without precalibration, independent of ear thickness, skin pigment, hemoglobin concentration, and light intensity (37).

For more detailed history of blood gas physiology and analysis, the reader may consult Astrup and Severinghaus (38, 39) and West's recent collection of essays (40).

	INTENSIVE CARE MEDICINE

Clinical use of blood gas analysis of pH and PCO₂ began with the polio epidemics as described previously, but commercial devices able to also measure PO₂ came surprisingly slowly into common use. Not so intensive care medicine. Like the goddess Venus, it emerged fully grown and ready for action. The birth and instant maturation of intensive care medicine occurred in 1952 during the last worldwide epidemic of poliomyelitis, a scourge that in Copenhagen, Denmark, was unprecedented in its number of victims, in the high attack rate among adults, and in the severity of the accompanying paralysis. The lessons learned while ventilating hundreds of patients who were unable to breathe by themselves prompted the rapid design, manufacture, and extensive deployment of the prototypes of modern ventilators. The need for prompt and accurate pH and PCO₂ measurements forced the relocation of blood gas analysis from the research laboratory to the ward, and accelerated the development of new techniques purely for clinical application. Finally, the fact that the newly developed teamanesthetist, internist, surgeon, and clinical physiologist, supplemented by nurses and medical studentswas a better organization than the pre-existing hierarchical system for coping with the huge problem that presented itself was accepted and was here to stay. Thus, all the basic elements of modern intensive care were formulated in Copenhagen during the late summer and fall of 1952. Although its history has been nicely recounted by Wackers (41), the highlights are worth repeating here in view of their importance and relevance.

Blegdamshospital, 1952

No one was ready for the epidemic of poliomyelitis that ravaged the earth in the early 1950s. And the staff at Blegdamshospital in Copenhagen, a center of medical and epidemiologic expertise in infectious diseases, was as ill prepared as everywhere else. During the first 3 wk of the epidemic, 31 patients with respiratory muscle paralysis or bulbar polio were admitted and ventilated with the then available respirators: one Emerson iron lung and six cuirass machines. Twenty-seven of the 31 patients died within 72 h, a mortality rate of 90%, and it was clear there was much more to come. A catastrophe was in the making.

The chief physician and epidemiologist at Blegdamshospital since 1939, Henry Cai Alexander Lassen (1900-74), later admitted, "Although we thought we knew something about the management of bulbar and respiratory poliomyelitis, it soon became clear that only very little of what we did know at the beginning of the epidemic was really worth knowing" (42). After the disastrous first 3 weeks, Lassen knew he had to do something, but he was not sure exactly what. He was advised to consult Bjorn Ibsen (Figure 4), a free-lance anesthetist at Copenhagen's university hospital. Lassen not only had to overcome a certain degree of professional pride in seeking "outside" help, he was genuinely skeptical about the contribution that Ibsen, an anesthetist, could make. Anesthesiology was just then emerging as a medical specialty, and it was held in low regard; moreover, its activities were confined to the operating room. But Lassen did invite Ibsen to participate in a decisive conference on August 25, 1952, at which the hospital's leading physicians met to discuss the looming disaster.

View larger version (106K): [in this window] [in a new window]   Figure 4.   Photograph of Bjorn Ibsen, Danish anesthesiologist, who established the first intensive care unit to treat patients with poliomyelitis during the 1952 epidemic in Copenhagen.

One of the things that was considered at the meeting was that polio patients died with high total carbon dioxide contents in their blood, as measured by the Van Slyke manometric method. According to convention at the time, this meant the patients had metabolic alkalosis, but when this was said, Ibsen immediately commented that the high values could just as well be explained by retention of carbon dioxide. After the session, Ibsen examined some patients, studied their records, looked at specimens from four autopsies, and became convinced that the patients had died from lack of ventilation. Blegdamshospital's physicians had focused on assisting their patients' breathing and supplying oxygen when needed, using the presence of cyanosis as a guide; in the process, they had ignored the accumulation of carbon dioxide from inadequate exchange of air.

During the discussion, Ibsen proposed to use hand-supplied positive pressure instead of the customary machine-generated negative pressure, and to gain access to the airway with a balloon-cuffed tube inserted through a tracheotomy. He intended to use a Waters to-and-fro rebreathing system in which the patient would be ventilated by hand; this required that someone had to be at the bedside to squeezeover and over again, hour after hourthe rebreathing balloon into which oxygen or air was flowing. Such a system had never been tried at Blegdamshospital before.

