by The Pandemic Century- One Hundred Years of Panic, Hysteria
By the end of August, EIS officers had combed the hotel from top to bottom. Samples removed to Atlanta for testing included Freon 11 and chilled water from the air conditioner system; dust from the AHUs, carpets, draperies, and hotel elevators; water from the hotel drinking fountains and ice dispensers; rodent control chemicals; bleaches and housekeeping supplies; and a variety of convention mementos including mugs, hats, badges, and packs of Merit cigarettes that had been included in the convention gift bags. Noticing that ventilation grilles from the subway discharged onto Broad Street, Fraser had also ordered inspections of the underground concourse. Finally, mindful that no epidemiological survey could be considered complete without a record of the weather, Fraser ordered up meteorological readings for the period from July 21 to 25. These showed that the convention had opened to sweltering conditions and that on July 22 there had been a sharp temperature inversion. The result was that rather than temperatures decreasing with elevation above ground level, as is usual, air at the upper levels of the hotel, including the roof, had become superheated. This unusual effect had lasted for a day and a half, ending at around noon on July 24, and, Fraser discovered, had been accompanied by slightly higher levels of carbon monoxide and other atmospheric pollutants.
From the beginning, one of the most popular theories was that the outbreak had been due to psittacosis. Although in 1976 the great parrot fever pandemic of 1930 was a distant memory, ornithologists and veterinary specialists had continued to study the epidemiology of the disease and its natural history. As tighter regulation reduced the incidence of outbreaks in bird breeding establishments and pet stores, the focus had shifted to occupational settings, such as turkey farms and poultry processing plants. At the same time, serological studies and a better understanding of the role of latent infections had brought a new appreciation of the disease’s wide host range. Indeed, in 1967, Karl Meyer had tabulated 130 species of bird that carried the disease. These included homing pigeons raised in backyard lofts and the pigeons in New York’s Central Park, half of whom were found to be harboring the chlamydia bacterium.
Fraser noted that the upper floors and roof of the Bellevue were popular pigeon roosts. In addition, a local Philadelphia character, known as the “pigeon lady,” had been seen scattering bread crumbs on Broad Street. Then, there was the report from a guest that she had heard a parakeet chirping in one of the rooms. Fraser’s task was not made any easier by the support given to the psittacosis theory by prominent medics. The most vocal was Dr. Gary Lattimer, a specialist in infectious diseases at the Sacred Heart Hospital in Allentown. In early August, Lattimer had examined four Legionnaires and, believing they had psittacosis, treated them with tetracycline, a broad-spectrum antibiotic that was known to be effective against psittacosis and rickettsial diseases.* The Legionnaires’ symptoms had immediately improved, prompting Lattimer to urge Fraser to issue a directive recommending tetracycline to other patients. Fraser refused, citing the absence of scientific proof and saying it would be irresponsible to recommend tetracycline over erythromycin and rifampicin. However, Lattimer would not back down. Instead he began holding press conferences and writing to well-known chlamydia experts, including Julius Schachter, Meyer’s former pupil and a professor of epidemiology at UCSF. In support of his theory, Lattimer cited the fact that psittacosis had a variable incubation period of three to eleven days, similar to the two to ten days’ period for Legionnaires’ disease. The mortality rate and symptoms were also similar, and, as with psittacosis, there appeared to be no secondary transmission. Finally, Lattimer pointed out that histopathological examinations revealed extensive alveolitis or inflammation of the air sacs of victims’ lungs. Together with the changes seen in the liver and spleen, these were “compatible in all aspects with those reported from previous human chlamydial epidemics.” Unfortunately for Lattimer, in September a panel of expert pathologists tasked with reviewing the autopsy evidence disagreed. Although the panel found that five of the core Legionnaires cases and the three Broad Street pneumonias showed patterns of “acute diffuse alveolar damage,” it ruled that such alveolar damage could also be the result of exposure to toxins. “No pathologic diagnosis could be made on the basis of these findings,” they concluded.
All hope of solving the mystery now rested on the microbiology studies. As the leading federal agency for disease control and a WHO reporting center for influenza, the CDC’s laboratories in Atlanta were considered second to none. Staffed by 625 scientists and technicians, the laboratories, which were located on the main Clifton Road site adjacent to Emory University, covered seventeen separate disciplines, including bacteriology, toxicology, mycology, parasitology, virology, vector-borne diseases, and pathology. Here technicians could use electron microscopy to directly observe infected tissue, culture bacteria on appropriate media, and inoculate diseased material in cell cultures, eggs, and small laboratory animals. In addition, they could screen sputa and sera for antibodies to a range of antigens.
