Pocket worthyStories to fuel your mind

The Source: An American Outbreak of Death and Panic

On the eve of America’s Bicentennial, a mysterious illness terrifies the country and sends disease detectives racing the clock to find answers.

Truly*Adventurous

Read when you’ve got time to spare.

a dark building against a pale sky

This American Legion conference did no credit to the splendor of the Bellevue-Stratford, Anna Taggart huffed to herself as she surveyed the scene in the opulent hotel she loved as though it were her own home.

The elaborately gilded and marbled threshold of the grand hotel had been crossed by every American President since Teddy Roosevelt, and its French Renaissance style architecture and opulent décor played host to the city’s glitterati on a regular basis at champagne brunches and high-society benefits. True, the 754 plush guest rooms rarely filled up these days, but the Bellevue remained a cultural monument in Philadelphia.

By 1976, the year of America’s bicentennial, Taggart, 60 years old and an institution herself, had been working at the hotel for 43 years. During that time she had mainly served as the chief elevator operator, proudly conveying such luminaries as Franklin D. Roosevelt and Katharine Hepburn to their accommodations. The legionnaires, as members of the American Legion were known, were not quite the same ilk as the guests she desired.

older woman looking at something off to the side

Anna Taggart

On July 20, the legionnaires, mostly men, began arriving en masse, and Taggart began an exercise in hiding her disapproval. As they rode her elevators, she pretended not to notice how they lugged coolers of cheap beer and spirits to store in their rooms, rather than partaking of the expensive drinks in the hotel’s sophisticated, softly lit bars.

She had a temper, and it wasn’t beyond her to tell one of the men to get his feet off the antique tables. But she was also a professional, and as chief elevator operator she had to set an example of poise and discretion. As instructed by hotel management, she turned a blind eye to the raucous behavior, and to the way some of them leered at her and the other female elevator operators. She even ignored the pranks; some legionnaires would shock unsuspecting victims riding in the elevator with a small electric buzzer, the kind you’d buy in a novelty shop. It was all she could do not to lose her cool. Ruffians, she concluded. A pox on her beautiful hotel.

Climbing out of the elevator on July 21, one of these ruffians, George Chiavetta, needed a smoke.

He wended his way through the throng in the hotel lobby, toward the front entrance. Lighting his cigarette on the threshold of the Bellevue-Stratford, he glanced around at other people milling around the sidewalk. Most were there for the Legion convention.

He observed one particular person sharing the sidewalk with him, who caught his eye because the man wore a suit of such a striking color. About forty years old and tall — maybe 6 feet — the stranger wore a bright blue, single-breasted leisure suit. He also waved around a rolled-up sheet of paper, which he kept pointing at people entering the building.

He was saying something as he did. What was it? Chiavetta moved closer to try to hear.

“It’s too late now,” the man in the royal blue suit proclaimed as he brandished his piece of paper. “You won’t be saved. The legionnaires are doomed.”

And then, perhaps noticing he was watched, the man abruptly turned and walked away.

Back inside, Maria Reeves — not her real name, she wasn’t so careless — was one of the few people in the Bellevue-Stratford’s lobby not affiliated with the American Legion convention.

Well, not officially.

She had reserved a room for these dates in July, though, because it made good business sense. A hotel teeming with bachelors and men traveling without their wives presented an opportunity too good to pass up. She hoped for a windfall.

Though only 24, she had been turning tricks for ten years. She catered to a high-class clientele, picking and choosing her marks and making good money. But she had fallen sick about six months earlier with a terrible pneumonia, and an extended recovery period wiped out her bank account. The Legion convention was an opportunity to replenish her savings.

With smooth olive skin, an expensive leather suitcase and a manner of dress befitting her guise of a young, wealthy widow, Maria blended effortlessly into the chic surroundings as she checked into the hotel late on the evening before the conference began.

For the next several days, she took advantage of the convention as efficiently as the military men snapped to attention in the meetings and exercises going on in other parts of the hotel. On the day the convention ended and her clientele headed home, she also checked out. She felt a little under the weather, but nonetheless dragged herself to the train station. It had been a long, albeit fruitful week. She would check into a hotel in the Catskills, she had decided, to get some rest.

On that same day, sixty-one-year-old retired Army Captain Ray Brennan also arrived home to Sayre, Pennsylvania, tired out after his week schmoozing and carousing in Philadelphia.

Not unexpected, his family supposed. Ask anybody; his family would tell you that all he lived for were these raucous meetings.

So it wasn’t so crazy that he felt a little low on energy since returning three days ago. But he couldn’t shake this bug. Maybe it was his heart; he’d had some trouble with it over the years. His sister wanted him to go to the doctor. He refused.

Breathing was difficult. His chest hurt, and he had a fever. As his lungs filled with fluid, he would have felt like he was drowning, gasping for air as the oxygen refused to enter his alveoli. His heart would ultimately stop in what everyone assumed, given his health problems, was a heart attack.

It was not.


Dr. Leonard Bachman, the energetic and ambitious 51-year-old Pennsylvania Secretary of Health, stopped into the office on Monday morning for what he planned to be a quick check-in with his staff before heading to his cottage in Maine. His mind already half-occupied by thoughts of his beloved sailboat, he had no idea that, over the weekend, the health department and the American Legion had become embroiled in a burgeoning crisis.

blurry shot of a man wearing glasses and a tie

The disease causing the alarm had first attracted notice the previous Thursday, two days after the legionnaire Brennan died from what everyone thought was a heart attack. That day, Dr. Ernest Campbell had admitted his patient Sam Morris to a hospital in Bloomsburg for high fevers and weight loss, symptoms Morris had developed since attending the State American Legion convention in Philadelphia. After a day of testing and observation, the doctor told Morris that he thought the symptoms might be caused by typhoid, a disease that could bring on fevers and malaise, and that Morris probably contracted it from eating or drinking something contaminated with salmonella bacteria. At this news, Morris mentioned a curious coincidence.

“Well doc,” he said, “There’s two other fellows on this floor that were at the same place, the same convention that I was at. Maybe they should get checked out too.”

Dr. Campbell agreed. The timeline made sense, and the common location made sense; these men undoubtedly ate and drank some of the same things while in Philadelphia together. He started Morris and one of the other two men on drugs for typhoid fever. The other, an elderly man who already had serious health problems, sadly succumbed to the disease before the drugs could be given.

By the time he was back in his office on Friday afternoon, Dr. Campbell was reasonably sure that all three men had typhoid, and that they must have contracted it at the conference. Around 4 PM, he called the state public health department to report a typhoid outbreak, but the nurse who answered the phone told him that since it was so close to the weekend, no definitive response could come until Monday.

On that same Friday, a hundred miles away, an infection control nurse in Chambersburg called the Pennsylvania State Department of Public Health to report an eerily similar situation: three hospitalized men, one of whom had died. All had been at the Philadelphia convention; all had the same symptoms.

Her complaint, though, got stuck in a bureaucratic maze from which it failed to emerge in a timely manner: a restriction on long-distance calls impeded the message from traveling to the right people before they went home for the weekend.

On Saturday, the Philadelphia City Health Department also fielded two calls, one from the pathologist at Chambersburg and another from a doctor at another hospital, both concerned about very ill people with a bad respiratory disease that no one could identify. They, too, heard that no investigation could start until Monday.

Meanwhile, in Williamsport, six more men wound up hospitalized with fever, fatigue, and chest pain. Like the others, they were in the grips of a terrible pneumonia that didn’t respond to conventional antibiotics.

The head of infection control in Williamsport knew that six men with the same odd disease, contracted after attending the same event, could not possibly be a coincidence. A pattern, an outbreak, maybe an epidemic. The infection control doctor made the decision to eschew politeness and called the district health director at his home on Sunday night. From there, the message reached the state epidemiologist’s office at the Department of Health.

And as that phone call went through, another phone message was being received.

Edward Hoak, the Adjutant of the Pennsylvania American Legion, had been away at a Legion event that weekend, and had just returned to headquarters in Harrisburg on Sunday night. While he was away, he had heard that a few men who had been at the Philadelphia convention were sick. Then, on Sunday, the news got worse; he returned to headquarters to find a note from his secretary that yet another Legion member had fallen ill. A few minutes later, she called him again to tell him someone else had died. He then phoned his assistant, only to learn that he was away at a funeral for yet another legionnaire.

