Legionnaires’ disease – narrowing the list of suspects

February 2006

In 1976, when the legionnaires’ disease outbreak occurred in Pennsylvania, the Bellevue-Stratford Hotel was the grand dame of Philadelphia. It was an old hotel, but like the Waldorf Astoria in New York, it had class. It catered to a special clientele, especially those who came from old Main Line money. Presidents and movie stars stayed there. Politicians and cultural organizations held important fund-raisers there. It was a swanky place to stay in if you were anyone who mattered.

Therefore, it was not too surprising to learn that the hotel management was none too pleased with the implication that the outbreak started in this grand old hotel. In fact, the managers threatened to sue David Fraser, MD, which Fraser found a little disconcerting.

“Well, that’s how I interpreted what their lawyer said,” said Fraser, who headed the field investigation for the CDC. “They were unhappy about the situation and were constantly expressing their views about how the investigation was going.”

The CDC, however, could not ignore the risk factors for this mysterious pneumonia: older age, men, smokers, a former delegate to the convention and time spent in the hotel lobby or on the sidewalk in front of the Bellevue.

“We found that about 15% of patients died,” said Stephen Thacker, MD, an Epidemic Intelligence Service (EIS) officer at the time. “That was the case-fatality rate, and those [people] were more likely to be older, to have chronic lung disease, [to be] smokers, the usual things you might expect [to make pneumonia more fatal],” he said.

In all, 221 people came down with legionnaires’ disease and 34 died during the outbreak.

Investigation cooperation

Although Bellevue hotel staff and officials cooperated with the investigation, they were resentful. “It was a high-stakes game, and I heard things about the financial circumstances of the owners. I tried to reach my own conclusions about why they might have been leaning on me so hard,” said Fraser. “Yes, there was pressure. There was the pressure from the press because there was public interest and because there was a big unknown. There were people dying, and members of the press would get their own theories: Either they’d pick them up from other people, or they’d get a bee in their bonnet. Then, they’d press us about competing theories. Congress was ready to get up on its soapbox, so that led to a hearing that was held in Philadelphia,” he said. “It was very uncomfortable.”

The City of Philadelphia was also concerned about the investigation. This was the bicentennial year, and tourism was up, as visitors came to see the birthplace of the nation. The president was due in the coming weeks on a campaign stop, and the pope was also scheduled for a visit in the near future. Were they in danger? What if they didn’t come?

The city approached it differently from the hotel, however, and it was quite cooperative, Fraser said. In addition to giving the CDC clear access to its city health department and personnel, city officials sent 40 police detectives to help. “What am I going to do with them? I had to come up with a job that the police detectives could do. I don’t know anything about police work, but they talk to people, and I figured they could probably do that and probably do it very well. So, we gave them this task. It was a lot of phone calls, a lot of chasing down people. They did a lovely job, and it was important to the investigation,” Fraser said, another first in an unusual outbreak.

By the time the EIS team consolidated its efforts in Philadelphia sometime during the second week of the investigation, they were pretty certain that a respiratory agent was the cause of the illness. However, it was not a known virus or bacterium.

They had eliminated the possibility of a food- or waterborne pathogen that had been ingested at one of the convention events, and the agent did not appear to spread from person to person.

Once they pinpointed the Bellevue-Stratford as the source of the outbreak, the EIS officers needed to know where people became exposed. “Well, it was very interesting. We did several surveys and a case control study, and we asked all the questions to complete a seven-page questionnaire. You know: Where were your meetings? Where did you party? The wives incidentally stayed at a different hotel,” Thacker said. “They stayed at the Ben Franklin.”

“They would come to the Bellevue-Stratford, but they stayed at the Ben Franklin. It didn’t matter where you were in the hotel, what meetings you attended, what you ate, where you ate. None of those sort of things were connecting. So, we started looking at, well, where did people go that we didn’t ask about,” he said.

“Dave [Fraser] and I actually talked about this. You know, we’d get back to the hotel around 8:00 or 9:00 at night. Then, he would call Phil Brachman, who was the director of epidemiology at the time, to update what happened all day. We asked, ‘What about the elevators?’ Because everybody went on the elevators,” Thacker said.

“So, we had the elevator hypothesis. We had people climbing and getting specimens and asking questions. Did you use the elevator? How often did you use it and so forth? Trying to get a dose response and that sort of thing,” he explained. The elevators proved to be a false lead, just like everything else.

Floating Theories

The CDC asked outside experts to review specimens and suggest possibilities, but no one had an answer. “There was a score of theories floating around,” Fraser said. “I think the fact that it was a legionnaires’ conference on the bicentennial had conspiracy theorists particularly alert to the possibility of terrorism. We didn’t see any evidence of that. Of course, it didn’t turn out to be the case. However, that doesn’t slow people down from having theories.”

