|
Professor of History
John.Heitmann@notes.udayton.edu
|
|
Recent Scholarship & Works in Progress
|
Absolutely the Right Tool for the Job: Atomic Absorption Spectroscopy and Childhood Lead Poisoning, 1955 - 1990
John A. Heitmann Department of History University of Dayton 300 College Park Dayton, OH 45469-1540
July 1, 2000
Lead and its impact upon the public health of young children is undoubtedly one of the most significant case studies on the social consequences of science and technology to take place in this century.(1) The historical literature on this topic, however, is far from complete, contains widely divergent views, and has significant gaps.(2) By asking a simple question related to the history of analytical chemistry and instrumentation, I hope to shed new light on not only on WHY it took so long before a consensus crystallized on perceived risks related to lead, but also HOW this issue has moved from the periphery to center stage in our own time.(3) During the late 1980s and 1990s, much has been done in the history of instrumentation as it relates to scientific theory and practice, the sociology of science, or as a "handle" to better understand the science-technology interface.(4) Nevertheless, rarely has the scientific instrument been explored beyond itself or its role in critical experiments that subsequently altered the course of science.(5) Historians of science have been largely remiss in studying the instrument as an everyday tool used routinely within the working laboratory. It is precisely this angle towards instrumentation as applied to the environmental history of lead that I wish to explore in this essay. In doing so, we go beyond an examination of the connections between instrumentation and experiment, and we explore how instruments and the knowledge they created helped shape public policy. The development of the right tool, used in the right way and in the right place played an important role in the "lead in the environment" story.(6) Thus, given the difficulties in determining lead in blood in micro-analytical quantities with speed, accuracy, and selectivity with a minimum of interferences and free from contamination, it is no surprise why it took so long for the lead "problem" to be clearly recognized as the public health threat that it was and can be. However, there existed other important factors as well. These included the rise of new political pressure groups after the mid-1960s, an emerging awareness of the threats to human health from trace contaminant pollutants, and a developing realization that racism extended to the environment. Despite the complexity of the topic and confounding factors, my approach of characterizing the related science and technology suggests one significant reason why it took so long before a firm federal policy commitment to its solution was first offered. Atomic Absorption (AA) instrumentation ultimately produced for the first time large amounts of reliable blood lead level data. Thus, by the late 1980s policy makers, weighing the risk of damage to the young and sensitive to political pressure groups, sponsored legislation that called for screening large elements of the population as well as remedying the unmaintained lead paint on the walls and woodwork of the nation's deteriorating housing stock. Further, AA instrumentation contributed greatly to the gradual evolution of thought regarding the affliction of lead poisoning in children. Until the 1960s, pediatricians worried about the stark symptoms of coma and seizures, anemia and peripheral neuropathy. For the most part, during the decades that followed, very different effects typically occurring at the low end of the dose-response curve became the focal point of concern, first among occupational and environmental epidemiologists. Subtle and difficult to discern, these neurological, behavioral, and intellectual deficits were largely invisible and were critically studied only with the aid of blood lead assays coupled with psychological testing and statistical analysis. Indeed, blood lead determinations were the benchmark or the "gold standard" before which risks concerning lead poisoning could not be delineated.(7) What blood lead data provided, then, was diagnostic clarity, particularly at the lower end
of the dose-response curve. The significance of the absorption of lead related to risk depended
on data that urine samples or an office visit to a physician could never provide. And precisely
because this data became better and better in terms of accuracy and precision, the minimal
threshold safe lead level in blood was gradually lowered. For example, in 1965 the threshold
level of lead in blood of a child considered having an elevated level was 65 µg/dl, but by 1970
the figure had fallen to 40 µg/dl. Subsequently this number was lowered to 30 µg/dl between
1975 and 1985, and since 1991 it has been 10 µg/dl.(8) This stepped decline, along with a parallel
drop in what was considered a safe daily intake total of lead, from 300 micrograms a day in 1971
to 35.7 micrograms per day in 1993, reflects how emerging scientific knowledge, recognized and
assimilated by regulatory agencies and policy makers, can have profound societal implications.
Yet as this case study demonstrates, the power to facilitate enduring change does not necessarily
follow scientific knowledge, for politics and economics proved decisive in how this story ends.
Atomic Absorption Spectroscopy and the Quest for Absolute Analysis Since this essay centers on the genesis and development of a technique for attaining absolute blood lead analysis by the application of one particular instrument, the atomic absorption spectrophotometer (AAS), it is appropriate to begin with a brief description of this device. This instrument measures the intensity of radiation of a specific wavelength that is absorbed by an atomic vapor. Its radiation source is that of a hollow cathode tube or electrical discharge lamp, and light of a single specific wavelength is passed through a cloud of atomic vapor that was created for the most part in two ways. In its initial design, AA atomization was achieved by the use of air-propane, air-acetylene, and later in the 1960s nitrous oxide-acetylene flames contained within a specially designed aspirated burner into which a solution of the material to be analyzed was injected as a fine mist of droplets by a nebulizer apparatus. Alternatively, somewhat later in time and important to our story was the use of an electric furnace, but in either case detection of the resulting changes in the intensity of the incident light beam was accomplished with a photomultiplier tube and associated electronics. Since intensity of radiation before and during sample vaporization is proportional to elemental concentration, it was not surprising that, once introduced, the AA quickly became an important tool for analytical chemists, especially those involved in trace metals analysis. The shift to AA on the part of the majority of workers engaged in blood lead analysis took place rather rapidly during the late 1960s, although it would take another 15 years before AA instrumentation, automation, software and sample handling procedures were refined. Key to this transition was the simple fact that AA, when compared to any competing analytical method, was superior in analyzing elements as they were found in nature, or in detecting and quantifying "things as they are."(9) Indeed, in 1933 chemist G.E.F. Lundell had stated that "there is no dearth of methods that are entirely satisfactory for the determination of elements when they occur alone. [However] The rub comes in because elements never occur alone, . . . ."(10) During Lundell's time, chemists, often responding to demands placed on them by workman's compensation boards, developed a number of methods for the determination of lead and other metals in trace amounts found in biological materials. By the mid 1930s, many analysts' method of choice employed an organic compound called diphenylthiocarbazone, or dithizone, a complex organic substance that reacted with a great number of metals, including lead, to form highly brilliant colored complexes. These complexes, when subjected to light of a specific wavelength, not only absorbed that light strongly, but also did so in proportion to concentration in solution, and this property formed the basis of a very popular analytical method at mid-20th century -- colorimetry. Dithizone was first prepared by Emil Fischer in 1878, but then neglected until the mid-1920s when Hellmut Fischer, a chemist employed by the Siemens company, began publishing a series of papers on the compound's applications in trace metals analysis.(11) A violet-black solid, dithizone, when dissolved in chloroform or carbon tetrachloride and shaken in aqueous solution, reacted with metals to form corresponding dithizonates that were soluble in the organic layer. Between the mid 1930s and the late 1950s a legion of analysts refined the dithizone method for lead and institutionalized its use in laboratories. Despite the rhetoric of its defenders and advocates, however, the dithizone method for lead was not only very slow and subject to interferences, but demanded a host of complex manipulations that included pH control, extraction, back extraction and the addition of masking agents -- all of which contributed to a compounding of errors and opportunities for further sample contamination. Particularly in the 1930s, dithizone was used to measure lead in urine, and once a bismuth interference problem was solved by the use of cyanide as a complexing agent, researchers discovered a pattern of high results due to reagent oxidation and contamination. For example, in 1937 two General Electric chemists wrote "Our earlier experience with the lead analysis seemed to show that lead is present everywhere; . . . satisfactory quantitative results on small samples can be obtained only through painstaking and somewhat tedious manipulation. For this reason it often will be advisable to use other methods of analysis. . . ."(12) While further improvements followed during the 1940s and the 1950s, one could not escape the fact that the dithizone procedure demanded large quantities of sample -- 5ccs or more of fresh blood, often taken from children never fond of needles -- lengthy pretreatment using acids and ashing, and the utmost patience of a skilled analyst who had to perform a series of extractions using purified reagents and acid-washed glassware.(13) It was a labor intensive determination that often took two days or more to complete five samples, and a procedure that was fraught with difficulties. In 1991, for example, one chemist remembered the time when high blanks (contaminated control samples) associated with dithizone lead analysis "were responsible for more profanity per hour in the labs . . . than any other single factor."(14) Given these shortcomings, until the mid-1950s only two laboratories in the U.S. were sufficiently staffed to routinely assay blood samples for lead using the dithizone method -- the Kettering Laboratories at the University of Cincinnati, largely funded by the tetraethyl lead industry, and the Baltimore City Health Department, which in 1933 began routine testing after incinerated battery cases caused a massive lead poisoning. The Kettering Laboratories, directed by Robert A. Kehoe, had a laboratory staff that included several chemists, more than a dozen technicians, and numerous support personnel. Beginning in 1937, Kehoe's group employed the dithizone method along with carbon arc spectroscopy to assay lead in the blood of both workers and children. Yet despite their resources and expertise, laboratory supervisor Jacob Cholak, reporting on the dithizone method, could claim that "In most analytical work, including this type, errors of ±20 percent are permissible. . . ."(15) Indeed, the challenges of such an operation were so formidable that in 1951 even New York City, with all of its scientific and medical institutional infrastructure and resources, decided not to establish a government sponsored laboratory for the diagnosis of poisonings, preferring to send out samples to commercial laboratories.