Adjutant Hoak decided to inform state officials about these illnesses and deaths first thing Monday morning. But some time before he did that, he placed a call to a reporter he knew.

It was Hoak’s duty to get the word out and make sure his legionnaires knew to get to a hospital if they were feeling ill. He just wanted to know, he told Paul Carpenter of the Associated Press, how he could get a small notice published. He would even be willing to pay for an ad.

Carpenter’s ears pricked up, fingers flying over his typewriter keys. American veterans with a mysterious illness, who had all just been to Philadelphia during the Bicentennial festivities. That was a story.

It was the next day that Pennsylvania Secretary of Health Leonard Bachman popped into his office, impatient to get on the road to Maine, and encountered the seeds of a public health crisis. And worse, the media had somehow found out about the situation before he had; calls from the press demanded information about what the AP reporter Carpenter had dubbed “Legionnaires’ Disease.”

Brass sailboat fittings languishing forgotten in his jacket pocket, Bachman abruptly cancelled his vacation plans. The first thing to do, he figured, was put together a press conference, and get control of the story the media was putting out.

So later that day, Bachman found himself reassuring the press that, although, yes, a rash of deaths had followed a meeting of the American Legion, “we can see no cause for any panic of any sort at this time.”

He had ordered an immediate statewide investigation, he told reporters.

But of what? What was causing the deaths? The press wanted to know. At this, Bachman conceded.

“We don’t know what we’re dealing with.”


Dr. David Fraser was only 32 and a rising star at the Center for Disease Control (later renamed the Centers for Disease Control), the US government agency in charge of public health. He was a senior epidemiologist in the Special Pathogens branch, which focused on rare and unknown diseases. Over the first half of 1976, he had a prime vantage point from which to watch — interested, but uninvolved — as the CDC got increasingly invested in a very public effort to prevent the next flu outbreak.

blurry shot of a man not smiling and wearing a tie

In January, two soldiers had died at Fort Dix in New Jersey from what turned out to be swine flu, a strain likely related to the one that caused the 1918 global flu pandemic that killed at least 50 million people. Public health authorities and politicians had gone on high alert for the disease, and the CDC assembled a special flu team. In a disused auditorium, a squadron of epidemiologists constructed a makeshift epicenter for what they expected to be a large and intensive flu control program. In a politically dicey move, the head of the CDC also persuaded the US Government to vaccinate every person in the United States against flu, a program that quickly turned into a bureaucratic cesspool.

Fraser, a direct and observant man with unruly brown hair, a dry wit, and an independent streak, was glad to be insulated from the hubbub over the flu, but watched the developments in the program with interest.

His was a relatively sexy role, as public health jobs go, but only when an epidemic had to be contained. The most glamorous thing he had done all week was write a memo to a CDC security official with whom he was locked in a battle over whether he could park his beloved racing bike inside the government building instead of outside at the bike rack.

Monday, August 3, however, turned out to be very interesting.

That day, a call arrived from Philadelphia directed to the new flu control unit. It landed in the hands of a young physician fresh out of training for the Epidemiologic Intelligence Service, or EIS. EIS officers act as “disease detectives,” as the media would later dub them: a group of highly trained public health experts regularly dispatched to investigate and control epidemics all over the world. The EIS officer picked up the phone and took down the report from the Philadelphia Veteran’s Administration Medical Center.

“Four dead? Twenty-six more in all?” he confirmed, scribbling notes.

As he hung up the phone, he banged on the partition separating his own cubicle from that of another newly-minted EIS officer.

“This is it!” he yelled. “Swine flu!”

The fledgling unit mobilized quickly. They set about gathering supplies for tests they planned to run and sent word up the chain to the most senior people in the CDC. The epidemic, they thought, had arrived.

Fraser, in Special Pathogens, got wind of the situation and took a characteristically analytical approach to it. Could be flu, he thought. But it could also be any number of other diseases. Setting his concerns over his racing bike aside, he considered that the unfolding situation could be more in his wheelhouse than the flu team’s. At the very least, it might give him and his staff something interesting to do.

He offered up his assigned EIS officer as an addition to the team being dispatched to Pennsylvania. Less than twelve hours later, after arranging a formal invitation from the State to join the investigation, one young investigator was en route to Philadelphia, another headed for Pittsburgh, and Fraser’s man made his way to Harrisburg.

Apprehension among the EIS officers ran high. Their parents and superiors came of age during the first half of the 20th century and plied them with stories not only of war, but also of infection. Even as early as the first day of the investigation, public fear of flu rose to near panic. The question was on the lips of every reporter, politician, and medical professional: was this the flu, and was it a crisis?

And at the CDC, the epidemiologists had another concern: if not the flu, this could be something even scarier. It could be one of any number of contagious diseases that killed swiftly; or, worse, a new disease that no one knew anything about.

As the EIS officer team grew larger than any other in history, CDC higher-ups also quickly realized that they needed an experienced epidemiologist to head the field team. So early Tuesday morning, Fraser boarded a plane to Harrisburg.

On a “very intense plane ride,” as he later recalled, he mulled over what he expected would be a “very complicated problem.”


Lawrence Altman wondered if there was really anything to get worked up about. As a health reporter for the New York Times, the 39-year-old Altman prided himself on his critical eye for news. He also held a unique status as the first qualified physician ever to be a regular contributor to a major paper. And to top it off, back in the 1960s, he had worked as an EIS officer.

blurry shot of a man smiliing and wearing a tie

Dr. Lawrence K. Altman

He knew Pennsylvania Health Secretary Leonard Bachman by reputation as an opinionated government official intent on increasing the profile of the health department through legislative advocacy and, more problematically, through the media. He knew that Bachman, earlier that year, had jumped the gun on a different outbreak; he had prematurely announced several cases of a newly-discovered disease called “pantosomatitis.” The report caused a momentary frenzy of activity over a supposed eight-state outbreak of the ailment, but the claim came to nothing; an investigation that lasted less than a week showed that the disease did not exist.

The pantosomatitis incident embarrassed health officials and left Altman reluctant to trust the hotheaded Bachman. Could the Pennsylvania Health Secretary’s recent public announcement, now the biggest story in the region, be another Bachman red herring destined to squander resources and garner a sheepish backtracking press conference?

Altman figured he would wait and see. However, within a few hours of the first Monday announcement, the death toll rose to 14.

As the victim count mounted, Altman became less and less skeptical that these were just a few overblown cases of pneumonia. It could be the real deal, the beginnings of a pandemic. Late Monday night, he filed his first story on the outbreak in Pennsylvania for the Tuesday morning edition of the New York Times.

Then Altman got in his car and flipped on the radio to keep himself awake. On every station, newscasters reported the dangerous new illness and speculated about the next flu epidemic. The reports accompanied him the whole way to Harrisburg that night; nothing else was on the radio.


By Tuesday, more than 70 cases of the mystery ailment had been reported. For each case that came in, an EIS officer in Philadelphia pushed a yellow pushpin into a large map. The cases accumulated so quickly that by the middle of the first day of the investigation he had run out of pins.

map of pinpoints around pennsylvania

The mood of the fast-paced investigation, in its dual epicenters of Philadelphia and Harrisburg, bordered on frantic, but remained methodical and detail-oriented. No clue could be overlooked, no lead uninvestigated. At that moment, for example, a veritable search party was on the trail of a bus bound from Philadelphia to Georgia carrying a band that had played at the Legion conference. One of the band members was sick and could possibly seed another outbreak. They had to find that bus.

Fraser, meanwhile, had arrived at the Health Department in Harrisburg where Secretary Bachman was in the middle of one of his twice-daily press conferences.

“As long as we don’t have the cause, no cause is ruled out,” Bachman said, further explaining that health officials needed at least two more days to conduct laboratory tests to determine whether the next flu pandemic may be afoot.

The press corps pushed Bachman to explain how public health officials would vaccinate the American people in time to contain a potential epidemic. Bachman, for his part, considered imposing quarantines, even martial law, to get people vaccinated and the disease contained.

When the press conference finished, Bachman and Fraser met for the first time and got down to business, discussing how the investigation should be organized. They needed to decide which parts the Health Secretary should spearhead and which parts the Special Pathogens epidemiologist should run.

Fraser would happily let Secretary Bachman handle the press; Fraser was there to do science and had little interest in the evolving media circus. He would take responsibility for the activities of the EIS officers, they decided, and Bachman would be the public face, continuing his press conferences while managing the media.