Brachman, who was director of the CDC’s epidemiology office at the time, remembers a story in the local paper. It featured a man with his back to the camera with his open attaché case on the counter. “The theory in the article was that, in opening up his attaché case, he had released some toxic substance that got into the air circulation system in the hotel,” said Brachman. “Oh, boy, everything under the sun was suggested.”

Souvenir packages, which contained cigarettes, were distributed to the legionnaires. “In those days, they handed out cigarettes like mad. Rumor was, they had been poisoned, and the people who were smoking them would inhale a toxic substance,” he added. “The papers were full of rumors.”

“We didn’t know what it was at that point. There was a lot of pressure that it was a toxin. Very few people at that point were thinking it was an organism,” Fraser said. “The real pressure was on it being nickel carbonyl or some other chemical.”

“It was getting pretty frustrating,” said Brachman. Sen. Richard S. Schweiker (R-Pa.) was concerned about the problem, and he relayed that concern to then-CDC director David Sencer, MD. “He would initially be calling back to Sencer to find out, ‘What’s going on? Why can’t you guys solve this? Don’t you know what you’re doing?’ Eventually I was the liaison between the CDC and the senator’s office. Once a week, I’d have to call up Washington and give a report. Well, the reports were always, ‘We still don’t have an agent,’” Brachman said.

Almost from the beginning, people were concerned about the air conditioning unit, located on the roof. It was old and leaky. Could it be the problem? It might even provide the answer to the Broad Street cases: Perhaps, water sprayed off the roof from the unit. Repeated looks at the unit, however, did not find a source.

“One of the things we observed is that the air conditioning in the lobby was right over the registration desk, so right where you came in,” said Thacker. “One of them, the flags were over it, and it wasn’t blowing. But in the days between the time of the convention and [our arrival], the hotel changed and cleaned the filter. So, we had no idea. We took specimens there, but the hotel cleaned it with an agent that turned out to be the killer of the bacteria, chlorine.

“So, we never proved it was the air handling system. We just suspected that it was that,” he said.

“The other interesting thing is that people who worked there didn’t get sick, including the guy that cleaned the air handling system. Later it was determined that some employees had antibodies,” Thacker said. “It clearly had to do with exposure as well. Maybe they were younger and healthier than the group of people who became ill.”

No answers

After 3½ weeks, it was time to return to the CDC. The EIS team left without an answer. “We knew a whole lot about this epidemic,” said Thacker. “We ended up with over 220 cases that fit our case definition. We knew that the laboratory tests that we knew of – bacterial, viral rickettsial, fungal – were negative for this. This was something new.

“Even though it looked like an infectious disease, it may not be, and we already had involved toxicologists studying specimens. Because they were legionnaires, because it was a bicentennial and because the pope was coming in two weeks, the FBI was involved,” Thacker said. “We got consultation from the CIA because there was a fear this might be a terrorist event, too. The police worked very effectively with us, particularly in Philadelphia. They had their own room in the health department. They helped us find things and find people, which, of course, they’re quite good at. It was quite an extensive investigation. We took environmental samples from all over the hotel to see if we could grow something. After the 3½ weeks studying specimens, we knew a whole lot about it, but we didn’t know what it was.”

That did not suit the city and state health officials, the media or Congress, and Fraser and others were called back to answer questions at a congressional hearing.

“That was in November, and it was very painful,” Fraser said. “Congressman [John M.] Murphy (D-Rep.) from Staten Island[, N.Y.,] headed that and spent a lot of time telling us how incompetent we were. Of course, we weren’t feeling good about the situation in November.”

Solving other outbreaks

At one point, Fraser’s supervisor, John Bennett, MD, took him aside and told him that when he solves the puzzle and finds the etiologic agent of legionnaires’ disease in Philadelphia, they would also find the cause of St. Elizabeths’ fever. In 1965, Bennett was an EIS officer who investigated an outbreak of pneumonia at St. Elizabeths Hospital in Washington, D.C., where at least 81 patients and staff were infected. There were 14 deaths. Bennett never found the source of the outbreak.

“John told me that fall, before Joe McDade found the agent, that when we found the agent for legionnaires’ disease, we would find the agent for St. Elizabeths’ pneumonia. There was no question in his mind it was the same thing. He was right,” Fraser said.

The CDC also found the cause of Pontiac fever. Pontiac fever was another unresolved epidemic that occurred in 1968 in Pontiac, Mich. There were about 144 cases but no deaths.

“Now, St. Elizabeths’ pneumonia was different from Pontiac fever. None of us thought that was the same agent because there’s no pneumonia with Pontiac fever. Nobody dies. They feel sick and all of that, but they don’t die,” he said. “It’s still a great mystery why those two diseases are caused by the same bug.”

Next month: Finding the agent and the aftermath.

For more information:

• Fraser DW. The challenges were legion. Lancet Infect Dis. 2005;5:237-241.
• Thomas G, Morgan-Witts M. Anatomy of an Epidemic. Garden City, NY: Doubleday; 1982.