(16) In the end, AA minimized these inherent liabilities, but its history illustrates among other things that no new instrument design is totally new. In commenting on its history, early AA pioneers are quick to point to the pre-history of the device, finding antecedents in William Hyde Wollaston's 1802 experiments and the 1817 work of Joseph Fraunhofer, who when observing the solar spectrum noted that the dark lines representing wavelengths of absorption were identical in position to characteristic radiation patterns emitted by certain elements found on earth. Henry Fox Talbot and David Brewster continued studies of this phenomena during the 1820s and 1830s, but it remained for the German chemists Robert Bunsen and Georg Kirchhoff to clearly describe emission and absorption spectroscopy. Yet, while a handful of chemists exploited the technique to devise methods for the analysis of mercury during the 1930s, it was more widely used by astronomers to study the elemental composition of the stars.(17) The potential of AA in terms of chemical analysis remained largely ignored until 1955, a date which is recognized by early practitioners as the starting point for the development of the instrument as we know it today. The theoretical, technical and commercial pioneer of atomic absorption spectroscopy was Alan Walsh of the Chemical Physics section of the Commonwealth Scientific and Industrial Research Organization (CSIRO) located in Melbourne, Australia. Simultaneously, a worker in the Netherlands, C.T.J. Alkemade, also published on the same topic in the Journal of the Optical Society of America, but he subsequently gave credit to Walsh due to the latter's persistence and energy in selling the idea to a skeptical audience. Walsh's seminal paper, entitled "The Application of Atomic Absorption Spectra to Chemical Analysis," appeared in a 1955 issue of Spectrochimica Acta and for the most part contained a theoretical discussion of the potential of the proposed method, particularly when compared to existing spectroscopic methods. Walsh's vision in the paper reflected objectives far beyond that of just simply developing another analytical instrument, however, for the author's primary purpose was to realize an absolute method, one in which the element to be determined was free from any interferences due to other materials in its presence. Walsh asserted that In spite of the remarkable advances in technique which have resulted in press-button analyses of high precision at fantastic speeds, there has been practically no progress whatsoever in solving the fundamental problem of devising an absolute method, i.e., a method which will provide an analysis without comparison with chemically analyzed standards or synthetic samples of known composition [my emphasis]. When analyses of miscellaneous materials are required, the task of providing the required range of standards becomes insurmountable and the spectrochemical method then loses its accuracy, since accurate analyses generally necessitate the use of standards which are closely similar in composition to the sample for analysis. In some analyses it is also essential that the sample and standards be similar as regards physical condition.(18)
In 1953, well before publication, Walsh had already opened discussions with the British instrument manufacturer Adam Hilger, Ltd. to construct a device based on his ideas. An agreement on licensing was slow to come, however, for it was not until 1959 when the first Hilger instrument was constructed. Unlike Walsh's prototype consisting of multi-element hollow cathode lamps, source modulation, a burner and simple monochrometer and detector design, the first Hilger unit mated a conventional Ultraviolet-Visible Wavelength spectrometer with a burner and hollow cathode attachments, and it did not employ source modulation. Inherently limited due to the failure to eliminate emission signals generated by the flame, it was a serious misstep at the critical commercial introduction phase. This subsequently placed Hilger at a disadvantageous position compared to its American competitor Perkin-Elmer Corporation (P-E), once the latter had made a firm commitment to market these instruments in quantity. In the United States, P-E began to express interest in the AA idea as early as 1956, and a research group under the leadership of Jim Robinson at the Esso Research Laboratory located in Baton Rouge was also actively pursuing this idea in collaboration with P-E.(19) With the 1959 signing of a license agreement, P-E product development moved steadily forward, and two years later the P-E Model 214 was introduced to the marketplace. Quickly establishing a commercial hegemony over its rivals, P-E pioneered numerous technical breakthroughs in every area of the instrument's design. Their hollow cathode lamp design was the fruit of collaboration not only with Walsh's CSIRO group, but also with MIT metallurgist John Wulf, who it was later recalled made P-E staff scientists' "heads spin with incredible detail on intermetallic compounds, eutectic alloys, sintered metallurgy and gas absorption by metals."(20) The efforts of P-E employees Walter Slavin and Herbert Kahn, however, were central to the firm's rapid ascendancy in AAS.(21) Slavin and Kahn, both possessing engineering design backgrounds, pioneered the concept of applications or problem solving marketing beginning in the late 1950s. In particular, Slavin's early experience in arc emission spectroscopy and its tedious procedures had convinced him that the route to selling instruments began by identifying contemporary difficult analytical problems and then solving them. By relying on reports from applications sales personnel in the field, new methods were subsequently developed. And in this way, Slavin became the first scientist to develop an extraction technique for the determination of lead in blood by the early 1960s, thereby firmly connecting the manufacturer to the clinical marketplace.(22) The P-E 214 was just the beginning of commercial success, as reflected by the P-E 303, 403, 503 and 603 AAs that were introduced between the mid 1960s and 1970s; by the latter date thousands of these instruments were sold annually. Therefore Walsh's vision went beyond that of instrumentation to the concept of absolute
analysis, and concurrent with early pioneering hardware efforts he was also at the center of
applications development that much later led to near absolute methods for the determination of
lead in blood. A 1961 report in Nature summarized the early work of a colleague of Walsh's at
the CSIRO, J.B. Willis, who demonstrated how AA could be used to determine traces of heavy
metals that included lead, mercury, bismuth, nickel, cadmium and zinc in biological materials.
Willis claimed that "sample preparation procedures are much simpler than those involved in
most existing methods: cadmium and zinc can be determined directly in urine and preliminary
experiments suggest that zinc may be determined in blood serum and whole blood by spraying
the diluted material directly into the flame."(23) Indeed, Willis' preliminary work during the early
1960s shaped the research agendas of many clinical analytical chemists then overwhelmed by
sample demands that were the consequence of an epidemic of childhood lead poisoning cases in
Baltimore, New York City, Chicago, and elsewhere.(24)
Baltimore -- The Urban Crucible In a number of important respects, the disease of childhood lead poisoning was defined in Baltimore during the first half of the Twentieth Century. Elizabeth Fee has told much of the story surrounding the remarkable career of Dr. Huntington Williams and his efforts to eliminate a serious problem that first became apparent in the 1930s and led to Baltimore's commitment to blood lead testing beginning in 1935.(25) Initially these determinations were done by a method called emission spectroscopy, certainly the very best procedure of the day but also expensive and demanding specialized expertise that few laboratories in the world could provide.(26) By the early 1940s, the far more cost effective dithizone procedure was routinely followed, and the results were a cause for concern. In 1942, Emanuel Kaplan and John McDonald, the Baltimore scientists responsible for blood lead testing, reviewed their data and concluded not only that most of the non-fatal lead poisoning cases went unreported, but also that children, rather than adult workers, composed a large percentage of these cases, which were often the consequence of youngsters under the age of five gnawing on either cribs or toys.(27) Although the 86 cases reported between 1931 and 1940 were far from trivial, they only foreshadowed a rise in the disease that occurred during the postwar decades -- this time the consequence not of the lead from paint on familiar objects, but rather from disintegrating housing stock and other sources. Total cases of childhood lead poisoning rose from 13 in 1946 to 133 in 1958, however the number dropped to a plateau of around 45 cases per annum by the early 1960s.(28) Clearly the trouble was not confined to Baltimore, and laboratory work to detect lead poisoning in children, including analytical methods development, began to take place in other major American cities. And while the solution in Baltimore to the obvious health effects associated with this public health problem was the consequence of many different strategies, including a far reaching educational program, careful laboratory analysis was essential and yet remained a bottleneck. Given the urgency to further study the consequences of the use of lead paint in urban dwellings and to minimize damage to children, a number of researchers, including Eleanor Berman of the Toxicology Section of the Cook County Hospital in Chicago, developed procedures for lead using atomic absorption during the early 1960s. Due to difficulties with the matrix -- the total chemical environment of the sample, organic and inorganic -- Berman and others committed to AA were forced to follow a series of steps involving the formation of lead complexes that reduced recoveries substantially and generally slowed analysis times. Indeed, Willis, Berman, the Kettering Laboratory Group (located in Cincinnati) and others were all using flame AA instrumentation during the 1960s. Since the technique employed multiple extractions involving ammonium pyrrolidine dithiocarbamate (APDC) and methyl isobutyl ketone (MIBK), workers in the field remained far from Walsh's goal of absolute analysis by the close of the decade.(29) Difficulties with light scattering in the flame led to multiple wavelength corrective measurements, and the stubborn emulsions formed during the extraction procedure demanded extensive centrifugation of the samples before instrumental measurements could be taken. At Perkin-Elmer, Walter Slavin led an analytical development methods staff that by 1966 had improved and simplified the original Berman procedure for the determination of lead in blood. This method was first published in the firm's widely disseminated Atomic Absorption Newsletter and then inserted into the AA methods "cookbook" for AA users. But in the real world of the clinical laboratory, the manufacturer's claimed accuracy and precision were undoubtedly much harder to produce.(30) What was achieved by the application of the MIBK extraction method in terms of number of blood lead samples run, however, was remarkable when compared to laboratory productivity statistics of the earlier era. To be sure, government money enabled more resources to be deployed during the late 1960s and early 1970s attacks upon the childhood lead poisoning problem. Nevertheless, there were clear and practical sample processing limits when using dithizone, and those limits vanished with the introduction of the AA in routine clinical laboratory analysis. For example, in the case of New York City, during the 1960s health officials had ignored Robert Kehoe's advice and initially had used a urine test based on ALA (for delta amino levulinic acid) for screening children. Prophetically, authorities subsequently discovered that the test gave a false diagnosis in about 30% of the cases and failed to detect 1/3 of the true lead poisoning cases.(31) In the wake of this failure of analytical method, Dr. Vincent Guinee, in charge of NYC's childhood lead poisoning program, became convinced that AA blood lead determinations were the proper course. Thus, with funding and adequate personnel, the number of blood lead assays rose from 489 tests in 1968, to 2,648 in 1969 and to 84,493 in 1970.(32) And of the latter total, some 2,134 specimens assayed (or 2.5%) fell in the high risk range of 70-100 µg/dl, and another 3.3 percent fell in the next level of 60-69 µg/dl range. Given this emerging public health problem, a political response was forthcoming with the passage of the 1971 Lead-Based Paint Poisoning Prevention Act that appropriated federal funds for the detection and elimination of lead based paint poisoning and for a related research and development program for the fiscal years 1971 and 1972.(33) A series of charges and counter-charges of foot-dragging on the part of the Nixon Administration and gross incompetence on the part of the NYC bureaucracy followed. Yet, despite all the controversy, federal funds did play a role in further developing AA blood lead analysis.(34) Additionally, research programs were funded that explored the many different ways that lead, perhaps the most ubiquitous trace metal, found its way into the atmosphere, drinking water, dust and soil. The lead poisoning in children problem was more than simply due to leaded paints, but also the by-product of the widespread use of, among other things, tetraethyl lead in gasoline and lead solder in seams of food cans. During the early 1970s, federally funded collaborative research programs, most notably