Both felt keenly the urgency of the rapidly expanding investigation. As a Washington Post reporter wrote that day, “The disease struck with unusual suddenness and killed so swiftly that a number of persons had died before the state health department even knew it had an outbreak on its hands.”

In very short order after the Associated Press broke news of the first deaths on Monday morning, the number of calls to the Philadelphia health department overwhelmed its secretarial staff. The department opened a dedicated line to receive reports and concerns related to the disease, and after only three hours had fielded more than 150 phone calls. As media coverage of the mysterious outbreak intensified, the rate only increased. The team manning the call-in line — secretaries as well as police officers and a hodgepodge of public health personnel called in to help answer phones — carefully took down a caller’s information and asked screening questions, passing along those that seemed to warrant attention to the medical teams investigating new cases.

Pennsylvania public health nurses were reassigned from their usual tasks, sent to work with the EIS officers sweeping the state looking for cases. Victims’ blood, sputum, urine, and autopsy specimens were carefully preserved and split between the public health laboratory in Pennsylvania and the lab at the CDC in Atlanta. Teams also collected samples from family members and people who lived in the same house as the victims, as well as from totally unrelated people, in case there were patterns to be found in the environment where the sick veterans were living.

Each day for a week, health workers visited every single hospital in Pennsylvania, calling back to report new cases. As reports piled up, the disease detectives collated the information they collected onto a large chart, known technically as a “line list.” Each sick person’s name claimed a single line, and then relevant data about them filled out other columns, nearly seventy in all. Scanning down the chart quickly, one could see whether everyone who got sick ate shellfish for dinner on the same date, drank their whiskey with ice, or stayed in the same hotel.

As Fraser looked at the line list on the first day, his analytical eye noted one odd pattern for an outbreak of pneumonia. The CDC had just begun its investigation, but there were almost no new cases reported that day. There was a barrage of reports of people already sick, or getting worse, but nearly no one seemed to have fallen ill that morning. Flu was extremely contagious; if it were the flu, he would expect new cases every day. They would have to ask careful questions of family members, roommates and other contacts of the victims to be able to draw any clear conclusions about the flu theory, but at this point, they needed to remain open to other possibilities.

Over the next week, Dr. Fraser organized the EIS officer team into an efficient operation aimed at defining this strange disease, finding out where it had come from, and stopping it from killing more people. Each of the more than a dozen officers working under Fraser had his own project to elucidate the source of the outbreak. Fraser put one in charge of conducting a massive survey of everyone who attended the Legion conference, since the disease seemed to be associated somehow with the event. The investigators hand-delivered the so-called Ten Thousand Survey (which actually ended up surveying about 4,500 people) to every Legion post — about a thousand — in Pennsylvania. The CDC tabulated and analyzed answers to the survey using the first computer they ever employed in a field investigation. The clunky 15”-by-15” white metal box arrived at the field headquarters in Harrisburg with a statistician to operate it.

Another EIS officer took blood from employees of hotels where the legionnaires stayed. Two more called recovering victims to ask additional questions. The CDC team, along with public health nurses and doctors from Pennsylvania, worked 14- and 16-hour days.

Every night, Fraser got updates from his team, and at the end of the day — often at midnight, or one, or two AM — he phoned CDC headquarters. He meticulously reviewed his team’s findings with his boss before dropping into bed for a couple of hours of sleep.

(During this overworked period, three overly enthusiastic EIS officers made the poor decision to take home confidential Health Department data to analyze on their own, without the team’s knowledge or permission. Commenting on this debacle of protocol later, Dr. Fraser called it a “dicey situation…[but] we decided not to penalize anyone. Their motives were good, even though the act was arguably, uh, mutinous.”)

Fraser had chosen an EIS officer called Stephen Thacker for an assignment he knew some would have found particularly irksome. Thacker, a sandy-haired young physician, was both a precise scientist and an affable colleague, even a bit of a ham. Fraser figured Thacker could deal with these media types crowding the Health Department, waiting for the next big development in the story. So on Wednesday of the first week of the investigation, he sent a troupe of reporters off with Thacker to interview patients.

The group arrived in Chambersburg, where some of the first victims had been identified, and went to the room of a recovering legionnaire.

“Hello,” Thacker opened, sitting on the patient’s bed. “I’m Dr. Stephen Thacker with the Health Department, and I am part of the medical team investigating this weird disease.”

Questions about symptoms — “When did you start feeling sick? Have you had a fever?” — accompanied questions about exposures: “Have you had contact with pigs?” and “Where did you stay in Philadelphia?” Questions about what the patient had eaten, whom he had had contact with, and where he had traveled, would add information to the line list.

As Thacker sat with the legionnaire, the journalists interjected questions about the data collection process: Why such urgency, they wanted to know. EIS officers were fanned out over the state. They got only a few hours of sleep a night. They rushed to the bedside of any sick person they discovered, to ask a long list of painstaking questions. What were they looking for?

“Well, we don’t know what’s going on,” said the mild-mannered Thacker. “It’s clearly a serious disease… The urgency is because we want to know what’s going on.”

What had he found out? What could he tell the public?

“I’ve talked with pathologists, patients, looking for patterns, growths,” he replied. “So far, I’ve found nothing.”

And then, “Isn’t it risky for you?”

Thacker remained candid: “Possibly.”

Within the first week of the investigation, tests for flu had come back definitively negative. In fact, testing during that first week ruled out all known bacteria and viruses.

This result perplexed the experts; flu and other respiratory viruses pass person-to-person in a way this disease did not. But the autopsy results and the symptoms were most like a viral pneumonia. Maybe this was some as-yet unknown virus, for which they could not test.

The pneumonia killing attendees of the Legion conference also showed other oddities that baffled the researchers. The fevers were very high, up to 107 degrees Fahrenheit. The patients struggled to breathe, so much so that one out of every five patients became “obtunded” — hard to awaken, their bodies on the edge of shutdown. When physicians examined them, their lung sounds evoked a viral infection, but the running nose and congestion that would typify most respiratory viruses weren’t present. The physical exam, in fact, registered as strikingly normal, with no telltale rashes, no enlarged tonsils, nothing to give a hint as to what bug might be causing this strange and severe disease.

Bachman’s head virologist at the Pennsylvania State Health Department, Dr. Jay Satz, pressed the point home in a press briefing on August 6, five days into the investigation. “If this is a virus,” he remarked, “it is a very unusual virus.”

He added, “I am treating this organism in my laboratory as an extremely dangerous one. I don’t know how dangerous it is. I only know it has killed over 20 people. I have to consider my own life and the lives of my technical people, and we are using every precaution that we have available to us in studying this agent.”

Dr. Jonas Salk, the inventor of the vaccine against poliovirus and one of the era’s most brilliant microbiologists, even called Dr. Satz to help him think through the disease. “He was as perplexed as much as I am with this situation,” Satz later told reporters.

And then on August 10, a new theory emerged. Health officials stressed that they considered several toxins to be possible causes of the outbreak, but Lawrence Altman found one poison hypothesis particularly convincing. He made the decision to highlight it prominently in the story he filed for the August 11 edition of the New York Times.

Dr. F. William Sunderman was a Philadelphia scientist and former medical director for the Manhattan Project, which developed the atomic bomb. In reading descriptions of the sick legionnaires, he thought of the symptoms of poisoning by a substance he knew well: nickel carbonyl, a highly toxic molecule used to produce magnetic tapes and mirror coatings, that can cause a syndrome much like the one seen in the legionnaires.

Sunderman took a special interest, because on July 28, his son, also an expert in nickel carbonyl poisoning, had received a letter that mentioned the nickel compound and threatened, in vague but menacing terms, the murder of prominent political figures.

And indeed, the elder Sunderman told Altman, the Philadelphia outbreak seemed consistent with a mass poisoning due to nickel carbonyl. Unlike most other toxins, this particular one takes a few days for symptoms to set in, so the timeline was consistent with the outbreak. Affected individuals first get a headache and chest pain a few days after the initial exposure, and develop severe lung inflammation. The fevers, he allowed, were not typical of nickel carbonyl poisoning, but they had been a feature of one outbreak he had been involved in investigating in Japan.

Odorless and colorless, the nickel carbonyl — Sunderman said — may somehow have gotten into an air conditioning system.


