Two sets of figures are commonly cited, and both are correct.

The epidemiological case count from the investigation was 182 people ill and 29 dead — 149 Legionnaires and 33 others who had been in or near the hotel without attending the convention.

Broader tallies of 221 ill and 34 dead appear widely and include additional cases counted in and around the outbreak.

Where a single figure is needed, the first is the count drawn directly from the investigation.

That second group in the first figure — the 33 people who were near the hotel but not part of the convention — was the detail that eventually pointed investigators in the right direction. Among them were a bank teller who worked across the street and a bus driver who had transported cadets in a parade. They had not been to the convention. Some had not been inside the hotel.

Whatever was making people ill was not confined to the building.

Six months without an answer

The investigation became the largest field inquiry the Centers for Disease Control had ever undertaken. It ran for six months, in public, on the front pages of the Philadelphia papers.

Nickel poisoning was considered. So was sabotage. Early on there was a genuine fear that this was the beginning of a swine flu pandemic, and that fear shaped the national response for weeks. By September the focus had moved away from an outside cause and toward the hotel itself.

Meanwhile the Bellevue-Stratford, one of the grandest hotels in the city, was emptying. It closed in November 1976 for lack of business.

January 1977: the bacterium is identified

The answer came from a CDC microbiologist named Joseph McDade, working through stored tissue samples after most people had stopped expecting one.

He had already examined the samples once and found nothing resembling a known pathogen. He went back and changed the question — instead of looking for something he recognised, he looked for anything he did not. He later described the process as searching for a lost contact lens on a basketball court with his eyes four inches from the floor.

In January 1977 he found it: a rod-shaped bacterium that had not been described before. It was later named Legionella pneumophila — literally, lung-loving.

Environmental investigation traced the outbreak to the cooling tower serving the hotel’s air conditioning. Mist from the tower carried the bacteria into the building through its air intakes, and outward into the street — which explained the bank teller and the bus driver.

How Legionnaires’ disease got its name

The disease is named after the American Legion because the Legion’s convention was where the first recognised outbreak occurred. The Legion did not cause it, was not responsible for it, and lost members to it.

Names given in the middle of a crisis tend to survive it. Fifty years on, an organisation of military veterans is permanently attached to a bacterium that was in the hotel’s cooling tower before they arrived.

It had been there all along

Once McDade knew what to look for, he looked backward. The bacterium turned up in stored samples from outbreaks that had been closed unexplained:

• 1965, Washington DC. An outbreak of pneumonia at St Elizabeths Hospital that infected 78 people and killed 16.
• 1968, Pontiac, Michigan. An illness that swept through a county health department building, sickening 144 people. It was mild and nobody died — it is now known as Pontiac fever, the less severe form of the same infection.
• 1974, Philadelphia. An outbreak at a convention of the Independent Order of Odd Fellows that killed two people. It had been held at the Bellevue-Stratford.

That last one deserves its own sentence. The same hotel had an outbreak two years before the one that made the disease famous, and nobody recognised it — because nobody yet knew there was anything to recognise.

What has changed since 1976, and what has not

The 1976 investigation produced real progress. The organism was identified. Diagnostic tests followed. Cooling tower design and maintenance practices changed worldwide in response.

What has not changed is the underlying situation. The source of the first recognised outbreak was a cooling tower. Fifty years later, cooling towers remain a leading source.

Reported cases in the United States have risen roughly ninefold since 2000; adjusted for the changing age of the population, incidence rose about fivefold. And in most of the country, the buildings operating those towers face no legal requirement to test them, report results, or maintain a water management programme.

The science was settled in 1977. The obligation still is not.

Explore further

• What is Legionnaires’ disease? — causes, sources and who is most at risk
• Outbreaks — how they happen and why most cases are never traced
• Community and workplace — cooling towers and the people they reach
• Why legislation — what prevention requires, and where the law demands it