those spearheaded by the Center for Disease Control (CDC), began to explore alternative ways to
atomize blood samples. A collaborative investigation involving the CDC, the North Carolina
Board of Health, and Baltimore City Hospitals, developed a procedure based on apparatus first
designed by H.T. Delves of the Hospital for Sick Children and London University.(35) The Delves
cup method, employed small 10 to 20 µl samples that were dried, treated with hydrogen
peroxide, and then placed in tiny nickel crucibles that were directly placed into the AA's air-acetylene flame. Difficulties involving serious physical interferences, faulty cups and an initially
flawed design in the sample delivery system were eventually overcome. The modified Delves
cup method, though remaining somewhat tedious and labor intensive, was a marked
improvement on the Berman MIBK procedure. In 1973, using this method, the City of
Baltimore initiated a citywide childhood lead poisoning screening program in which blood was
drawn using a finger prick technique.(36) The Delves method, and minor modifications,
represented more than an improvement in technique, for it was an important transition in the
analytical chemistry of lead at low levels in terms of analyst deskilling. A 1971 article in Perkin-Elmer's Atomic Absorption Newsletter stated that "An important advantage of the Delves
Sampling Cup method . . . is its simplicity. The method demands little time and operator
training, and eliminates the complex sample preparation previously required for blood lead
analysis. With the Delves method, . . . one person can analyze 30-50 samples per hour. It is
unnecessary to clean the sample between crucibles. . . ."(37) Yet, as costs spiraled and sources of
funding became scarce, the Delves method gradually was supplanted by the graphite furnace
technique, in part because the latter could be automated, but also because in its final design the
graphite furnace offered the possibilities of absolute analysis free from matrix interferences.
The Cold War Connection The key to unlocking the atomization bottleneck that prevented the realization of the goals of absolute analysis and sample automation lay not in the U.S. or Australia, but in the Soviet Union. In 1955, Boris V. L'vov, a recent physics graduate of Leningrad University and researcher in the isotope laboratory of Leningrad's State Institute of Applied Chemistry, read Walsh's seminal paper on AA.(38) L'vov had been familiar with emission spectroscopy for some time and was well aware of the method's inherent limitations, including matrix effects that forced the analyst to prepare standards of similar composition to that of the sample. L'vov reasoned that flame atomization simply could not lead to absolute analysis because of incomplete atomization of the sample. Thus, he turned to a technique involving the use of a graphite furnace (GFAAS) with the promise that this would lead to total sample evaporation in a predetermined volume. As L'vov later remarked, "The major goal of this research was to determine the conditions suitable for absolute measurements, regardless of the technical and methodological difficulties." L'vov's first paper on the subject was in Russian, and it was largely ignored in the West. In Germany, however, Hans Massmann picked up on this work, developed designs for graphite furnaces, and subsequently presented his research at a conference in Dresden. Unlike L'vov's design, Massmann used the walls of the furnace to evaporate his sample, thereby losing the advantage of the graphite cuvette to produce analytical results independent of matrix. Yet it was Massmann's design that was commercialized first by a German subsidiary of Perkin-Elmer located near the Bodensee; in 1970 they introduced to the market a line of AA furnaces called the HGA-70.(39) The Massmann furnace design did lead to a 100 fold gain in sensitivity when compared to flame AA, but the goal of absolute analysis remained elusive. Commenting on the status of graphite furnace technique related to lead in blood analysis in 1974, a skilled analyst remarked that "direct analysis in a flameless atomizer of untreated biological materials can at best be considered but a semi-quantitative technique."(40) The shortcomings of the graphite furnace technique in its initial stage of development, however, were largely overcome during the mid to late 1970s, primarily due to L'vov's persistence. Convinced that the difficulties associated with GFAAS were tied to the Massmann design, L'vov was finally able to do something about it after moving to the Leningrad Polytechnical Institute in 1975, where he rather quickly established a formal collaborative relationship with Perkin-Elmer. Consequently, L'vov then had access to a modern P-E Spectrometer and HGA-76 graphite furnace, the first Western instruments in his hands after working nearly 20 years in this field. Thinking that the real problem somehow lay with uneven heating of the sample during evaporation and resulting interference from halogens in the vapor phase, in 1978 L'vov suggested to co-worker Larissa Pelieva that a small platform be inserted inside the graphite sample tube.(41) The platform furnace was thus born, and along with a small constellation of other innovations that were introduced at about the same time, a system of analysis emerged based on the so-called stabilized temperature platform furnace (STPF).(42) Employing autosampling, pyrocoated tubes, chemical matrix modifiers, fast electronics and Zeeman effect background correction, STPF minimized difficulties associated with interference, calibration, and standard solution preparation.(43) The STPF process was also automated in terms of sample digestion and injection by the mid-1980s, in part the result of efforts by Japanese researchers studying levels of metals in blood of farmers.(44) And although AA matrix effects were not totally eliminated by the spring of 1985, theoretical breakthroughs pointed to a very close correlation between experimental and calculated characteristic masses. Further refinements in the movement toward absolute analysis have followed since the
mid-1980s, with a new emphasis on the part of the technique's advocates to "educate regulatory
agencies, plant engineers and analysts."(45) Indeed, CDC state-of-the-art procedures for the
determination of lead in blood as of 1996 directly reflected L'vov's platform design along with
several of the modifications that had their origins in Slavin's laboratory. The P-E model 5000
spectrophotometer, along with a HGA 500 furnace fitted with a L'vov platform and AS-40
autosampler, are specifically called for in the 1996 CDC procedures manual that was employed
to analyze blood samples associated with the Third National Health and Nutrition Examination
Survey (NHANES III) of 1988 to 1994.(46)
A Rival Method: Anodic Stripping Voltammetry (ASV) While this essay's primary focus is that of the historical relationship between changes in AAS methods and emerging public policy on lead levels in blood, it would be remiss not to mention an electrometric analytical method that played an important tangential role in this story. Beginning in the late 1960s, and evolving from the then well developed area of Polarography, Anodic Stripping Voltammetry (ASV) emerged for a time as a second (and for some analysts a preferred) means of determining lead in blood.(47) In several respects, ASV is similar to AAS. Indeed, while electrometric methods, including ASV, count electrons, spectroscopy measures numbers of photons. And in both cases the fundamental particles measured are directly proportionate to the material present in the sample. ASV involves the use of a microelectrode, typically a hanging drop of mercury (HDME), that works in conjunction with reference electrodes, with the complete system's voltage controlled by a potentiostat. When properly set up and adjusted, the ASV apparatus reduces specific metals in solution, including lead, and plates them selectively on the HDME's surface. Thus concentrated, metallic elements are then sequentially stripped, or oxidized, away from the electrode surface by reversing the applied electric potential. The resulting current is measured on a strip chart recorder as a voltammogram, characteristically a Bell shaped curve whose peak height is proportional to the bulk concentration of the metal dissolved in solution. Although the historical roots of ASV are complex, credit for its development is normally given to G.C. Barker, who worked in Great Britain during the 1950s and whose persistent efforts led to the marketing of an expensive device named the Mervyn-Harwell Square Wave Mark III Polarograph in 1958.(48) During the 1960s, the technique's basic ideas found a number of adherents in the U.S., and by the early 1970s several instrument companies, most prominently Princeton Applied Research (PAR), marketed commercial devices that were low in cost and comparatively easy to operate. "Ease of operation," however, is a vague and relative term, as Jud Flato, the originator of the benchmark PAR 170, rather fondly recalled criticism that claimed his device was "the world's only $10,000 pH meter . . . obviously . . . designed by a musician because only a piano player could handle it. . . ."(49) Despite reservations about dealing with finicky electrodes and complex electronic equipment, ASV had its boosters. And they had several very persuasive arguments, including the fact that even the most sophisticated instruments with accessories, including the PAR 170, were many thousands of dollars less than a modestly equipped AAS unit. Furthermore, at least until the mid-1980s, performance data pointed to ASV sensitivities and analytical resolutions equal or better than Graphite Furnace AAS. Indeed, ASV results normally served as a basis of comparison to AAS in terms of accuracy and precision. Ultimately, the rivalry between these two instrumental methods provided much of the stimulus for subsequent critical improvements in technique and instrument design related to the determination of lead in blood at very low levels.(50) By the early 1990s, however, AAS emerged the winner over ASV, not simply because it was cheaper or better, but for a constellation of complex factors. In practical terms, ASV methods involved an extra digestion step that Graphite Furnace AAS in its most refined state did not. Hence, the perception among working chemists, who were already suspicious of a unfamiliar technique and possibly looking for reasons to avoid it, was that ASV was slower and demanded one more laboratory manipulation that was time consuming and prone to increasing the chances for error. Further, ASV was not only perceived by some as a step slower, but also perhaps more dangerous, since many electrode combinations employed mercury, a highly toxic material with a significant vapor pressure even at room temperature.(51) And to complicate matters, during the critical decision making period of the late 1980s, a crisis in confidence in ASV data surfaced, the consequence of the discovery that one specific instrument consistently failed to process data accurately, resulting in low values at blood lead concentrations below 40 µg/dl and higher readings at levels above 40 µg/dl.(52) Thus for very complex reasons, by the early 1990s, the Graphite Furnace had proved to be the right tool for the job of blood lead analysis, but success in the endeavor to produce data with confidence was also due to concurrent developments in what may be called "laboratory hygiene." By the 1930's, the dithizone procedure had called initial attention to the many ways in which extraneous sources of lead could enter the sample under analysis. By the early 1970s, chemists probed deeper and deeper layers of the trace element microcosm at the picogram (10-12) and nanogram (10-9) levels, and far more was learned concerning manifold sources of contamination.(53)