About a week after her lucrative stint cozying up to the legionnaires, Maria Reeves holed up in a New York City high rise. She felt increasingly unwell.

After her stay at the Bellevue, she had gone to the Catskills as planned, and then come to the city to visit a friend. Now, she was alone in the friend’s apartment, with a high fever, shaking chills, and difficulty breathing. Though she knew she needed a doctor, she feared what might happen if it came out why she was having symptoms so similar to the sick legionnaires’. A few years earlier, a doctor had betrayed her confidentiality and reported her to the police, causing her legal headaches that she did not want to repeat.

Eventually she managed to leave the apartment, planning to get on a train home to Philadelphia. But when the cab picked her up, she couldn’t seem to get her intention across to the driver. He threw her out of his taxi, thinking her drunk or on drugs or both.

She wandered through midtown Manhattan, eventually drifting into a Macy’s. Someone alerted the police to a woman wrapped in a winter coat in the heat of August, roaming the store, not in her right mind. The policemen waited for her at an exit and, more observant or perhaps more compassionate than the cabbie had been, coaxed her into their car to bring her to a hospital.

What came later was a blur.


Legionnaire George Chiavetta read the papers. He knew that all the tests for viruses and bacteria were negative. He had read that some investigators thought that the victims — his compatriots, with whom he had clinked glasses in Philadelphia — might have been poisoned.

Chiavetta remembered coming out of the elevator and stepping outside for a cigarette, remembered that peculiar man in the royal blue suit who had threatened the legionnaires from the street. The man’s hissed words came back to him: You won’t be saved. The legionnaires are doomed.

He had seen the man a couple more times during the convention, always wearing that suit and always carrying that rolled-up paper. The last time he saw him the man had a glassy look in his eyes and a red face. He looked like he was up to no good.

Come to think of it, Chiavetta realized, he also seemed to have had something in his pocket that night. It looked like a tobacco pouch, but anything could have been in there.

Suspecting that he held the key to a plot to murder the legionnaires, Chiavetta picked up the phone to call the authorities.


Dr. Joe McDade, a 36-year-old microbiologist, joined the CDC a few months prior to the outbreak. He had met his boss, the eminent Dr. Charles C. Shepard, at a conference where “Shep” immediately offered him a job. Shep, a brilliant microbiologist who had revolutionized leprosy treatment, had been impressed with the young McDade, whom he assessed as a sharp researcher, and who specialized in investigating a type of unusual bacteria called rickettsiae. Shep knew that the CDC had no rickettsial specialists, and needed one; so McDade started as the CDC’s sole rickettsial researcher at the end of 1975.

old portrait of a man with a mustache wearing thick glasses

Dr. Joe McDade

After years living in Egypt and Ethiopia as an Army researcher, McDade had fallen out of the habit of reading daily newspapers or watching television, so was relatively oblivious to the media storm around Legionnaires’ Disease. He overheard some water cooler chat about a major outbreak, but since the symptoms of the disease didn’t suggest infection by rickettsiae, he didn’t concern himself with the details. Being new to the agency, a “lowly backbencher” in a relatively obscure branch of the lab division, he didn’t expect to be involved.

But as the investigation went on, and the net widened to consider less probable causes of the pneumonia, McDade, to his surprise, was caught in it. Unlike the news frenzy that he had blocked out, the frenzy inside the walls of the CDC became impossible to ignore. “Just about everybody in the Bureau of Labs seemed to be involved,” McDade recounted when asked later about this period, blue eyes sharp and serious behind his sensible round glasses. Speaking in a quick, clipped accent reflective of his military background and his fastidious scientist’s demeanor, he described the procedure he followed when he, too, was asked to weigh in on his area of expertise. He tested for a rare but known disease called Q fever, which is a pneumonia caused by rickettsiae.

“[Q fever] didn’t seem quite plausible to me at the time,” he explained in a lecture at the CDC later. “First of all, Q fever was associated with contact with domestic animals or their environs… but more importantly, Q fever was rarely fatal.”

Nevertheless, he decided to go through the prescribed diagnostic algorithm, which involved taking a physical specimen from a human patient and inoculating it into a guinea pig. If the guinea pigs got sick, tissues taken from them were then cultured in egg yolks where the bacteria would multiply wildly. Then, under a microscope, a researcher would confirm that the yolks were growing rickettsiae, and make the diagnosis.

McDade inoculated guinea pigs with lung tissue taken from one of the legionnaires. It caused a fever in the guinea pigs, but much more quickly than McDade expected to see; Q fever would cause disease about 7–12 days after inoculation, while these guinea pigs got sick after two or three days.

The short time between infection and disease seemed to rule out Q fever, but McDade nonetheless systematically smeared tissue samples from the unfortunate rodents onto glass slides. He also made a solution that he dropped onto petri dishes containing different combinations of nutrients and salts that different bacteria liked to grow in. He prepared egg yolk cultures too, adding an antibiotic that wouldn’t kill the rickettsiae, but would eliminate any contaminant bacteria that might be present, so that his cultures would be unsullied when he reviewed them.

petri dishes

“The results did not seem particularly unusual,” he said later. “Nothing grew in the embryonated eggs, which suggested that no rickettsiae were present… which was not unexpected. I found nothing growing on any of the bacteriologic media. And when I looked at the stains and smears of the guinea pig tissues, I would see an occasional rare rod-shaped bacterium… probably normal bacterial flora from human lung tissues, because basically that was all that the people in the bacteriology labs were finding in their cultures of tissues that they were studying from the epidemic.”

Being a lab scientist and not a medical doctor, he didn’t know much about pneumonia-causing bacteria, but he did know that rare, single organisms on a slide were almost always contaminants. A single pink rod-shaped bacterium generally had no significance.

Having ruled out Q fever, McDade wrapped up his assignment and prepared to move on. He had a fuzzy notion from the rumor mill that a toxin, maybe nickel carbonyl, could be the cause, which seemed more probable.


Tips flooded in as test after test came back negative and the country became ever more anxious for an answer. One coroner suggested that the autopsies he had done showed signs of ethylene glycol poisoning, and posited that the victims may have ingested antifreeze in moonshine during the conference. Heavy metal poisoning likewise was considered, but ultimately it too was discarded after tests for seventeen different metals came out negative.

And then, as the churn of tests and investigations continued without yielding answers, the Veterans of Foreign Wars put out a press release demanding a criminal investigation into the outbreak, which they charged was a conspiracy to kill legionnaires.

In a heated response to this press release, the editorial board of the Philadelphia Inquirer remarked on August 8 that “those who see conspiracies everywhere can always find them,” prompting outraged letters to the editor that week that further reflected a paranoia among some veterans.

One interested party wrote from Seattle. “Let’s say a mentally deranged person with a fixation that the American Legion is somehow the cause of our wars felt it his duty to do something about it. As a roving hot dog hawker, he might have a supply of ordinary and wholesome hot dogs for the average citizen but quite another kind for persons wearing badges or ribbons identifying them as being associated with the American Legion Convention.”

A writer from New Jersey voiced a different concern: “The apparent difficulty in determining a chemical toxin must inevitably lead to the suspicion of a chemical warfare agent. The timing of the incident and the victims both suggest foul play… If Army chemical warfare personnel do not look for it, nobody may be able to document such a secret agent. What an inviting prospect for the government!”

After all, a decade prior, the New York City subway system had been seeded with a chemical agent in a vast joint CIA-Army test on the public. And just the previous year, in 1975, the country had learned that the CIA secretly dosed its own people with LSD, resulting in the death of at least one of its own employees. Could the outbreak be the unintended consequences of another government experiment?

It also didn’t help that the Army declined a request to send a specialist in chemical and biological warfare to help with the investigation, asserting that they had nothing to do with the outbreak.

And while frightened veterans entertained thoughts of foul play, others had their own ideas. A woman from Bala-Cynwyd, Pennsylvania, floated an off-the-wall theory that the whole thing could be put down to soda bottles with screw-on caps rather than the clamp-down, cork-lined caps used until the 1960s.

“If someone with dirty fingers handled the bottles, then surely,” she wrote in a letter, “organisms that could cause mischief would be cultured.”


As the Philadelphia Inquirer engaged in a war of words with the VFW and the Army stonewalled requests from Pennsylvania for help and information, Altman, the New York Times reporter, remained in frequent contact with the Sunderman lab in Connecticut, where they ran the nickel tests. There had been great excitement after the first batch of victims’ samples all came back positive for the poison.