Lassen gave permission to go ahead, and the next day, August 26, a 12-yr-old girl who was thought to be dying from severe polio was tracheotomized, intubated, and manually ventilated. It was stormy but it worked. The total carbon dioxide content of her serum fell from about 40 to less than 20 mM/L, which showed that Ibsen's prediction was correct. Using an instrument called a carbovisor that measured expired carbon dioxide, Ibsen demonstrated to the local physicians how he could manipulate the patient's carbon dioxide levels by varying the frequency and force of his manual compressions of the bag. He could also make the clinical manifestations attributable to carbon dioxide retention come and go. The coup de grâce was applied when the patient was returned to a negative pressure ventilator and her exhaled carbon dioxide gradually began to rise.

Poul Astrup, director of the clinical laboratory, prevailed on Radiometer A/S in Copenhagen to provide a pH electrode that they had recently developed for biologic measurements on small samples, including blood. The next day, Astrup was able to measure pH in blood directly with the new electrode, and he quickly confirmed Ibsen's conclusion that patients with terminal stage bulbar polio were acidotic, not alkalotic.

Jolted from his initial skepticism, Lassen, now convinced, devoted himself energetically to implementing the new treatment. Teams of internists, otolaryngologists, and anesthetists were organized to deal with the flood of new patients. And hands were needed, lots of hands, to squeeze the bags to ventilate the patients 24 h/d during the 2 to 3 mo it usually took for them to recover the ability to breathe. At the peak of the epidemic, 40 to 50 new patients were admitted every day and 70 hospitalized patients required manual breathing assistance. Initially, the "hands" belonged to medical students (Figure 5), who worked four shifts of 6 h each. When the need exceeded the number of available medical students, dental students were recruited. According to one observer, in total, approximately 1,500 students contributed 165,000 hours of life-preserving service, squeezing rubber bags (43).

View larger version (119K): [in this window] [in a new window]   Figure 5.   Photograph of a young girl with polio, showing Ibsen's life-saving system: a tube supplying oxygen, a cuffed tracheotomy tube, and a medical student squeezing the rebreathing bag.

After Ibsen's persuasive demonstration of the real clinical hazards of underventilation, regular arterial blood samples were taken from manually ventilated patients for measurements of pH, using the new Radiometer electrode, and total carbon dioxide by Van Slyke's method. After calculating PCO₂ from the Henderson-Hasselbalch equation, the medical students were given instructions, when needed, about how to modify the frequency and intensity of their ventilatory efforts (44).

Aftermath

Primitive though it was, Ibsen's approach was highly successful. The mortality rate of ventilated patients dropped from 90% at the beginning to 25% at the end of the epidemic, and the world took notice. Physicians from all over Europe visited Blegdamshospital and were impressed by what they saw. Everyone, though, locals and visitors alike, recognized the need to replace the medical students, good as they were, with machines capable of delivering constant positive-pressure ventilation. It turned out that the Swedish physician-engineer Carl-Gunnar Engström had designed and built a volume ventilator in 1950 that incorporated a negative-pressure "suck" during exhalation to compensate for any impairment of venous return imposed during the positive-pressure inspiratory phase. Moreover, the machine had been successfully used in 1951 to treat a single patient with chronic poliomyelitis in whom a negative-pressure respirator seemed to be inadequate. In the autumn of 1952, Engström's volume-controlled, positive-pressure ventilator was taken to Blegdamshospital for the treatment of patients with bulbar polio. It performed extremely well, so well in fact that it persuaded Swedish health officials to plan for their own inevitable epidemic by ordering several machines. These were manufactured in time to be available the following summer when poliomyelitis struck Stockholm with a ferocity similar to that experienced in Copenhagen. But Stockholm was ready. All patients with bulbar or respiratory polio were treated with mechanical Engström ventilators and Swedish medical students were not needed.