Cleaning Up The Laboratory During its initial introduction, chemical traditionalists claimed that instrumental chemists committed to the AA technique had apparently either neglected or paid scant attention to the intricacies of the art of wet chemical analysis, and thus error was piled on error with the addition of each impure reagent, sampling system and improperly cleaned labware. One critical observer remarked "unfortunately analysts who use such instruments without chemical separations are generally those least experienced and qualified to do so. Analysts who do use chemical separation techniques before determining metals with instruments such as AA generally fail to use proper clean laboratory techniques and therefore contaminate their samples badly. . . ."(54) This commentary and criticism on what might be regarded as "analyst deskilling" was, according to Davis Baird, rather commonplace in the literature of analytical chemistry during the transition from wet to instrumental analytical methods between 1920 and 1950.(55) Nevertheless, in the case of Graphite Furnace AA and similar techniques that examined trace materials at the very lowest of levels, an entirely different laboratory technology had to develop, and it was largely based on the relatively new and key material, Teflon®. In part, a major symposium convened by the National Bureau of Standards on accuracy in trace analysis in 1974 sorted out procedures related to cleanliness. At that pivotal meeting, geologist Clair C. Patterson, a central figure involved in the lead in the environment controversy of the late 1960s and early 1970s, described his efforts at Cal Tech to design a totally clean lead- free laboratory. Derived from nuclear geochemistry techniques that had evolved over a quarter of a century related to U/Pb systems analysis and geochronology, Patterson had designed and built a "clean" laboratory, and in the process had institutionalized sampling procedures that eliminated most or all of the problematic sources of contamination. While Patterson's laboratory consisted of a complex of six different work areas each with its own purpose, one material, Teflon®, stood out for its near universal use in preventing contamination from entering the assay process. Among other things, Teflon® was used to coat electrical wires connected to hot plates, which in turn were covered with thin sheets of FEP Teflon®. Patterson recommended the use of Teflon® watch glasses, beakers, dishes, water distilling apparatus, heated chambers and funnels; and he cautioned against the use of labware that had been the basis of the "wet" chemistry of the past, namely Pyrex®, Kimax®, polycarbonate, nylon, and platinum.(56) Additionally, polyethylene was used in a variety of other situations in Patterson's laboratory, including flooring in the so-called "clean room" that one entered before changing from his street clothes and going into the laboratory proper, where the air had been carefully filtered through a series of four specially constructed filter boxes that removed fine particles down to a few hundredths of a micrometer in diameter. While most clinical laboratories adapted only some of Patterson's CIT biogeochemical laboratory design features and routine procedures, the high standards now being followed caused confidence levels in the data generated to be far higher than in previous decades. Patterson continued this clean laboratory crusade into the early 1980s, when at an international conference he stated boldly that a number of published researchers were "swimming in pools of industrial lead [where] they tend to make blind, incorrect, and insufficient subtractions for 'lead blanks' supposedly added to their sample during analysis. Usually they underestimate this laboratory contamination such that the unaccounted excess adds another unrecognized quantity about 500-fold greater than natural lead originally present in the samples."(57) By explicitly naming the names of researchers who had published sloppy work and others who had picked up on this flawed data and made sweeping generalizations concerning the environment, Patterson proved to be a powerful example within the scientific community of how it policed itself. Patterson's efforts, crucial to future analytical work aimed at very low levels, made working chemists far more aware of possible sources of unwanted contamination. These subsequent and at the same time difficult changes in laboratory practice were clearly recognized by policy makers who might not have fully understood the science related to lead in the environment, but who did connect these improvements with an enhanced confidence that analysts were getting their numbers right at last, or at least more precisely with far smaller standard deviations. One more factor played an important role in developing confidence in the blood lead
numbers generated by the late 1970s, and this was due to an intensive quality control effort on
the part of collaborating laboratories to check and cross-check analysts' procedures and methods
with the purpose of enhancing blood lead assay performance related to accuracy and precision.
Following a system similar to that which British laboratories had worked out beginning in the
late 1960s, the CDC, under legal mandate after 1978, organized a blood level proficiency testing
service in which 95 laboratories submitted data derived from a number of different methods,
including Delves cup, more traditional MIBK extraction, and graphite furnace AAS.(58) Based on
a fast turnaround time in issuing samples, evaluating and statistically processing results, and in
formulating a scoring system by which laboratories were rated, these external quality assessment
schemes became effective between 1979 and 1981, thus correcting what had been earlier
characterized as a "disastrous situation [in which] the experts were not very good at routine
work, [and] the routine laboratories could not reach the hitherto regarded as state of the art values
of the experts. The data obtained [was] usually far from the expected accuracy and precision
limits ."(59) Ultimately, then, questions about reliability of analytical methods were addressed, and
efforts shifted to improving methods of sampling and the establishment of standards, for no
sample of blood is known to be lead free.(60) When these techniques were combined, then, the
probability of measurement errors were reduced from an estimated high of 50% to far more
acceptable levels of between 7% and 15%.(61)
From Better Data to Public Policy It was thus propitious timing that in 1982 a research team that included Kathyrn R. Mahaffey and Joseph L. Annest published an important study in the New England Journal of Medicine on blood lead levels obtained from samples taken during the second National Health and Nutrition Survey, commonly referred to by the abbreviation NHANES II, collected between 1976 and 1980.(62) Using a modified Delves cup AAS method, these investigators had produced evidence not only that blood levels were uniformly higher than initially thought, but that their numbers, based on standard deviations, were at near unshakeable levels of confidence. And what this data suggested was that particularly in urban areas, blood levels in children, especially low income black children, were substantially higher than their white, suburban counterparts. Indeed, almost 1/5th of the black children living in major cities had blood levels exceeding 30 µg/dl, levels well beyond which a group of researchers including Herbert Needleman were claiming were significant in terms of increasing the risk of permanent learning and behavioral disorders.(63) And Needleman was now far from alone. By the middle 1980s AAS blood level determinations were used to study fundamental aspects of the mechanism of lead poisoning in the brain, relatively large populations of pregnant woman and their children after birth, and young children living near lead smelting operations. This unprecedented number of case studies and the ability to explore low levels with accuracy and precision opened a new round of scientific debates centered on implications that even very low levels caused low birth weights, cognitive and behavioral dysfunctions, as well as elevated blood pressure in adults.(64) While there was little immediate response from either executive or legislative branches of the federal government, a number of high level EPA administrators, recovering from their own series of internal scandals, moved decisively to follow up on this work by producing a four volume report entitled Air Quality for Lead that proved pivotal to legislation that followed.(65) The principal author of this document was Lester Grant, Director of the EPA's Environmental Criteria and Assessment Office. Included on a Lead Subcommittee were such influential figures as Johns Hopkins physician J. Julian Chisolm and Paul Hammond from the University of Cincinnati's Kettering Laboratory.(66) In terms of analytical procedures, the experts recommended Atomic Absorption Spectroscopy, a decision that was perhaps not a surprise since AA specialist Rodney Skogerboe from Colorado State University headed the group responsible for writing the section on sampling analytical methods. Additionally, Clair Patterson's influence on the report was clearly evident, particularly in terms of the importance of a clean laboratory.(67) But it was the conclusion that children had possible mental and behavioral deficits when exposed to lead even at levels between 15-30 µg/dl that was most radical. Thus, Grant concluded that even assuming a continuing decline in blood lead levels among the overall population due to the gradual phase-out of leaded gasoline, "Dose-population response information for heme synthesis effects, coupled with information from various blood lead surveys, eg. NHANES II study, indicate that large numbers of American children (especially low-income, urban dwellers) have blood levels sufficiently high (in excess of 15-20 µg/dl) that they are clearly at risk for deranged heme synthesis, and possibly, other health effects of growing concern as lead's role as a general systemic toxicant becomes more fully understood."(68) The 1986 EPA Air Quality Report set in motion further EPA efforts to influence both the
CDC and Congress. Despite the mounting evidence of the deleterious effects of lead on humans
even at very low exposures, the Regan administration was initially unresponsive. Indeed, a 1987
report contracted by the Public Health Service was so "red lined" by upper level administrators
as to cause its authors to resign in protest. Yet the AAS generated data was revealing. As Ellen
Slibergeld, chief toxic scientist for the Environmental Defense Fund and a member of the EPA's
Clear Air Advisory Committee stated, "We caught hold of the tail of the tiger, . . . But now we
are starting to see the rest of the beast, . . . And it is bigger than we thought." (69) Indeed, for many
scientists AAS results indicated that it was not just black children in the ghetto that were being
poisoned, but white middle class children and adults alike, as no one was safe from a ubiquitous
material found in the air, soil, water and interior of residences. And it was this conclusion was
the key to legislation that evolved.(70) Thus by 1991 the Center for Disease Control dropped the
minimum allowable threshold for lead in blood from 25µg/dl to 10 µg/dl, reflecting the notion
that lead was a hazard to children at any level and that it could be found virtually anywhere.