But excitement quickly turned to dismay. They realized that control samples, taken from an autopsy of a patient who did not have Legionnaires’ Disease, were also positive. In other words, all of the samples had been contaminated somehow with nickel, and all the tests therefore were invalid. They planned to run them again, but this time on autopsy samples collected using plastic knives.

Altman checked in at a coroner’s office to get their perspective on the new instructions to use plastic knives. But to the reporter’s surprise, the coroner used a metal knife. No, the man told Altman, he had received no such instruction from the investigators. He would use his normal equipment and ship samples off in plastic canisters to be analyzed, like he always did.

Altman confirmed with the health department that the instruction had indeed been sent out, but clearly it hadn’t reached the furthest margins of the investigation. Something, frustratingly, had been missed. If nickel carbonyl was the cause of the outbreak, the mistake might have just put the whole investigation at risk.


Though the verdict was still out on whether a toxin was the culprit, Fraser’s team grew more convinced that whatever caused the disease could not be passed from person to person. The “epi curve,” a graphic representation of the outbreak, showed a sharp increase in new cases and then a rapid decline. This pattern is classic for a “common source” epidemic — something acquired by all the victims within a short window, indicating that they had all been exposed to some specific disease-causing agent. (In this way, Ernest Campbell’s initial typhoid hunch was astute, as typhoid does often cause common-source outbreaks.)

As the investigators proposed and discarded cause after cause, the situation rang nagging bells in Fraser’s head, and in the heads of the other senior CDC epidemiologists.

In 1968, a strange fever had struck office workers at the health department in Pontiac, Michigan. EIS officers sent to investigate the case got sick themselves, one after another after another, and finally realized that they could only enter the building safely while wearing face masks; this observation marked the first clue as to how the disease spread. The air inside the building seemed to be causing a common-source outbreak.

A two-year investigation identified two other important aspects of the infection: first, that animals didn’t seem susceptible to it, with the exception of guinea pigs. And secondly, that the problem seemed to be traceable to the building’s air conditioning system. But even with these clues to go on, the CDC had never found the cause of the disease: not from autopsying the guinea pigs, not from running tests on the air conditioning system. Without a source, they would be unable to prevent outbreaks of it in the future.

As August ticked by with no new answers found, the press asked a scientist at CDC whether the 1976 situation had prompted the center to review files on the disease known as Pontiac fever. According to a report in the New York Times, the unnamed official had smiled and said that it hadn’t been necessary. Everyone at CDC knew the Pontiac fever case “by heart,” he had explained. It was, after all, “one of their very few failures.”


Confusing the public and the press, the number of cases seemed to increase and decrease willy-nilly over the first two weeks of the investigation. The jockeying happened as Dr. Fraser’s team tried to tease out who on the line list suffered from the terrible Legionnaires’ Disease and who just had a run-of-the-mill pneumonia that they could have caught anywhere, in any year, in any city.

With the CDC in Atlanta receiving upwards of 3,000 calls a day from all over the country with questions, tips, and general panicked inquiries from the public, pressure grew for the scientists to understand this disease and tell people how to protect themselves.

Without a bug to culture or a definitive lab test to run, one of the investigators’ most urgent priorities was to define two sets of criteria that they could use to identify victims of the mysterious disease. First, every victim must have the right symptoms, and second, they must have been in the right place at the right time.

Within a few days, Fraser’s team had finalized the symptom criteria. To qualify as a Legionnaires’ Disease patient, a person needed a fever above 102 degrees Fahrenheit and a cough or a chest X-ray that looked like pneumonia. But the epidemiologic criteria were trickier. Using the line list — the complicated chart of patients and their exposures, that by then had grown to more than 12 feet long — they had to narrow down whether a particular meal, a particular building, or something else altogether tied the victims together.

And then, on August 8, a member of the Fraternal Order of the Odd Fellows called the CDC operation. He wanted to let them know that he had attended a different conference at the Bellevue-Stratford Hotel two years before, in 1974. Sixteen people had gotten sick afterward, he said. Three had died.

The CDC raised its institutional eyebrows at that report and opened an investigation of the 1974 Odd Fellows meeting. The call supported an idea that had already been on the minds of the investigators: Was it possible that the hotel was the common link? Could something coming from the most elegant establishment in Philadelphia, the playground of movie stars and presidents, the so-called “Grande Dame of Broad Street,” be killing people?

The hypothesis raised further questions. Had whatever struck people at the Legionnaires convention actually lurked in the environment all along? If so, why hadn’t it killed anyone between 1974 and 1976? And would it kill again?

As the line list grew, it became clear that the only thing that every victim had in common was indeed the Bellevue-Stratford Hotel. All of the victims of the mysterious pneumonia had spent time in or near this hotel. There were even a few cases — a bus driver who dropped off a group of conventiongoers, a bank teller who worked across the street — among people who had only stood in the hotel lobby, or outside on Broad Street. These passers-by shared no other risk factors with the ill legionnaires, and yet they became sick too.

So, after a week or so tracking the epidemic, with the specter of another, bigger, more deadly Pontiac looming in his consciousness, Dr. Fraser decided that there was only one logical thing to do. Logical, albeit foolhardy, and possibly highly dangerous.

They would go meet the beast where it dwelt. He and his team would move into the hotel.

back leg shot of people wearing suits and holding briefcases

The churn of conventions and events at the Bellevue had continued after the legionnaires had left, starting with the International Eucharistic Congress, which brought almost a million people to Philadelphia. But even as the Pope prepared to fly to the US to attend, news broke that people were getting sick and dying. The Vatican cancelled.

Within two weeks, both the Houston Astros and the San Francisco Giants also cancelled plans to stay at the Bellevue-Stratford. Patronage of the Grande Dame of Broad Street, with her ornate marble staircases and delicate Tiffany stained glass, slowed to a trickle.

Dr. Fraser later reflected, tongue partially in cheek, that part of the reason he moved his CDC team into the hotel was that, “There were rooms available, the price was good, and… maybe there’s a residual chance… that the agent is still being disseminated, and maybe we could catch a break.” Or maybe they would just catch pneumonia.

Fraser knew, like every epidemiologist in the world, the basic tenet of a common source epidemic: in order for the disease to stop spreading, the source must be found and controlled.

In the middle of the nineteenth century, Dr. John Snow first demonstrated this rule of common source epidemics in a quirky historical coincidence that struck a chord for the team of EIS officers arriving at the hotel’s main entrance on Broad Street. Dr. Snow became the first person to map an epidemic, overlaying the places that the disease was found onto a map of a city to try to find clusters or patterns that could point to a source. In Dr. Snow’s case, he plotted the spread of cholera through Cambridge, England. Using this innovative technique — considered the first epidemiologic study in history — he concluded that many more cases of cholera occurred among people who drank water that came from a specific water pump — located on another Broad Street — than among those who did not.

Dr. Snow convinced town authorities to remove the pump handle from the Broad Street water pump. The cholera epidemic ground to a stop.

As Fraser and his team milled around the lavish but empty Bellevue lobby waiting to check into their rooms, it seemed very likely that their outbreak could be traced to the hotel where they were standing, a Broad Street water pump for a new century. They also had no idea whether another crop of people connected to the hotel — perhaps them — could fall sick at any moment. And they remained acutely aware that they still didn’t know how to remove the pump handle.

Despite the risk, the CDC team asked recovered victims to return to the hotel and walk the investigators, step by step, through the places they had gone during the convention. Samples from every conceivable surface went to labs in Harrisburg and Atlanta for analysis.

More information accumulated every day, but nothing seemed to bring the investigators closer to narrowing down a source they could eradicate. The Ten Thousand survey revealed people who had shared rooms and meals with the victims but not fallen ill themselves, making it difficult to pinpoint what it was about the Bellevue-Stratford — or if it was truly something in the hotel at all — that made people sick.

Almost no hotel workers seemed to get sick, with the exception of one single air conditioning repairman. How could only visitors be susceptible to the strange illness? It seemed for a while that the disease may have come from the sidewalk outside the hotel where people had watched a parade. Was the cause some substance that had originated from the parade?

The public as well as the scientists fixated on the Bellevue. A boy who had attended a magicians’ conference just before the Legion meeting admitted over a hotline that he had thrown a smoke powder into a vent, and he fretted that he had caused the deaths. Another woman reported that she had heard parakeets chirping in a room and wondered whether they could carry disease. (The magician’s powder and ornithosis — an infectious disease that originates with birds — were both quickly ruled out.)