As the word spread, new ventilators were designed, built, and marketed with extraordinary speed. By 1953, hospitals searching for positive-pressure machines could choose from among several different models and trademarks. Within a few years, the switch from negative pressure to positive pressure for mechanical ventilation was complete throughout Europe, but not in America. When polio swept through the United States in the early to mid-1950s, we had only old-fashioned tank respirators, and our patients were the worse for it.

Another formative lesson learned from the Blegdamshospital experience was the vital importance of the clinical laboratory in the care of critically ill patients. Monitoring the adequacy of ventilation proved crucial. Shortly after the epidemic in Copenhagen, Astrup completed work on his equilibration method for measuring PCO₂, and Radiometer A/S began manufacturing instruments that were soon used throughout the world to measured pH and PCO₂. Astrup and his colleagues also developed new conceptual insights for evaluating acid-base balance, such as "standard bicarbonate" and "base excess," which were rapidly assimilated and applied in intensive care units that were mushrooming everywhere. As could be expected from the early successes, these prototype units began welcoming not just polio patients, but those with respiratory failure from any cause, lung disease, drug overdose, central nervous system impairment, and after thoracic or other surgery, in some instances using paralysis, sedation, and artificial ventilation for a few days while healing proceeded. And not long afterward, highly specialized units began to appear in which facilities and expertise were focused on single groups of patients: premature babies, patients with burns, patients with heart attacks or arrhythmias, patients after open-heart surgery, patients with neurosurgical or neurologic disorders, and on and on, depending on local needs and politics.

	THE FUTURE

Groucho Marx's old lament that "even the future isn't what it used to be" has a heightened contemporary sting to it owing to the intrusion of managed care into modern medicine. It is difficult to predict how the struggle will play out between managed care's zeal to cut costs in order to make a profit and industry's unrestrained development of costly instruments and medications that American physicians and patients eagerly assimilate. On the one hand, there is bound to be increasing pressure to reduce beds and services in intensive care units, notorious guzzlers of funds. On the other hand, it is equally certain that there will be no let-up in the burgeoning number of high-tech expensive devices and products that enhance monitoring capabilities, minimize or eliminate human error, and provide new therapeutic benefits. Some sort of equilibrium will undoubtedly be reached.

Since the mid-1960s most critical care facilities have been able to obtain rapid blood gas analysis, usually using locally placed instruments. Because these devices were designed to be self-calibrating after about 1970, it became common practice for physicians, nurses, and respiratory therapists to withdraw blood samples from indwelling arterial lines, carry them over the apparatus, inject the sample, and read the answer usually as printed by the analyzer. This practice facilitated rapid therapeutic decision making and led to the widespread feeling that a blood gas analysis was the most useful laboratory test in critically ill patients.

Unfortunately, this situation was too good to last. There are changes taking place now that alter the availability and methods of blood gas analysis in many locations. Regulatory agencies have decreed that only licensed technologists may operate these automated blood gas analyzers, and this has resulted in reduced availability of rapid analysis in many locations, because the cost of keeping a licensed technologist at hand 24 hours per day cannot be justified. There may have been economic motivations behind this change, which has resulted in the transfer out of intensive care, operating rooms, and other locations of the blood gas service and resulting revenue to clinical pathology laboratories. Pressure on the regulatory agencies for this change is said to have originated with pathologists.

At the same time, technologic changes have permitted introduction of bedside, or hand-held blood gas analysis devices. These devices use disposible analytic cartridges, operate with newer optical sensors in some cases, and do not fall under the category of the installed blood gas analysis, so it can be performed by physicians, nurses and respiratory therapists, without laboratory technology certification. In general this change may have reduced accuracy because the devices are not calibrated automatically just before analysis is done and cost more per analysis. It remains unclear whether cost containment will further impede availability and accuracy in this field.

The introduction of pulse oximetry has greatly reduced the frequency of blood gas analysis for two reasons: because of the constant threat of hypoxia, and because acid-base status changes occur more slowly than oxygenation failure. An interesting technological development may lead to another change in practice. Pulse oximeters are becoming so small, rapidly responding, and battery operated that they can be kept by the clinician (in a pocket or bag) and used whenever needed to assist in decision making, whether in an ICU or an office. Such devices were recently used by climbers on Mount Everest.