Further, the CDC called for a universal screening of young children in all U.S. communities.
Full Circle? Nevertheless, while policy was set in the early 1990s, and AAS data had played such a large role in setting the key criteria legislation, the story does not end on a happy note. Strategies that followed during the decade of the 1990s were not fully implemented, and indeed in many cases ignored, due to institutional obstacles, lack of genuine concern and funding constraints.(71) In 1998, the comprehensive program of 1991 was narrowed to discrete target areas, much to the disappointment of the activists like Herbert Needleman and Ellen Silbergeld, who maintained that universal testing was the only just route to follow. Indeed, Silbergeld's unheeded argument to continue casting the wide net for childhood lead poisoning was reinforced by her assertion that AAS methods gave the CDC an edge that one rarely has in the public health screening process: "The test for elevated BLL's is highly accurate: a positive venous blood lead test almost certainly indicates, at minimum, low level poisoning, . . . . [Further] 80% to 90% of laboratories participating in proficiency testing programs had results . . . that were within 4µg/dl of the actual level."(72) Clearly a commitment had been made to screen for prostate and colon cancer using far less reliable methods, but in the case of lead exposure and children the issue was rather quickly marginalized. Economics and cost benefit analysis, local physician disinterest in areas outside of the
large Eastern cities, HMO regulations that provided no incentives, and weak state and local
government administrations led to a half-hearted screening effort during which few gains were
made. In Baltimore, the pioneer of childhood lead poisoning diagnosis and prevention programs,
the entire monitoring system was found to be in utter shambles at the turn of the century.(73) In a
fashion no different from the early 1950s or 1970s, politicians continued to spout posturing
rhetoric, but their words only clouded a tragedy. It seemed that more than sixty years of
government activity on both the federal and local level had resulted in nothing enduring.
Enforcement in Baltimore had lapsed, and several thousand children, now under the 10µg/dl
guideline, were in harms way. As late as spring of 2000 new state and federal funds have been
directed to the problem, but only time will tell if the childhood lead poisoning problem will be
finally resolved. Certainly the science was and is there to reveal the beast, but will the beast be
tamed?
Conclusion In a recent essay, Christian Warren has argued that the lead paint poisoning issue was firmly rooted in the Progressive Era and that forces of continuity have had a powerful effect on what followed after the 1920s.(74) For Warren and others the story was essentially over by the middle of the 1920s, as politics and economics, along with manufacturer and consumer attitudes crystallized and fundamentally remained static for the next several decades. In the meantime, the lead problem went "round and round." This interpretation, however, fails to adequately deal with changes in related science and technology. If we focus on these two elements within a complex political, social and economic system, the story as it unfolded during the 20th century was hardly over by the mid-1920s. Changes in analytical chemistry, led to significant consequences in a broader sphere.(75) Indeed, an instrument -- AAS -- and instrumental chemical analysis in general played critical roles in shifting policy associated with one of the most significant environmental and public health questions of the late 20th century. Historians of science, and more specifically historians of chemistry, have often relegated the field of analytical chemistry to the status of a poor stepchild when compared to the far more glamorous areas of physical, organic, and even inorganic chemistry. Yet, in an age in which environmental concerns now rank high on our list of societal priorities and when bureaucracies are more dominant than ever, numbers (and particularly good numbers derived from methods that ideally aim to reach absolute analysis) possess an inherent power to influence, but not dictate, the course of social and political history. END NOTES 1. Edward Tenner, Why Things Bite Back: Technology and the Revenge of Unintended Consequences (New York, 1996), pp. 3-24. In his opening chapter, Tenner briefly discusses the emergence during the 1920s of the term "foolproof," and he provides examples of a number of technologies that later defied the products' purported claims. White lead carbonate, one of the most important of the lead paint pigments, was precisely touted as such a substance and thus earned loyalty from both consumers and professional painters. 2. To my knowledge, this essay represents the first scholarly exploration into the history of the analytical chemistry associated with the lead paint/childhood lead poisoning topic. Sources that provide historiographical background for this work include Samuel P. Hays, "The Role of Values in Science and Policy: the Case of Lead," in Explorations in Environmental History (Pittsburgh, 1998), pp. 291-311. I would contend that Hays' model of linking scientists' values with their institutional affiliations is far too simplistic. A recent publication that deals not with childhood lead poisoning but that of workers, particulary in the period 1910 to 1925, is Christopher Sellers, Hazards of the Job: From Industrial Disease to Environmental Science (Chapel Hill, 1997). D.E. Jacobs, "Lead-Based Paint as a Major Source of Childhood Lead Poisoning: A Review of Evidence," in Michael E. Beard and S. D. Allen Iske, eds., Lead in Paint, Soil and Dust: Health Risks, Exposure Studies, Control Measures and Quality Assurance, ASTM STP 1226 (Philadelphia, 1995), pp. 175-187, contains a number of factual inaccuracies, including a description of international and foreign legislation to ban white lead during the 1920s. Peter Reich's The Hour of Lead: A Brief History of Lead Poisoning in the United States, Over the Past Century and of Efforts by the Lead Industry to Delay Regulation (Washington, D.C., 1992), was written with the express purpose of attempting to influence litigation and legislation. All of these sources are to varying degrees in debt to William Graebner, "Private Power, Private Knowledge and Public Health: Science, Engineering, and Lead Poisoning, 1900-1970," in R. Bayer, The Health and Safety of Workers (New York, 1988), pp. 15-71, who argued in a manner similar to that of David Noble in America by Design (New York, 1978) that scientists and engineers were coopted by the capitalist system and thus constrained in their autonomy. More useful in the reconstruction of a balanced account are: Jacqueline K. Corn, "Historical Perspective to a Current Controversy on the Clinical Spectrum of Plumbism," Health and Society (Winter, 1975), pp.93-114 and Marjorie Smith, "Lead in History," in Richard Landsdown and William Yule, eds., Lead Toxicity: History and Environmental Impact (Baltimore, 1986), pp. 7-24. 3. On this controversial matter from the perspective of housing rather than childhood poisoning, see Environmental Protection Agency, Lead In Your Home: A Parent's Reference Guide (Washington, D.C., 1998); EPA, Testing Your Home for Lead in Paint, Dust and Soil (Washington, D.C., 1998); U.S. Department of Housing and Urban Development, Office of Policy Development and Research, Comprehensive and Workable Plan for the Abatement of Lead-Based Paint in Privately Owned Housing: Report to Congress (Washington, D.C., 1991); Cassandra Chrones Moore, Haunted Housing: How Toxic Scare Stories are Spooking the Public out of House and Home (Washington, 1997); and Ellen Ruppel Shell, "An Element of Doubt," Atlantic Monthly (December, 1995), 24, 26, 28, 36, 38-9. 4. For example, see Robert Bud and Deborah Jean Warner, eds., Instruments of Science: An Historical Encyclopedia (New York, 1998); Nicolas Rasmussen, Picture Control: The Electron Microscope and the Transformation of Biology in America, 1940-1960 (Stanford, CA, 1997); Yorgos Goudaroulis, "Can the History of Instrumentation Tell Us Anything About Scientific Practice?" in Kostas Gavroglu, ed., Trends in the Historiography of Science (Dordrecht, 1994); Peter Morris, "Development of High Resolution NMR Spectroscopy as a Structural Tool," in Robert Bud, ed., Invisible Connections: Instruments, Institutions, and Science (Bellingham, Washington, 1992); Lissa Roberts, "Condillac, Lavoisier, and the Instrumentalization of Science," Eighteenth Century: Theory and Interpretation, 33 (1992), pp. 252-271; Jeffrey L. Sturchio, "Arnold Beckman and the Revolution in Instrumentation: Building a Business of Science," Today's Chemist, 1 (1988), 16, 18, 31; and Yakov M. Rabkin, "Technological Innovation in Science: The Adoption of Infrared Spectroscopy by Chemists," Isis, 78(1987), 31-54. 