Shortly after the outbreak became public knowledge, most people had started avoiding the Bellevue altogether. One reporter noted a couple walking by while he stood in front of the hotel in the weeks after the convention. He wrote that the couple stopped when they noted the sign announcing the hotel’s name.

“They studied the building for a moment,” the reporter wrote later, “and then veered toward the curb so as to walk as far away from the door as possible.”

Another reporter listened in on a conversation among a group of bellmen, looking up toward the ornate ceiling, where a CDC investigator collected samples of soot from a ventilation shaft. “If it’s up there,” one remarked to his coworkers, “we’re dead.”


David Fraser, known at the CDC for his unflappability, was a little shaken as the lawyer stood to go. On August 14, the epidemiologists had officially added a new criterion for getting a diagnosis of Legionnaires’ Disease: having been inside the Bellevue-Stratford Hotel. The attorney charged in shortly afterward, bearing a threat of legal action if Fraser was not correct about the source of the epidemic. The fiscal promise of the Bicentennial summer had been smashed to pieces; once media accounts started linking the Bellevue-Stratford to the disease, occupancy dropped drastically. Conventions canceled. Money hemorrhaged from the hotel’s accounts. Livelihoods were at stake.

sketch of the top of an ornate building

Dr. Fraser knew that he would be supported by the full weight of the federal government if he successfully demonstrated the source of the epidemic. So he had better be right that it was the Bellevue.

The victim count had stabilized: 29 people had died, 16% of the 182 who had come down with symptoms of the disease. One out of every five or six people who got this pneumonia would die of it. That death rate rivaled the 1918 flu pandemic and was twice as high as the death rate for a common bacterial pneumonia. People who had not been physically inside the Bellevue but showed the same symptoms were classified as having “Broad Street pneumonia.” Counting these meant 34 more people with the disease, and five more who had died.

A few more cases had been found among participants in the International Eucharistic Congress the week after the legionnaires left Philadelphia, but to everyone’s immense relief new cases seemed to stop cropping up by mid-August.

Fraser understood that the managers of the Bellevue found themselves in a near-impossible position as the media had a field day with the implication of the Grand Dame of Broad Street in the epidemic. But his role was as an epidemiologist overseeing talented scientists working hard to apply their unique skills to this quagmire. He could not be responsible for the Bellevue’s public relations problem.

He also couldn’t help the 19 violations of the plumbing code that the CDC investigation uncovered at the Bellevue, which in turn got picked up by the media. The hotel had connected their air conditioning system to their drinking water system, which was unlawful although probably not dangerous. That certainly wasn’t the CDC’s fault.

On the 18th of August, the day after the news of the plumbing violations, Secretary Bachman made a point of saying in one of his still-frequent press conferences that “there appears to be no evidence of on-going risk to persons staying at the Bellevue-Stratford Hotel.” But by the 19th, the opulent and normally packed Bellevue-Stratford dining room had only four people in it for dinner. The usual breakfast crowd of a couple hundred dwindled to just a handful. A newspaper reporter described a salon worker buffing her own nails and polishing those of her manager.

By the end of that week, the epidemiologists wrapped up their work at the Bellevue and headed back to their normal posts, further emptying out the enormous hotel that, just a few weeks prior, had been full nearly to capacity.


By the end of August, the intense public pressure on Fraser to find an answer reached a fever pitch. Congress had asked the CDC for weekly updates on the investigation, and one Senator’s office seemed to take pleasure in publicizing every negative report.

And then, on August 26, Lawrence Altman filed a story for the New York Times describing a hopeful development. After the contamination setback, the repeat tests run in the lab of William Sunderman, Jr., came back positive. There appeared to be five times more nickel in the victims’ tissues than in the tissues of people who did not have Legionnaires’ Disease. The Sundermans needed to do more testing, but the results buoyed the nickel carbonyl theory.

Altman informed the Philadelphia Health Commissioner of the results. “If that’s true,” the commissioner responded to the reporter when asked for comment, “that would be very significant. We would still have the problem, if it is nickel, of how it got into people.”

Dr. Sunderman, Jr., whose father had first posited the nickel poisoning theory, speculated in response that people may have inhaled it. Sometimes it was used to make carbon copy forms; maybe a pile somewhere had been burned. After all, there was a sanitation workers’ strike in Philadelphia during the Legion conference.

Altman, sure he was onto something, began filing near-daily updates on all the new lab results he could find, mentioning the nickel carbonyl theory whenever possible. As time went on, the tone of his coverage reflected the public impatience with the investigation. On August 30, he penned an article describing in detail the frustration in Dr. Sunderman’s lab at the limited specimens they had received from CDC for testing. Without more tissue, results would stay inconclusive.

Then, in a somewhat jarring pivot, Altman provided a detailed criticism of the quantity of information the CDC released to the general public about the results of lab testing. He capped off the article with a jab at Bachman, reporting that the Secretary was at a cabin in Maine — which was factual, although he had also been in daily contact with the investigators.

Altman’s reporting got under Dr. Fraser’s skin. The epidemiologist chafed at the continued coverage of the nickel carbonyl theory, especially from a former EIS officer. Nothing about the theory rang out as convincing. Autopsy practices still possibly tainted the lab results, and no one had yet proposed a plausible way for nickel carbonyl to have entered the legionnaires’ systems. A Washington Post reporter approached him about the issue.

“My editor asks me about when I’m going to file a story about nickel carbonyl,” the Post reporter told Fraser. “And I’ve told him that I don’t see a story there. Dave, tell me, am I missing something?”

“You’re not missing anything.” Fraser responded in consternation. “This is an idea that’s taken on its own life, quite unrelated to the science, either in the laboratory or the epidemiology, so we need to not be swept up by it.”

Still, the nickel results pointed to a real and rather urgent problem: the amount of material available for testing dwindled every day. It would soon run out.

Meanwhile, the CDC labs in Atlanta remained all-hands-on-deck to determine the cause of the epidemic. In a move totally unprecedented and still never repeated, every laboratory at the institution had been ordered to put a hold on other projects and pick up samples from a central repository of autopsy specimens from the Philadelphia outbreak, a room that one longtime CDC scientist would later describe as the “grand central autopsy suite.” The agency asked every lab to test the samples for whatever it was they specialized in. It was in this vast, desperate wave that McDade initially ran his tests. But like the rickettsial studies, every single test came back negative.

On September 8th and 9th, a group of expert pathologists convened to review the glass microscope slides prepared by the CDC labs. They sought to inject a fresh perspective into the investigation. They concluded that the lung disease seemed to have more in common with a condition called “diffuse alveolar damage” and less in common with bacterial illnesses. Alveolar damage, they thought, pointed toward a toxin, or perhaps an as-yet unknown virus.

One dissenter, a virologist, still believed that the disease could be bacterial. There had been a few bacteria on the slides of lung tissue he had looked at. But the other pathologists disagreed. Like McDade, the rickettsial researcher, the group thought that the small pink rod seen on the slides, unremarkable and present only sporadically, represented nothing more than a contaminant.

The issues surrounding the investigation came into even sharper relief when, on September 15, the New York Times ran an editorial stridently criticizing the CDC’s approach to the epidemic and questioning the initial assumption that the cause of the outbreak was a contagion, like a virus or a bacterium.

“Unfortunately, however,” the article read, “because of the assumptions on which the epidemiological study began early last August, samples of victims’ tissues and body excreta which are needed to investigate different toxins are now exceedingly scarce… Even laymen may wonder why toxicologists played so little role in the early state and Federal investigations, and may hope that the initial errors made in studying legionnaires’ disease will not be repeated in future investigations of outbreaks of mysterious illnesses. The Center for Disease Control has not added to the luster of its record by its performance here.”


Anna Taggart’s blood boiled. The elevator operator had watched her beloved hotel decline in public stature for weeks. The wealthy and beautiful no longer bustled about the lobby; instead, Taggart cringed as scientists climbed the walls to peer into vents or collect samples of leaking air conditioner fluid.

Those legionnaires had brought their own disease. She was sure of it. This hotel had never made any of its employees sick. Those rowdy war veterans had brought in their own food and drink. It didn’t make any sense, all of this circus over the Bellevue. She watched disdainfully as a TV reporter stopped people in the street outside the hotel to ask them all about the disease inside the Bellevue-Stratford. Utter nonsense.