Pulse oximetry has also severely cut into the use of transcutaneous PO₂ and PCO₂ electrodes for monitoring. This dip seems to have partially reversed because the continuously available values from transcutaneous PCO₂ measurement have proved useful in both neonates and adults in some critical care situations, even when skin PO₂ is recognized to be an inadequate index of arterial oxygenation. Transcutaneous PCO₂ devices provide a very stable and sensitive monitor of changes produced by subtle manipulations of artificial ventilation, by the use of continuous positive airway pressure and positive end-expiratory pressure, and by the administration of sedative and opioid medications. Our prediction is that use of transcutaneous PCO₂ electrodes will increase.

A final comment relates to determination of gut PCO₂ in the critical care of patients who are in shock. The concept has been called gastric tonometry (45). It is based on the idea that mucosal PCO₂ is normally 5-10 mm Hg higher than arterial PCO₂. In shock, as flow stagnates, mucosal surface PCO₂ rises far above arterial PCO₂, for three reasons: (1) oxygen consumption and carbon dioxide production tend to remain normal, so as flow falls, the arteriovenous difference rises, elevating surface PCO₂; (2) blood supply to mucosal surface in the gut consists of capillary loops that permit some countercurrent carbon dioxide exchange, so at low flow, carbon dioxide generated near the surface partly diffuses from venules into arteries, causing it to collect at the surface; and (3) as oxygen supply falls to some critical level, mucosal cells make lactic acid, and these hydrogen ions react with tissue bicarbonate to generate carbon dioxide gas in solution, raising local PCO₂. With severe ischemia gut surface PCO₂ rises to the 200-300 mm Hg range (46). Various devices are being introduced to monitor this effect as a continuous index of gut, and hence body, circulatory adequacy. A new, disposable, inexpensive method of measuring PCO₂ has recently been invented (T. I. Tønnessen, Oslo, personal communication), which would be ideal for introduction on the tip of a nasogastric tube, inserted into either the stomach or the small intestine. We predict widespread use of this and similar devices in critically ill patients in the near future.

In other aspects of monitoring, there are sure to be new ways of assessing cardiovascular function noninvasively, continuously, and accurately, including such important variables as blood pressure, cardiac output, regional blood flow, especially to the brain, and even pulmonary arterial and left atrial pressures. Similarly, advances in the speed and fidelity of imaging methods, such as computed tomography and magnetic resonance, will make them safer and more readily available to patients in intensive care units. In the field of therapeutics, current efforts to construct a durable artificial heart and a simplified extracorporeal lung-replacement instrument will finally be realized. We can also predict that future progress will intensify the debate about what we are actually doing in our intensive care units: are we salvaging meaningful lives or are we prolonging inevitable deaths? We hope that along with the technological advances that are going to occur, more attention will be devoted to the personal and societal costs of critical care medicine and to its ethical underpinnings.

	SUMMARY

TOP INTRODUCTION CONCLUSION REFERENCES

Substantial progress in the acquisition of scientific knowledge concerning blood gas transport, which began in the 17th century, led to the discovery of oxygen in air and carbon dioxide in smoke, the presence of these gases in the bloodstream and the role of the lungs in getting them in and out of the body, and finally, how to measure them in blood. These basic research achievements were clinically applied in dramatic and successful fashion in 1952 during the polio epidemic that ravaged Copenhagen, Denmark. An inspired anesthetist, Bjorn Ibsen, after making the right deductions from scanty information, introduced a radical type of therapy that incorporated several novel features: a team approach by experts, a separate facility for trained personnel and special equipment, and a clinical laboratory for essential monitoring. This radical and effective way of treating seriously ill patients launched the proliferation of intensive care units and led to the inauguration of the now flourishing specialty of critical care medicine, where science and clinical medicine continue their powerful partnership.

	Footnotes

Correspondence and requests for reprints should be addressed to John F. Murray, M.D., International Union Against Tuberculosis and Lung Disease, 68, boulevard Saint Michel, 75006 Paris, France.

	References

TOP INTRODUCTION CONCLUSION REFERENCES

1. O'Malley, C. D. O. 1953. Michael Servetus: A Translation of His Geographical, Medical and Astrological Writings with Introductions and Notes. American Philosophical Society, Philadelphia. 204-205.