5. For a survey on the 20th century history of instrumentation, see Frank Greenaway, "Instruments," in Trevor I. Williams, ed., A History of Technology, Volume VII (Oxford, 1978), pp. 1204 - 1219; Adele E. Clarke and Joan H. Fujimura, eds., The Right Tools for the Job: At Work in the Twentieth Century Life Sciences (Princeton, NJ, 1992); Davis Baird, "Analytical Chemistry and the 'Big' Scientific Instrumentation Revolution," Annals of Science, 50 (1993), 267-290. Perhaps the best known example of how instrumentation influenced the course of recent environmental policy is that of James E. Lovelock and his electron capture detector, the data from which proved crucial to the CFC accumulation in the atmosphere controversy. 6. This case study clearly reflects the transition from a modern industrial to a post-modern risk society that occurred in the last 30 to 40 years of the twentieth century as characterized in Ulrich Beck's Risk Society: Towards a New Modernity, trans.by Mark Ritter (London, 1992). 7. I am indebted to Dr. Peter C. English of the Duke University School of Medicine for this insight. 8. Center for Disease Control, Preventing Lead Poisoning in Young Children: a Statement by the Center for Disease Control, CDC, Atlanta, 1991; taken from Erik Millstone, Lead and Public Health: the Dangers for Children (Washington, 1997), p.17. On the significance of these thresholds, see Liora Salter, Mandated Science: Science and Scientists in the Making of Standards (Dordrecht, 1988), pp. 12-14. It is beyond the intention and the scope of this paper to examine the nature of the health risks posed by lead exposure at different levels. My major point is that until the analytical procedures were in place, large scale population studies were unfeasible. 9. S.R. Koirtyohann, "Atomic Absorption Spectrometry from an Academic Perspective," Analytical Chemistry, 63 (1991), 1025A. 10. G.E.F. Lundell, "The Chemical Analysis of Things as They Are," Industrial and Engineering Chemistry, Analytical Edition, 5 (1933), 221. 11. Hellmut Fischer, "Die Metallverbindungen des Diphenylthiocarbazons und ihre Verwendbarkeit für die chemische Analyse," Veröffentlich. Seimens-Konzern, 4 (1925), 158-170; Über den Nachweis von Schwermetallen mit Hilfe von "Dithizon," Zeitschrift für Angewante Chemie, 42 (1929), 1025-1027. Subsequent papers that mark the historical evolution of the method with respect to lead include: E.S. Wilkins, Jr., C.E. Willoughby, E.O. Kraemer, "Determination of Minute Amounts of Lead in Biological Material. A Titrimetric-extraction Method," Industrial and Engineering Chemistry, Analytical Edition, 7 (1935), 285; C.E. Willoughby, E.S. Wilkins, Jr., and E.O. Kraemer, "Determination of Lead. Removal of Bismuth Interference in the Dithiozone Method," Industrial and Engineering Chemistry, 7 (1935), 285; Herman A. Liebhfsky and Earl H. Winslow, "Diphenylthiocarbazone (Dithizone) as an Analytical Reagent," Journal of the American Chemical Society, 59 (October, 1937), 1966-1971; H. KraftStrom, K. Wulfert and O. Syndes, "Determination of Lead in Whole Blood," Biochemisches Zeitschrift, 290 (1937), 382; and C.E. Willoughby and E.S. Wilkins, Jr., "The Lead Content of Human Blood," Journal of Biological Chemistry, 124 (1938), 639. 12. Liebhafsky and Winslow, "Diphenylthiocarbazone (Dithizone) as an Analytical Reagent," 1968-9. 13. For example, see K. Bambach and R.E. Burkey, "Microdetermination of Lead by Dithizone. With an Improved Lead-Bismuth Separation," Industrial and Engineering Chemistry, 14 (1942), 904; J. Schultz and M.A. Goldberg, "Photometric Routine Estimation of Traces of Lead by Dithizone," Industrial and Engineering Chemistry, 15 (1943), 155; Lawrence T. Fairhall, Industrial Toxicology, (Baltimore, 1949), pp. 96-99; J. Cholak, D.M. Hubbard and R.E. Burkey, "Microdetermation of Lead in Biological Material. With Dithizone Extraction at High pH," Analytical Chemistry, 20 (1948), 671; and Methods for Determining Lead in Air and Biological Materials (New York, 1955), p. 8. 14. Koirtyohann, p.1030A. 15. J. Cholak, "Comments on Report LTD-45-57 (December 14, 1945)," in Robert A. Kehoe Archives, Cincinnati Medical Heritage Center, Box 19, File Miscellaneous Data -- Lead. There exists considerable textual material in the Kehoe Archives comparing dithizone and arc spectroscopy data during the 1930s and 1940s. For example, see Box 19, File 3, "Lead in Urine and Blood of Children;" D.M. Hubbard to Dr. Kehoe, January 11, 1937, Box 37, File Reagents, Equipment, etc. Used in Lead Analyses Methods; "Lead Studies on Blood and Urine from Employees of the Lowe Brothers Company, Dayton Ohio," n.d. [1940], Box 37, File The Lowe Brothers Company Dayton Ohio (Correspondence, Analytical Data, and Reports). 16. "Report of the Subcommittee on Laboratory Facilities for Diagnosis of Poisonings," n.d. [1952?]. 17. K. Muller and P. Pringsheim, Die Naturwissenschaften, 18 (1930), 364; T.T. Woodson, Review of Scientific Instruments, 10 (1938), 308. 18. A. Walsh, "The Application of Atomic Absorption Spectra to Chemical Analysis," Spectromchimica Acta, 7 (1955), 108. 19. J.W. Robinson, "Atomic Absorption Spectroscopy," Analytical Chemistry, 32 (1960), 17A-29A. 20. Walter Slavin, "Atomic Absorption Spectroscopy: Why Has It Become Successful?" Analytical Chemistry, 63 (1991), 1037A. 21. By 1970 Perkin-Elmer's success with the 303 AAS design completely dominated a global market. Only a few years earlier and prior to an industry shakeout, serious competition included the Australian firm Techtron, English firms Hilger & Watts, Unicam, Southern Analytical and Evans Electroselenium, and American competitors Beckman, Jarell-Ash, and Optica. See Ivan Rubeka and Bedich Moldan, Atomic Absorption Spectrophotometry, trans. P.T. Woods (Cleveland, 1969), pp.78-79. 22. Conversation with Walter Slavin, June 30, 2000. 23. Nature, 192 (1961), 929. 24. J.B. Willis, "Determination of Lead and Other Heavy Metals in Urine by Atomic Absorption Spectroscopy," Analytical Chemistry, 34 ( 1962), 614-617. 25. The first article on childhood lead poisoning in the U.S. was written by Johns Hopkins pediatrician Kenneth D. Blackfan, "Lead Poisoning in Children with Especial Reference to Lead as a Cause of Convulsions," Am. J. Med. Sci., 153 (1917), 877-887. Huntington Williams' career-long campaign is well described in Elizabeth Fee, "Public Health in Practice: An Early Confrontation with the 'Silent Epidemic' of Childhood Lead Poisoning," Journal of the History of Medicine and Allied Sciences, 45 (1990), 570-606. The complexity of the post-WWII Baltimore community is suggested in Martin Millspaugh and Gurney Breckenfeld, The Human Side of Urban Renewal (Baltimore, 1958). 26. The development of this method can be directly traced to the outbreak of a lead poisoning epidemic in Baltimore during 1932 due to the burning of battery casings. See P.G. Shipley, T.F. McNair Scott and H. Blumberg, "The Spectrographic Detection of Lead in the Blood as an Aid to the Clinical Diagnosis of Plumbism," Bulletin of the Johns Hopkins Hospital, 51 (1932), 327-8; Harold Blumberg and T.F. McNair Scott, "The Plasma-Cell Partition of Blood Lead in Clinical Lead Poisoning," Bulletin of the Johns Hopkins Hospital, 56 (1935), 311-316; Harold Blumberg and T.F. McNair Scott, "The Quantitative Spectrographic Estimation of Blood Lead and its Value in the Diagnosis of Lead Poisoning," Bulletin of the Johns Hopkins Hospital, 56 (1935), 276-293. 27. Emanuel Kaplan and John M. McDonald, "Blood Lead Determinations as a Health Department Laboratory Service," American Journal of Public Health, 32 (1942), 481-485; John M. McDonald and Emanuel Kaplan, "Incidence of Lead-Poisoning in the City of Baltimore," Journal of the American Medical Association, 119 (1942), 870-872. 28. "Chronology of Lead Poisoning Control, Baltimore, 1931-1971," Baltimore Health News, 48 (December, 1971), 34 ff. 29. James O. Pierce II, and Jacob Cholak, "Lead, Chromium and Molybdenum by Atomic Absorption," Archives of Environmental Health, 13 (1966), 208-212. 