By the end of August, CDC technicians had scanned hundreds of tissue samples and used fluorescent antibodies against over a dozen different microbes. With the exception of one patient, who tested positive for mycoplasmal pneumonia, none of the sera showed significant antibody responses. Nor did tests of nasal and throat washings reveal the presence of chlamydia, Y. pestis, or more exotic bacteria and viruses, such as Lassa and Marburg. At one point, technicians got excited when three guinea pigs died of a mixed bacterial infection after inoculation with a lung suspension from a patient. However, it later transpired that the bacteria were typical of those found in patients after treatment with antibiotics or in postmortem overgrowth, and when the lung suspension was passed through a bacterial filter in an effort to exclude everything except a virus, it was no longer pathogenic for guinea pigs.† As the tests came up blank, the scientists tried different methods. One was to put blood samples in test tubes with antibodies against various microbes and look for a positive reaction. Mindful of the toxic chemicals theory, the scientists also subjected lung, liver, and kidney samples from deceased Legionnaires to radioactive assays for poisoning with twenty-three heavy metals, including mercury, arsenic, nickel, and cobalt.
After influenza and psittacosis, the next most likely suspect had been Q fever. Caused by Coxiella burnetii, an obligate intracellular parasite midway between a bacterium and a virus, Q fever used to be classed as a type of rickettsia. However, unlike other rickettsial diseases, such as typhus and Rocky Mountain spotted fever, which are transmitted by the bites of arthropods, humans typically get Q fever when they breathe in dust infected by contaminated animals (the principal animal reservoirs are cattle, sheep, and goats). Common symptoms are fever, severe headache, and a cough. In about half of patients, pneumonia ensues, and hepatitis is frequent enough that the combination of pneumonia and hepatitis is usually considered diagnostic. Unlike with typhus, a rash is rare and, though Q fever is an acute illness, patients usually recover even in the absence of antibiotics.
The researchers to whom it fell to test for Q fever were Charles Shepard, the head of the CDC’s Leprosy and Rickettsia Branch, and his assistant, Joe McDade. A bespectacled, blue-eyed scientist with a reputation for meticulous research, McDade had only just joined the CDC a year earlier. Thirty-six years old, he had previously been stationed with a Naval Medical Research Unit in North Africa, where he had worked on rickettsial diseases. In theory, this made McDade the perfect person for the job. However, at the time he had no experience of public health microbiology and looked to Shepard and other more experienced CDC hands for direction. After his overseas posting, McDade found the work in Atlanta laborious and a little dull. Departing from standard tests and other procedural deviations were not encouraged, he recalled. Instead, he was expected to follow prescribed algorithms and testing procedures, entering the results in a matrix that would hopefully line up with the epidemiological evidence and bring about a resolution of the mystery. Working with lung tissue from deceased Legionnaires, McDade’s fir
st job was to grind up the autopsy material and inoculate it into guinea pigs. Q fever has an incubation period of a week to ten days, so the next stage was to wait. If a guinea pig developed fever, McDade euthanized the animal, removed some of its tissue, and injected it into an embryonic egg. In this way, he hoped to obtain sufficient numbers of bacteria that could be stained and examined.
Part of the reason for McDade’s lack of enthusiasm was that at the time “everyone was looking for influenza or known causes of bacterial pneumonia,” and there was no evidence that the Legionnaires had been exposed to livestock, making Q fever highly unlikely. Sure enough, when he inoculated the material into guinea pigs they developed fever within two or three days, far earlier than if the organism had been C. burnetti. Modifying his procedure, McDade euthanized the guinea pigs prematurely and removed a section of their spleens. He then made impression smears on glass slides and stained them to see what organisms he could observe through the microscope. At the same time he used some of the tissue to prepare a suspension and streaked it on an agar plate to see if anything would grow on the media. Finally, he added antibiotics to the mixture to inhibit the growth of any contaminants that might be lurking in the tissue and inoculated the material directly into embryonated eggs so as to grow rickettsia should they be present.
He found no evidence of rickettsia; all the eggs remained perfectly healthy for more than ten days. Nor was he able to recover any bacteria from the agar plates. However, as he peered at the smears through the microscope, McDade would occasionally spot a rod-shaped, Gram-negative bacterium, “one here and another there.” Distrusting his observations, McDade shared the slides with more experienced colleagues, only to be told that guinea pigs were “notoriously dirty animals” and what he had seen was most likely an “experimental contamination.” “I was told there was an accumulating body of evidence that no bacteria were involved and what I had was an anomalous observation,” he recalled. Instead, McDade was told to look for a virus.
As McDade and Shepard’s efforts faltered, other scientists, abetted by politicians in Washington, revived the toxic metal and chemical contaminant theories. The leading advocate of the toxic metals theory was Dr. William F. Sunderman Jr., the head of laboratory medicine at the University of Connecticut School of Medicine. Early on in the outbreak, Sunderman and his father, William Sunderman Sr., professor of pathology at the Hahnemann Medical College, Philadelphia, had urged the public health authorities to collect urine and blood samples from suspected cases so they could be analyzed for toxic substances. The leading suspect in the Sundermans’ view was nickel carbonyl. A colorless, odorless metal, nickel carbonyl is widely used in industrial operations and is highly toxic. Symptoms can present themselves anywhere from one to ten days after exposure and typically include a severe headache, dizziness, and muscle pains. In the first hour after exposure, victims may also complain of shortness of breath and a dry cough. Without treatment, exposure can result in acute pneumonitis and bronchopneumonia with high fever.