Not. Today.

Disregarding instructions from hotel management not to comment to the media, Taggart marched out the front door of the hotel, onto the street, and right up to the interviewer.

man holding a video camera

“Put down that horn,” she ordered the surprised news reporter, and then proceeded to lay into her about the media coverage of the Bellevue. They were so quick to criticize and to jump to conclusions, she said. It wasn’t fair.

She then turned to address the hotel directly.

“You are a lady,” she intoned. “You have grace. And these punks, this bimbo,” she gestured to the reporter, “these squirts,” motioning toward the camera crew, “aren’t going to take that away from you.”

Turning back to the reporter, she finished her tirade with a warning.

“Don’t twist my words around or I’ll personally punch you in the mouth.”

Then she stalked back into the hotel.


Philadelphians with loyalty to the ailing Grande Dame of Broad Street initiated personal missions to save her. Even the governor of Pennsylvania stopped by the hotel, dining there on August 23 as a publicity boost, but to no avail. That very day, a bartender from the hotel entered the hospital with pneumonia and died several days later. Despite doctors’ assurances that the pneumonia was not the same one that had killed 29 people, the media linked the death to Legionnaires’ Disease and to the hotel.

A few days after the bartender’s death, which garnered nationwide news, a committee calling itself the Friends of the Bellevue formed and started giving out bright orange lapel pins to everyone entering the building that read “I Love the Bellevue.” A few weeks later, on October 3, the owner of a jewelry shop inside the hotel took out a full-page ad in the Sunday edition of the Philadelphia Inquirer to publish a love letter to the institution. She enumerated the dignitaries who had been “impressed by your majesty.”


“Kings have squired you,” she wrote, “and queens have admired you, but you love your Philadelphians best!”

And truly, Philadelphians seemed the only ones still in love with the Bellevue by the fall, as 30 of the 50 conventions scheduled for that summer were cancelled. Even a meeting of a psychic society held in the hotel specifically to try to divine the cause of the outbreak saw fewer than half the expected number of participants.

On October 8, the Friends of the Bellevue staged a $50-a-plate United Way benefit portrayed as a “get well party” for the ailing hotel. Media accounts compared the lavish, champagne-drenched event, attended by more than a thousand well-heeled politicians and socialites, to the last time the Bellevue had been so busy and festive, back in July, at the Legion conference. The legionnaires, by contrast, had been a “noisy, ill-dressed, no-tipping bunch who filled the bathtubs with beer-on-ice and shunned the hotel’s bars,” wrote a reporter for the New York Times, conveying comments of doormen and bartenders.

“It hasn’t been this nice for so long,” one waitress remarked wistfully, tears in her eyes as she carried an ice cream cake and champagne out to the party.

And, as many present might have predicted, given the Bellevue’s $10,000-a-day losses, that was the last time it would be that nice for a very long time. On November 11, 1976, the day that the hotel’s losses totaled $1 million, management announced that the Grande Dame of Broad Street would close its doors to the public.

Five hundred employees, including Anna Taggart and others who had practically grown up in the Bellevue, would be out of work without severance. Tears filled Anna’s eyes as she reiterated her opinion to the media after the announcement that the legionnaires did this to themselves, and that the hotel had nothing to do with it.

The following Monday, a scientific symposium hosted by the American Lung Association was one of the last events held in the Bellevue function rooms. Attendees discussed progress in the study of Legionnaires’ Disease.


By the time the Bellevue took her dying breath in November, the CDC investigation had ground to a standstill. And to add insult to injury, yet another bout of bad press plagued its laboratories and administrators.

A New York Congressman named John Murphy, who had been vocally critical of the investigation for months, hauled in investigation officials, including Fraser, to answer questions about the lack of progress on the Legionnaires’ Disease outbreak. One of the first to testify in the two-day hearing was George Chiavetta, the legionnaire who had spotted the suspicious man in the royal blue leisure suit. Sitting next to his lawyer, Chiavetta described a man dressed in unusual fashion, with something in his pocket.

newspaper clipping of a dour-looking man

“I told my wife that the man might be some sort of a nut or something,” Chiavetta elaborated. He went on to explain that he had reported it to the police, but he didn’t think the police officer ever received a reply from the health department.

Chiavetta’s testimony didn’t entirely have the effect he had hoped. In response, another legionnaire explained to the media that a lot of unusual people hung around the hotel during the conference. “I even got accosted by a couple of old ladies,” the second legionnaire told a reporter. “I don’t think it had anything to do with the illness.”

Later in the hearing, Secretary Bachman testified, as did other Pennsylvania officials involved in the investigation, but the panel of CDC officials was the main event. When their turn came, Fraser and another CDC representative flanked Dr. David Sencer, the head of the CDC at the time.

With his usual exacting care, Fraser had prepared remarks for Sencer to deliver. The speech attentively and precisely represented the state of the science, which remained inconclusive as to the exact cause of Legionnaires’ Disease, despite bringing to bear the largest number of people and resources ever committed to a CDC investigation. The statement also eloquently expressed compassion for the victims and their families and then acknowledged medical science as imperfect; some diseases just could not be understood with current science and technology.

sketch of a man with a pencil mustache and side part

The Congressional subcommittee shot back. Rep. Murphy opened his remarks with a sideways comment that, yes, the panel understood well enough the fallibility of the CDC. “To be charitable,” the Representative said at one point during the hearing, “they botched it up.”

The Representative once again bemoaned the agency’s failure to consider toxic substances sooner, limiting the testing that could be done. He invoked the nickel carbonyl hypothesis, although at that point even Altman had stopped reporting on it after the lack of definitive evidence rendered it a dead end.

A Congressional report widely covered in the press read: “It is totally unacceptable that in a country of 220 million people… supposedly with the most advanced technology in the world, we find ourselves in a position of not knowing what happened in Philadelphia and even worse, not being in a position to prevent it from happening again.”

Later in November, Fraser and his assigned EIS officer, Ted Tsai, completed their report on the epidemiology of the Philadelphia outbreak. It was a 71-page behemoth called an “EPI-2” in CDC lingo, a comprehensive review of the data compiled during the field investigation. (The name is in contrast to an “EPI-1,” a brief initial outline of the problem prepared at the beginning of the investigation.)

In an effort to inject new life into the still-ongoing lab research, they distributed it widely. They needed people to keep thinking.


A copy of the EPI-2 landed on researcher McDade’s desk in the leprosy and rickettsia lab. He admired the report’s thoroughness, not thinking he could add to the science. Nonetheless, he would periodically ponder the investigation — now poised to be CDC’s most conspicuous failure to solve an outbreak — through the fall and winter.

By Christmas, public frustration remained, though in muted form. In an interview he gave in December, Health Secretary Bachman refused to give up hope on the five-month investigation. “As people we are not acclimated to failure,” he admitted. “But we should keep in mind that God, in all his omnipotence, has not yet deemed it appropriate to share all the secrets of the universe with us.”

At the end of December, Dr. McDade was at a Christmas party with his wife in their neighborhood when a stranger started talking to him about the investigation. McDade relates that the man became a little aggressive over the subject. In McDade’s account, the stranger said something along the lines of, “Everybody knows all you scientists at CDC are kind of…” Here, McDade pauses, and raises his expressive eyebrows. “Weird.”

“But you know,” the man continued, “we count on you to figure these things out. There’s something out there that’s killing people, and it could do it again, and we don’t know what’s causing it. That’s really scary.”

McDade took the man’s comments to heart. But what could he do to allay his neighbor’s fears? As a lab scientist he had done what he could to contribute to the investigation.

A week or so later, as was his habit of years, he took the opportunity of the quiet week between Christmas and New Year’s Eve to clean out his lab.

“I remember going into the back lab where the microscope was,” he described later for an audience at the CDC. “There was a wooden slide box that had the slides from the guinea pig tissues, and I thought, well, you know, I should probably take one more look at these before I file them away. So I started looking.”

He was frustrated by the tiny, pink rod-shaped bacteria he saw scattered over his slides, the ones he had previously dismissed as contaminants. He was a methodical, detail-oriented scientist, and not knowing what they were was driving him a little crazy.

He later described searching for the tiny pink rods as akin to, “looking for a contact lens on a basketball court with your eyes four inches above the ground.”