30. Concurrent efforts at Perkin-Elmer to market infrared spectrometers are well described in: Rabkin, pp. 46-48; S. Sprague and W. Slavin, "A Simple Method for the Determination of Lead in Blood," Atomic Absorption Newsletter, 5 (1966), 9; and Perkin-Elmer Corporation, Analytical Methods for Atomic Absorption Spectrometry (Norwalk, CT, 1964-), Pb 5, Looseleaf, Revised May, 1966. Subsequently, Slavin's method was improved upon by D.W. Hessel, "A Simple and Rapid Quantitative Determination of Lead in Blood," Atomic Absorption Newsletter, 7 (1968), 55-6. Despite all of the claims made by analytical chemists, it seems that few procedures could be duplicated with the accuracy and precision claimed in publications; thus, it is no surprise that as late as 1973 a National Academy of Sciences report called for the development of "better analytic methods to measure lead in . . . body tissues, especially blood." See typescript copy, National Academy of Sciences, "Report of the Ad Hoc Committee to Evaluate the Hazard of Lead in Paint," Prepared for the Consumer Product Safety Commission, November, 1973, p. 35. 31. "Lead Poison Cases Worst Ever at 260 Cases," New York Times, May 12, 1970; Sol Blumenthal, Bernard Davidow, David Harris, and Felicia Oliver-Smith, "A Comparison Between Two Diagnostic Tests for Lead Poisoning," American Journal of Public Health, 62 (1972), 1060-1064. 32. Dr. Bernard Davidow, Chief, Food & Drug Laboratory, to Dr. Morris Schaeffer, Assistant Commissioner, NYC, March 7, 1969; Vincent F. Guinee, "The Position of New York City on the Level of Appropriations for Control of Childhood Lead Paint Poisoning," typescript, submitted to Sub-Committee on Labor-HEW of the House Appropriations Committee, June 16, 1971. I am very skeptical of the 84,000+ assay figure, but I have not seen evidence to contradict this figure. If true, I would also wonder about the quality of this work, given the fact that so much of the blood lead data from this period was later called into question or discredited. 33. 92nd Cong. 1st Session, H.R. 2627. 34. "Lead-Based Paint Poisoning: An Example of Administration," Congressional Record -- House, June 1, 1971; Fred Loetterle, "Manes Launches Lead Poison Probe," New York Daily News, 1971. 35. H.T. Delves, "A Micro-sampling Method for the Rapid Determination of Lead in Blood by Atomic-absorption Spectrophotometry," The Analyst, 95 (1970), 431-438; William F. Barthel, Ann L. Smrek, Gloria P. Angel, John A. Liddle, Philip J. Landrigen, Stephen H. Gehlbach, and J. Julian Chisolm, "Modified Delves Cup Atomic Absorption Determination of Lead in Blood," Journal of the Association of Official Analytical Chemists, 56 (1973), 1252-6. 36. Albert Chang, George W. Schucker, Elkins Dahle, William Smith, Emanuel Kaplan, Dove Cresswell and Joseph Gordon, "Lead Poisoning Screening in Child Health Clinics," unpublished paper, presented at the Annual Meeting of the American Public Health Association, November 7, 1973. 37. Frank J. Fernandez and Herbert L. Kahn, "The Determination of Lead in Whole Blood by Atomic Absorption Spectrophotometry with the "Delves Sampling Cup Technique," Atomic Absorption Newsletter, 10 (January-February 1971), 4. 38. On this episode, see Boris V. L'vov, "A Personal View of the Evolution of Graphite Furnace Atomic Absorption," Analytical Chemistry, 63 (1991), 924A-931A; 39. On the history of Perkin-Elmer and its German subsidiary, see Thomas P. Fahy, Richard Scott Perkin and the Perkin-Elmer Corporation (N.P., 1987), pp. 113-23. On current developments at P-E, see Phillip E. Ross, "The Making of a Gene Machine," Forbes (February 21, 2000), 98-104. 40. Eleanor Berman, "The Challenge of Getting the Lead Out," in LaFleur, ed. Accuracy in Trace Analysis: Sampling Sample Handling, Analysis (Washington, 1974), II, p. 716. One serious difficulty when using the graphite furnace as originally designed was discovered in the early 1970s with the possible formation of nonvolatile compounds, including spinels, carbides, nitrides, or phosphates as a result of the reaction of metallic elements with other substances in the initial matrix. See G. Tolg, Talenta, 21 (1974), 327. This problem was later largely overcome by the addition of a matrix modifier to the sample. 41. Walter Slavin and D.C. Manning, "The L'vov Platform for Furnace Atomic Absorption Analysis," Spectrochimica Acta, 35B (1980), 701-714. 42. Perkin-Elmer Corporation, Techniques in Graphite Furnace Atomic Absorption Spectrometry (Ridgefield, CT, 1985), pp. 134, 196. 43. See D.C. Manning and Walter Slavin, "Determination of Lead in a Chloride Matrix with the Graphite Furnace," Analytical Chemistry, 50 ( 1978), 1235-1238; J. Fernandez, S.A. Myers, and Walter Slavin, "Background Correction in Atomic Absorption Utilizing the Zeeman Effect," Analytical Chemistry, 52 (1980), 741-746; E. Pruszkowska, G.R. Carnrick, and W. Slavin, "Blood Lead Determination with the Platform Furnace Technique," Atomic Spectroscopy, 4 (March-April, 1983), 59 - 61; W. Slavin and G.R. Carnick, "Interferences in Graphite Furnace AAS Continuum Background Correction. A Survey," Atomic Spectroscopy, 7 (January-February, 1986), 9-13: Walter Slavin, "Atomic Absorption Spectrometry," in James F. Riordan and Bert L. Vallee (eds.), Metallobiochemistry, Part A ,Volume 158 (San Diego, 1988), pp. 117-145. 44. Takao Watanbe, et. al., "A Semiautomated System for Analysis of Metals in Biological Materials and its Application to Mass Determination of Cadmium in Blood," Toxicology Letters, 37 (1982), 231-238; Takao Watanabe, Hiroyshi Fujita, et. al., "Baseline Level of Blood Lead Concentration Among Japanese Farmers," Archives of Environmental Health, 40 (1985), 170-176. 45. J.A. Holcombe and J.A. Hassell, "Absolute Analysis," Analytical Chemistry, 62 (1990), 169R. 46. Elaine W. Gunter, Brenda G. Lewis, and Sharon M. Koncikowski, Laboratory Procedures Used for the Third National Health and Nutrition Examination Survey (NHANES III), 1988-1994 (Hyattsville, MD, 1996), pp. VII-H-1 - VII-H-15, in NHANES III Reference Manuals and Reports, October, 1996, CD ROM HE 20.7039/2: M31/CD. 47. On the polarographic method for determining lead in blood, see I.M. Koltoff and James J. Lingane, Polarography (New York, 1952), II, 531-532. On the history of ASV, see Janet Osteryoung and Carolyn Wechter, "Development of Pulse Polarography and Voltammetry," in John Stock and Mary Virginia Orna, eds., Electrochemistry, Past and Present (Washington, D.C., 1989), pp. 380-395. 48. Barker's published papers remain to be carefully examined. They include G.C. Barker and I.L. Jenkins, AERE C/R 924 (British Government Report, 1952); G.C. Barker and I.L. Jenkins, Analyst, 77(1952), 685-696; G.C. Barker, CRC-440 (Canadian Government Report, 1950); G.C. Barker, AERE C/R 1553 (British Government Report, 1954); G.C. Barker and D.R. Cockbaine, AERE C/R 1404 (British Government Report, 1957); G.C. Barker, R.L. Faircloth and A.W. Gardner, AERE C/R 1786 (British Government Report, 1958). 49. E.P. Parry and R.A. Osteryoung, "Evaluation of Analytical Pulse Polarography," Analytical Chemistry, 37 (1965), 1634-1637. The fascinating story of Princeton Applied Research and its efforts to design and market a variety of instruments including the PAR 170 is recounted by Jud Flato in Herbert A. Laitinen and Galen W. Ewing, A History of Analytical Chemistry (Washington, 1977), pp. 274-279; quote taken from p. 278. 50. On comparisons, see Joe Boone, Tom Hearn and Sue Lewis, "Comparison of Interlaboratory Results for Blood Lead with Results from a Definitive Method," Clinical Chemistry, 25 (1979), 389-393. One hundred and thirteen laboratories compared 6 different methods using an NBS lead in blood reference with the conclusion that in terms of accuracy extraction AAS was best, followed by ASV, Graphite Furnace AAS and Delves Cup AAS; however, ASV was the most precise of all procedures examined. See also H.W. Nurnberg, "A Critical Assessment of the Voltammetric Approach for the Study of Toxic Trace Metals in Biological Specimens," in W. Franklin Smyth (ed.), Electroanalysis in Hygiene, Environmental, Clinical and Pharmaceutical Chemistry (Amsterdam, 1980), p. 357; H.W. Nurnberg, "Potentialities and Applications of Voltammetry in the Analyses of Toxic Trace Metals in Body Fluids," in S. Facchetti, (ed.), Analytical Techniques for Heavy Metals in Fluids (Amsterdam, 1983), p. 212; M. Stoeppler, "Atomic Absorption Spectrometry -- A Valuable Tool for Trace and Ultratrace Determinations of Metals and Metalloids in Biological Materials," Spectrochimica Acta, 38B (1983), 1559; National Academy of Sciences, Committee on Measuring Lead in Critical Populations, Measuring Lead Exposure in Infants, Children and other Sensitive Populations (Washington, 1993), p. 198. 51. My thanks to University of Dayton colleague Gerald Keil for calling my attention to this point. This assertion is supported in Janet Osteryoung, "Developments in Electrochemical Instrumentation," Science, 218 (1982), 264. 52. Sandy Roda, Robert Greenland, Robert Bornschein, and Paul Hammond, "Anodic Stripping Voltammetry Procedure Modified for Improved Accuracy of Blood Lead Analyses," Clinical Chemistry, 34 (1988), 563-567. 