2. Malpighi, M., and 1661. James Young, trans, . 1929. -30. De Pulmonibus. Proc. R. Soc. Med. 23: 7-11 .

3. Black, J. 1754. Dissertatio Medico Inaugularlis, de Humore acido a cibis orto, et Magnesia alba. Edinburgh. 1-46.

4. Boyle, R. 1680. The sceptical chymist or chymico-physical doubts and paradoxes. Henry Hall, Oxford.

5. Lavoisier, A. L. 1762-93. Oeuvres de Lavoisier. Imprimerie Impériale, Paris.

6. Magnus, H. G.. 1837. Über die im Blute enthaltenen Gase, Sauerstoff, Stickstoff und Kohlensãure. Ann. Phys. Chem. 40: 583 .

7. Magnus, H. G.. 1845. Über das Absorptionsvermögen des Blutes für Sauerstoff. Ann. Phys. Chem. 66: 177 .

8. Meyer, L. 1857. Die Gase des Blutes. Dieterich, Göttingen.

9. Ludwig, C.. 1865. Zusammenstellung der Untersuchungen über Blutgase. Zeitschrift kaiserlich königlich Gesellschaft der Ärzte in Wien 1: 145 .

10. Pflüger, E. F. W.. 1872. Über die Diffusion des Sauerstoffs, den Ort und die Gesetze der Oxydationsprozesse im tierischen Organismus. Arch. Gesamte. Physiol. 6: 43 .

11. Bohr, C. 1905. Blutgase und respiratorische Gaswechsel. In W. Nagel, editor. Handbuch der Physiologie des Menschen. I:54.

12. Krogh, A.. 1910. On the mechanism of the gas-exchange in the lungs. Scand. Arch. Physiol. 23: 248-278 .

13. Berzelius, J. J. 1806-08. Förelãsninger över djurkemien, Vol. 1-2. Marquard, Stockholm.

14. Hoppe-Seyler, F.. 1862. Über das Verhalten des Blutfarbstoffes im Spektrum des Sonnenlichtes. Arch. Pathol. Anat. Physiol. 23: 446 .

15. Hüfner, C. G.. 1894. Neue Versuche zur Bestimmung der Sauerstoffkapazit des Blutfarbstoffs. Arch. Pathol. Anat. Physiol. 55: 130 .

16. Bert, P. 1878. La Pression Barometrique. G. Masson, Paris.

17. Bohr, C., K. A. Hasselbalch, and A. Krogh. 1904. Ueber einen in biologischer Beziehung wichtigen Einfluss, den die Kohlensãurespannung des Blutes auf dessen Sauerstoff-bindung übt. Scand. Arch. Physiol. 16: 402-412 .

18. Severinghaus, J. W.. 1979. Simple, accurate equations for human blood O₂ dissociation computations. J. Appl. Physiol. 46: 599-602 [Abstract/Free Full Text].

19. Henderson, L. J.. 1909. Das Gleichgewicht zwischen Basen und Sauren im tierischen Organismus. Ergebn. Physiol. 8: 254 .

20. Henderson, L. J. 1928. Blood. Yale University Press, New Haven.

21. Hasselbalch, K. A.. 1917. Die Berechnung der Wasserstoffzahle des Blutes aus der freien und gebundenen Kohlensaure desselben und die Sauerstoffbindung des Blutes als Funktion des Wasserstoffzahl. Biochem. Z. 78: 112 .

22. Arrhenius, S. A.. 1887. Über die Dissociation der in Wasser gelösten Stoffe. Z. Physik. Chemie 1: 631-658 .

23. Ostwald, W. F. 1896. Elektrochemie, ihre Geschichte und Lehre. Veit & Co., Leipzig.

24. Nernst, W. H.. 1889. Die elektromotorische Wirksamkeit de Jonen. Z. Physik. Chemie 4: 129-181 .

25. Heyrovsky, J.. 1922. Electrolysis with the dropping mercury electrode. Chemicke. Listy. 16: 256-304 .

26. Sørensen, S. P. L.. 1909. Enzymstudien II. Mitteilung über die Messung und die Bedeutung der Wasserstoffionen-konzentration bein enzymatischen Prozessen. Biochem. Z. 21: 131-304 .

27. Hasselbalch, K. A., and C. Lundsgaard. 1912. Elektrometrische Reaktionsbestimmung des Blutes bein Korpertemperatur. Biochem. A. 38: 77 .