And then, in a far corner of a single slide, something stood out. Not a single bacterium, unlikely to be significant, but a small cluster.

“[That] said to me,” he remarked, “this is something which is actually growing there. I don’t think that this is something that I can overlook. So I’m hooked, I have to do something about it.”

He thought back through the rickettsia protocol and considered ways to get this organism to grow so that he could identify it.

The protocol called for giving the egg cultures a little dose of an antibiotic to kill contaminants. But maybe if he skipped the antibiotic dose, these tiny pink bacteria might grow, and he could find out what they were. If he identified a common contaminant then he could be satisfied that this pink bug was not the cause of the epidemic. So he settled on a strategy: He would implement a variant of the protocol without the antibiotic, inoculating the yolks with the guinea pig serum to observe what grew.

Several days later, he looked again at samples from the yolk cultures. This time they teemed with long, pink-stained, rod-shaped bacteria — bacteria he didn’t recognize as any familiar contaminant and that hadn’t shown up in any other cultures in any lab in Atlanta or Harrisburg.

In addition to McDade’s doggedness, the discovery involved a strong measure of chance. Scientists who didn’t work with rickettsiae didn’t use the pink dye that McDade employed in his work. Most bacteria were identified with different stains, stains that these bacteria, which were bright fuchsia on McDade’s slide, apparently didn’t pick up; it would have been almost impossible to see them on these more common slides. The bacteria also didn’t thrive on petri dishes that used conventional nutrients that made most bacteria grow; but they grew in guinea pig lungs, and they grew in the egg yolks that McDade used for his rickettsial research.

The scientist rushed to tell Shep, his boss, who ran the finding up the chain and got permission to use precious autopsy samples for more experiments. McDade got to work. He devised an experiment that would show whether or not the Legionnaires’ Disease patients had antibodies to the observed bacteria in their blood when they died.

A few days into the new year, he and his lab technician finished the experiment. If the right antibodies were there, they should see a fluorescent green glow under a UV microscope. If they were not, the slides would look dark.

They glowed apple green.

Over the next two weeks, McDade and his laboratory staff ran confirmatory tests. On Friday, January 14, 1977, they went to report their findings to David Sencer, the head of the CDC. As they sat in Sencer’s office, McDade and Shepard hesitated for a few moments before they spoke. They knew what they had, but could hardly believe that their lab had solved this seemingly unsolvable problem. Then they came out with it: they had isolated the organism responsible for the Legionnaires’ Disease outbreak.

Sencer, as the head of a public health organization, immediately wanted to publish the results, but the meticulous lab scientists wanted time to repeat the experiments, perhaps in a different lab, to be absolutely certain their discovery was not a fluke. Sencer told them they had until Tuesday. Then the CDC would publish their results in a special Tuesday edition of the weekly Thursday news bulletin, the Morbidity and Mortality Weekly Report, alongside a press release.

The rickettsia lab worked through the weekend to confirm and write up their findings. On Tuesday morning, with their report already at the printer, Shepard reported to Sencer that they had also tested specimens from Pontiac and from another outbreak of pneumonia in a psychiatric hospital in Washington, DC, that had never been solved. Those samples were positive too.

The presses were held, and the story was amended.

A scheduled conference call included Pennsylvania health officials, the US Surgeon General, the National Institutes of Health, and everyone at the CDC who had anything to do with the epidemic investigation, from the lowliest “backbencher,” as McDade described himself, to the heads of the laboratory and epidemiology divisions. Shepard and McDade then attended a press conference where they announced their findings. There, they distributed the special Tuesday edition of the MMWR, the only special edition of that publication ever to be published.

Larry Altman, who had so doggedly chased the nickel carbonyl theory, and also, while in the EIS had happened to be the editor of the MMWR, was on assignment in Geneva reporting on a smallpox outbreak. He did not cover the press conference.


When McDade found antibodies to the bacterium that soon became known as legionella pneumophila in samples from the Pontiac outbreak, another piece of the puzzle fell into place.

The complex egg yolk culture method required serum from live animals; it was not easy to take an environmental sample — of soil, dust, or carpet fabric — and prove that legionella lived there. So even with the organism identified, a crucial question remained: how had the Bellevue-Stratford made people sick?

In the Pontiac case, investigators never identified the bacterium, but they did conclude that the building’s cooling system had somehow acted as the source of the disease. CDC experts had painstakingly modeled the airflows out of the top of a cooling tower on top of the building and then down the sides of the structure, which would have exposed people in the building to anything being blown out of the cooling tower.

In the case of the Bellevue-Stratford, those airflows, the investigators posited, had then swept around to the front of the building, onto Broad Street, and into the front door, where they settled in the lobby.

By the time McDade, Shepard, and their research assistants had developed the right set of nutrients to allow scientists to culture the finicky bacterium in a petri dish, the air filters at the Bellevue had been changed and the cooling tower cleaned, so they could never demonstrate the source with absolute certainty. CDC officials later theorized that, twice in the Bellevue’s history, this small, hard-to-grow bacterium met exactly the right set of weather conditions to allow the bug to multiply and be carried on air currents out of the cooling tower, down to the street, and into the lungs of the victims.

They would never completely understand why hotel employees did not get sick during the outbreak, but a combination of reasons likely factored in: on average, the hotel employees were younger and healthier than those who died, and many of those who were tested later had antibodies to the bacterium, meaning that they had likely been exposed but not gotten ill. Employees who had started work before the Odd Fellows conference in 1974 were much more likely to have antibodies to the bug, so may have gotten sick during that earlier conference and by 1976 had antibodies in their blood to protect them.

The Bellevue, back in business since 1978, has never had another outbreak.


It is unclear but unlikely that Maria Reeves, the sex worker, was ever added to the case list of Legionnaires’ victims. Treated in a hospital in New York, she had been beyond the reach of the health workers scouring Pennsylvania hospitals, and she had not disclosed her presence in Philadelphia to the intensive care doctors who treated her while she was ill.

Only several years later would she be identified and interviewed for a book about the outbreak. She still did not want to use her real name. She only had vague memories of what happened in the hospital that August, of blood tests being taken, of examinations by nurses and doctors. She did know that she was treated with an antibiotic called erythromycin.

She left the hospital having survived her second bout of severe pneumonia in one year, a very narrow escape given her harrowing condition when she arrived in the hospital in New York.

Maybe it wasn’t luck. Ultimately, grateful to God for allowing her to live, Maria experienced a religious awakening, became a devout Christian, and gave up sex work for good.


In the years since Dr. McDade first identified legionella pneumophila, there have been multiple large outbreaks of Legionnaires’ Disease, some more fatal than others. The bacteria constitute part of a large family of ubiquitous organisms that live in many moist, warm environments but for which the biology remains only partially understood. They live in urban centers, in cooling towers, air conditioning systems, swimming pools and fountains, but they also live all over the natural world, in natural pools, hot springs, and soil. They require very specific conditions and nutrients to grow in a lab, but they have always been in our environment, hiding in plain sight.

Legionella now has its own program at the CDC with a specialized staff. In addition to responding to outbreaks as needed, they produce handbooks and guides on the prevention of Legionnaires’ Disease, mainly through maintenance of cooling towers and other water-containing systems. Legionellosis — the more modern, formal name for Legionnaires’ Disease — remains a dangerous infection, treatable with one of several antibiotics when caught in time. The most common one is azithromycin, close cousin of erythromycin, the drug that just by chance was tried on Maria Reeves, and may have saved her life.

Forty years after the outbreak, McDade and Fraser participated in an event marking the milestone at the CDC. A researcher in the audience, paraphrasing science fiction legend Isaac Asimov, remarked that the legionella story epitomized how “most advances in science are not accompanied by ‘eureka,’ but rather by, ‘Well, that’s weird.’” In response, McDade ruminated on the importance of pushing science forward by thinking carefully about the strange and anomalous. The things, in his words, that “just drove you nuts.”

several people sitting on a stage facing an audience, with one person in particular highlighted to stand out

ALEXANDRA CORIA, MD is a doctor, writer and professional globetrotter, currently living in Brooklyn, NY.

How was it? Save stories you love and never lose them.


Logo for Truly*Adventurous

This post originally appeared on Truly*Adventurous and was published March 30, 2020. This article is republished here with permission.

If you're new to Truly*Adventurous and want more incredible true stories:

Get the T*A newsletter