53. On this topic, see Morris Zief and James W. Mitchell, Contamination Control in Trace Element Analysis (New York, 1976). 54. Clair C. Patterson and Dorothy Settle, "The Reduction of Orders of Magnitude Errors in Lead Analyses of Biological Materials and Natural Waters By Evaluating and Controlling the Extent and Sources of Industrial Lead Contamination Introduced During Sample Collecting, Handling, and Analysis," in Philip D. LaFleur, ed., Accuracy in Trace Analysis: Sample, Sample Handling, Analysis (Washington, 1974), I, p. 322. 55. On the issue of computers and their impact on the nature of work, including the issue of deskilling, see Shoshana Zuboff, In the Age of the Smart Machine: The Future of Work and Power (New York, 1988). 56. Patterson and Settle, pp. 337-346. 57. Clair C. Patterson, "British Mega Exposures to Industrial Lead," in Michael Rutter and Robin Russell Jones, eds., Lead Versus Health: Sources and Effects of Low Level Lead Exposure (Chichester, 1983), p.27. 58. On the UK's External Quality Assessment Scheme for General Clinical Chemistry and its statistical methods based on mean running variance index score (MRVIS), see Iain L. Marr and Malcolm Cresser, Environmental Chemical Analysis (Glascow, 1983), p.246; specific efforts in the UK related to lead in blood standards are discussed in H.T. Delves, "Use Reference Samples Rather than Reference methods," Analytical Proceedings, 21 (1984), 391-394. CDC efforts are traced in "OSHA Criteria for Laboratory Proficiency in Blood Lead Analysis," Archives of Environmental Health, 37 (1982), 58-60, and Roger L. Boeckx, "Lead Poisoning in Children," Analytical Chemistry, 58 (1986), 275A-287A. 59. M. Stoeppler and H. Nurnburg, "Critical Review of Analytical Methods for the Determination of Trace Elements in Biological Materials," in A. Berlin, A.H. Wolff and Y. Hasegawa, (eds.), The Use of Biological Specimens for the Assessment of Human Exposure to Environmental Pollutants (The Hague, 1979), p. 326. See also, Joe Boone, Tom Hearn, and Sue Lewis, "Comparison of Interlaboratory Results for Blood Lead with Results from a Definitive Method," Clinical Chemistry, 25 (1979), 389-393. 60. An area of future work is planned around the topic of the setting of standards by the NBS related to lead in blood as well as lead in paint manufacturing and in the environment. This study should complement a number of the essays contained in M. Norton Wise, The Values of Precision (Princeton, 1995). The first attempt on the part of the National Bureau of Standards to develop a lead in bovine blood standard is described in D.A. Becker and P.D. LaFleur, "Production and Certification of NBS Standard Reference Materials," in Delbert D. Hemphill (ed.), Trace Substances in Environmental Health -- IV (Columbia, MO, 1971), pp. 433-435. It was only during the early 1980s that commercial reference standards were marketed by the German firm Behring, which introduced lyophilized bovine blood with assigned values for lead, cadmium and mercury. See J. Angerer, K.H. Schaller and R. Heinrich, Arbeitsmedin, Sozialmedicin, und Praventivmedicin, 115 (1981). The importance of NBS blood lead standards to the development of the definitive analytical method as used in NHANES III (some 60,000 samples collected between 1988 and 1994) is described in Dayton T. Miller, Daniel D. Paschal, Elaine W. Gunter, Phillip E. Stroud and Joseph D'Angelo, "Determination of Lead in Blood using Electrothermal Atomisation Atomic Absorption Spectrometry with a L'vov Platform and Matrix Modifier," The Analyst, 112 (1987), 1701 - 1704. 61. Joseph L. Annest, "Trends in the Blood Levels of the US Population: The Second National Health and Nutrition Examination Survey (NHANES II) 1976 - 1980," in Michael Rutter and Robin Russell Jones, Lead Versus Health: Sources and Effects of Low Level Lead Exposure (Chichester, 1983), p. 56. 62. Kathryn R. Mahaffey, Joseph L. Annest, Jean Roberts, and Robert S. Murphy, "National Estimates of Blood Lead Levels, United States, 1976-1980, Associated with Various Demographic and Socioeconomic Factors," New England Journal of Medicine, 307 (September 2, 1982), 573-579. To date there have been three NHANES surveys, during which the federal government officially sampled the blood of its citizens, the first conducted between 1971 and 1973 (NHANES I), the second taking place between 1976 and 1980 (NHANES II) and the third sample collection occurring between 1993 and 1995 (NHANES III). Only the last two sampled blood for lead, and while NHANES III indicated that a general decline had taken place in lead in blood levels, there still remain issues of hazardous exposure. See Milstone, Lead and Public Health, pp.75-77. 63. The entire neurotoxicity controversy can be explored in Millstone, Lead and Public Health, pp. 24-73. Undoubtedly, this episode will become the basis for someone from a future generation writing a dissertation in the history of science. 64. "New Guidelines for Lead Issued," New York Times, February 8, 1985, 18; Marjorie Anders, "675,000 U.S. Children Suffer from Man-Made Disease: Lead Poisoning," Los Angeles Times, November 24, 1985, p. 5; J. Raloff, "Signs of How Lead Toxicity Begins," Science News, 130 (July 26, 1986), 54; "Even Low Lead levels in Mom Affect Baby," ibid., 130 (September 13, 1986); "Excess Lead: Its Evolving Definition," ibid., 130 (November 22, 1986), 333. 65. United States Environmental Protection Agency, Air Quality Criteria for Lead, 4 vols., June 1986, documents EP1.23/9:600/8-83/028AF-DF. 66. Ibid., I, iii. 67. Ibid., I, p.63; II, p. 4-19. 68. Ibid., I, p.160. 69. Michael Weisskopf, "Authors protest Report on Lead Poisoning; Researchers Resign, Call Charges Misleading," Washington Post, June 13, 1987, A14; "Persistent, Pervasive Pollutant; Found in Soil, Water and Food, Lead Poisoning Seen as Health Peril," ibid., June 15, 1987, A4. 70. "Low Lead Levels Can Impair Kids, New Study Says," Toronto Star, August 26, 1988, B3. On the legislative history, see Centers for Disease Control, Strategic Plan for the Elimination of Childhood Lead Poisoning (Washington, 1991), Appendix III, pp. 1-2. 71. Sally Squires, "Lead Exposure Threshold Reduced," Washington Post, October 8, 1991, Z8; Phillip J. Hilts, "Lower Lead limits Are Made Official," New York Times, October 8, 1991, C3; Gerry Sikorski, "Posturing About Lead Poisoning," Washington Post, October 21, 1991, A19; Steven Lee Myers. A history of the failure of universal screening is Herbert Needleman, "Childhood Lead Poisoning: The Promise and Abandonment of Primary Prevention," American Journal of Public Health, 88 (1998), 1871 - 1877. 72. See especially Eric W. Manheimer and Ellen K. Silbergeld, "Critique of CDC's Retreat from Recommending Universal Lead Screening for Children, Public Heath Reports, 113 (1998), 39-46, and the CDC's response to this criticism by Nancy M. Tips, Henry Falk and Richard J. Jackson, "CDC's Lead Screening Guidance: A Systematic Approach to More Effective Screening," Public Health Reports, 113 (1998), 47-51. Herbert Needleman, "Childhood Lead Poisoning: The Promise and Abandonment of Primary Prevention," American Journal of Public Health, 88 (December 12, 1998), 1871-1877. 73. Jim Haner, Offensive Against Lead Paint Promised," Baltimore Sun, January 9, 2000, 1A; Timothy B. Wheeler and Jim Haner, "$50 Million Pledged to Fight Lead Poisoning," Baltimore Sun, January 29, 2000, 1A; Timothy Wheeler, "State, City, Propose Bill to Require Lead Testing," Baltimore Sun, February 12, 2000, 2B; Ivan Penn and Jim Haner, "Plan Calls for Stricter Lead Test Standards," Baltimore Sun, February 29, 2000, 3B; Gerald Shields, "Lead Fight Takes Wide Approach," Baltimore Sun, March 23, 2000, 1B; "Promising a New Start on Lead Poisoning," Baltimore Sun, April 13, 2000, 24A. 74. Christian Warren, "Toxic Purity: The Progressive Era Origins of America's Lead Paint Poisoning Epidemic," Business History Review, 73 (1999), 705-736; Gerald Markowitz and David Rosner, "'Cater to the Children': The Role of the Lead Industry in a Public Health Tragedy, 1900-1955," American Journal of Public Health, 90 (2000), 36-46. My use of the phrase "round and round" comes from the title of an essay that Warren draws on, Barbara Berney, "Round and Round it Goes: The Epidemiology of Childhood Lead Poisoning, 1950-1990," Milbank Quarterly, 71 (1993), 3-39. 75. The notion of science and technology at Western civilization's center is discussed in Ian Varcoe and Steven Yearly, "Introduction, The Centrality of Science and Technology," in Varcoe, Maureen McNeil, and Yearly (eds.), Deciphering Science and Technology: The Social Relations of Expertise (London, 1990), pp. 1-28. |
|