28. Van Slyke, D. D., and J. M. O'Neill. 1924. The determination of gases in blood and other solutions by vacuum extraction and manometric measurement. J. Biol. Chem. 61: 523 [Free Full Text].

29. Astrup, P.. 1956. A simple electrometric technique for the determination of carbon dioxide tension in blood and plasma, total content of carbon dioxide in plasma and bicarbonate content in `separated' plasma at a fixed carbon dioxide tension. Scand. J. Clin. Lab. Invest. 8: 33 [Medline].

30. Siggaard Andersen, O., K. Engel, K. Jorgensen, and P. Astrup. 1960. A micro method for determination of pH, carbon dioxide tension, base excess and standard bidarbonate in capillary blood. Scand. J. Clin. Lab. Invest. 12: 172-176 .

31. Stow, W., R. F. Baer, and B. Randall. 1957. Rapid measurement of the tension of carbon dioxide in blood. Arch. Phys. Med. Rehabil. 38: 646-650 .

32. Severinghaus, J. W., and A. F. Bradley. 1958. Electrodes for blood PO₂ and PCO₂ determination. J. Appl. Physiol. 13: 515-520 [Free Full Text].

33. Clark, L. C. Jr.. 1981. Measurement of oxygen tension: a historical perspective. Crit. Care Med. 9: 960-962 .

34. Huch, R., A. Huck, and D. Lübbers. 1973. Transcutaneous measurement of blood PO₂ (tcPO₂). J. Perinat. Med. 1: 183-190 [Medline].

35. U.S. Department of Health and Human Services. 1980. Evaluation report on completed contracts awarded in response to RFP NHLI 75-1, `Development and clinical evaluation of blood gas sensors for continuous monitoring (adult and neonate)'. U.S. Government Printing Office, Washington, DC. NIH Publication No. 81-2207. 1-16.

36. Milliken, G. A.. 1942. The oximeter: an instrument for measuring continuously oxygen saturation of arterial blood in man. Rev. Sci. Instr. 13: 434-444 .

37. Aoyagi, T., M. Kishi, K. Yamaguchi, and S. Watanabe. 1974. Improvement of an earpiece oximeter. In Abstracts of the 13th Annual Meeting of the Japanese Society for Medical Electronics and Biological Engineering. Osaka. 90-91.

38. Astrup, P., and J. W. Severinghaus. 1986. The History of Blood Gases, Acids and Bases. Munksgaard, Copenhagen. 1-332.

39. Severinghaus, J. W., and P. B. Astrup. 1987. History of Blood Gas Analysis. International Anesthesiology Clinics, Little Brown, Boston. 24: 1-224.

40. Astrup, P., and J. W. Severinghaus. 1996. Blood gas transport and analysis. In J. B. West, editor. Respiratory Physiology: People and Ideas. Oxford University Press, New York. 75-107.

41. Wackers, G. L.. 1994. Modern anaesthesiological principles for bulbar polio: manual IPPR in the 1952 polio-epidemic in Copenhagen. Acta Anaesth. Scand. 38: 420-431 [Medline].

42. Lassen, H. C. A. 1956. Management of poliomyelitis. In H. C. A. Lassen, editor. Management of Life Threatening Poliomylitis, Copenhagen, 1952-1956: With a Survey of Autopsy Findings in 115 Cases. E. S. Livingstone Ltd, London.

43. Hansen, J.. 1953. Den okonomiske baggrund for poliobekaempeisen. Ugeskr. Lager. 115: 471-473 .

44. Astrup, P. 1956. Laboratory control of gas exchange. In H. C. A. Lassen, editor. Management of Life Threatening Poliomylitis, Copenhagen, 1952-1956: With a Survey of Autopsy Findings in 115 Cases. E. S. Livingstone Ltd, London. 111-120.

45. Fiddian-Green, R. G.. 1995. Gastric intramucosal pH, tissue oxygenation and acid-base balance. Br. J. Anaesth. 74: 591-606 [Free Full Text].

46. Schlichtig, R.. 1996. Tissue-arterial PCO₂ difference is a better marker of ischemia than intramural pH (pHi) or arterial pH-pHi difference. J. Crit. Care 11: 51-56 [Medline].