In mid-September, the younger Sunderman had studied six lung tissue samples from patients with Legionnaires’ disease and found that five contained unusually high levels of nickel. However, while this suggested the patients may have inhaled a toxic substance, nickel concentrations in other tissues and organs, such as the liver and kidney, were normal. To exclude the possibility that the elevated readings were due to accidental contamination, he would also need to test urine and blood from Legionnaires, but unfortunately, in the confusion of the early days of the outbreak, public health officials had failed to collect and preserve specimens for future testing. Despite these caveats, at a congressional hearing in November chaired by John M. Murphy, a Democrat from Staten Island, New York, the Sundermans were highly critical of the CDC and the “flaws” in its investigation—flaws that they attributed to the “zeal” of public health authorities to see Congress enact legislation indemnifying vaccine manufacturers against prosecution arising from the swine flu immunization program. Sunderman Sr. was particularly critical, agreeing with a recent newspaper article in the Washington Post that had accused the CDC of “eagerness bordering on mania . . . to find swine flu in Pennsylvania.” Indeed, in his congressional testimony, he went further than his son had been prepared to do, stating definitively that the outbreak was due to nickel carbonyl poisoning. Congressman Murphy was similarly critical, stating it was “inconceivable” that no one could say with any certainty whether the outbreak had been “murder; a virus; accidental introduction of a toxic substance; or a . . . convergence of factors yet to be determined.” In particular, he described the lack of coordination between the CDC and other agencies as a national “embarrassment,” telling House committee members that “nothing was done to search for toxic evidence until it was almost too late.” Pointing out that “many experts [had] recognized the toxicological symptoms very early,” he argued that the possibility of “foul play” could not be excluded and that toxic substances may have been placed on telephones, food, or the Legionnaires’ ice cubes. “It is entirely possible that a terrorist group or single fanatic might possess the technology to distribute a deadly poison or bacteria among a large group,” he concluded.
It was not the first time Murphy had sought to stoke paranoia about the antiwar movement. In October, aides to his committee had leaked a story to the Washington Post saying that congressional investigators believed “a demented veteran or paranoid anti-military type” with some knowledge of chemistry may have been responsible for the Legionnaires’ deaths. Such stories played to the suspicion and anxiety that infected American society in the mid-1970s which, arguably, has only become more pronounced with time. A decade earlier, historian Richard Hofstadter had coined the term paranoid style to describe “the sense of heated exaggeration, suspiciousness, and conspiratorial fantasy” he detected in extreme right-wing movements, such as the 1964 campaign for the White House by Barry Goldwater, the militantly anticommunist Republican senator from Arizona. By the 1970s, this paranoid style was arguably no longer confined to the Right, but following the assassination of leading lights of the civil rights movement, was also beginning to infect the Left, hence the popularity of theories blaming the deaths of Jack and Bobby Kennedy and Martin Luther King on the CIA, the Mafia, and the Ku Klux Klan, or some combination of all three.
The early 1970s was also a time of growing anxiety about nuclear energy and the dangers of environmental and chemical pollutants, such as Agent Orange, the highly toxic herbicide sprayed on the Vietnamese countryside, which was just beginning to give rise to cancers and other unexplained health problems among Vietnam vets and their children. As Laurie Garrett has argued, from the perspective of the Left, “events in Philadelphia fit neatly with the then vogue view that an unregulated chemical industry was raining toxic compounds upon the American people.” By contrast, the Right was more inclined to view the outbreak as an act of sabotage, or as the Philadelphia Veterans of Foreign Wars put it, “a sneak attack against the finest kind of Americans.”
This sense of moral panic did not escape Bob Dylan, who incorporated some of the wilder speculation into a song, “Legionnaires’ Disease.” Written for his touring guitarist, Billy Cross, the song opened with the verse: “Some say it was radiation, some say there was acid on the microphone / Some say a combination that turned their hearts to stone.”
In retrospect this panic seems irrational, laughable even. After all, unlike cholera and plague, Legionnaires’ disease was not contagious. Nor was it a disfiguring disease like smallpox, or one freighted with metaphors of waste and decay, like cancer and tuberculosis. On the other hand, the fact that its identity was unknown made it ripe for the projection of society’s worst fears. Like Jack the Ripper, the mysterious killer had descended suddenly and unexpectedly on the Bellevue, then vacated the scene just as mysteriously. In the process, it had left few clues—or at least, none that the CDC’s disease detectives had been able to parse—turning what was usually considered a safe and secure
location into dangerous ground. It was a blow from which the Bellevue, already in financial trouble before the outbreak, would never recover. With newspapers using phrases like “mysterious and terrifying disease” and “the Philadelphia killer” to describe the outbreak, guests canceled their reservations one by one. The result was that on November 10, the management announced the Bellevue was no longer able “to withstand the economic impact of the worldwide, adverse publicity” and closed its doors to further business.