Upper East Side Legionnaires' Killed 11: Is It Safe Now?
For most people who spent time on the Upper East Side this summer, the danger has passed. The last case was diagnosed July 21, and every cooling tower that grew live Legionella has been cleaned. What still matters is the two weeks after any exposure. If you develop fever, cough, muscle aches, diarrhea, or confusion within 14 days of being in Carnegie Hill or Yorkville (ZIPs 10028, 10075, 10128), ask specifically for a Legionella urinary antigen test plus PCR or sputum culture, and push for antibiotics fast.
Legionnaires' isn't contagious. You can't catch it from a neighbor, and drinking tap water won't give it to you. The bacteria live in warm building-water systems and reach you as inhaled mist, usually from a rooftop cooling tower. Once you've breathed it in, whether it barely registers or kills you depends almost entirely on who you are, not what strain landed in your lungs.
A rooftop, not a person, and not the tap
The outbreak clustered in ZIPs 10028, 10075, and 10128, with 94 confirmed cases and 11 deaths. City investigators tested 183 cooling towers on 160 buildings; 59 grew live Legionella on culture and all have since been disinfected and rechecked. The specific source tower hasn't been publicly named, and molecular typing from the city Public Health Lab is expected in September.
Cooling-tower plumes typically carry Legionella a kilometer or two under ordinary weather. Longer reaches happen. An industrial air scrubber in Sarpsborg, Norway seeded cases at homes over 10 kilometers away, and a Pas-de-Calais outbreak spread at least six kilometers from a single tower. Person-to-person transmission has been documented in one case report, ever. Swallowing contaminated tap water doesn't cause Legionnaires'. The risk is inhaling the mist, not drinking the water.
Who actually gets sick, and who dies
A healthy 30-year-old who breathes the same aerosol as a 70-year-old with emphysema may barely notice it. The 70-year-old can die.
Incidence climbs steadily with each decade after 50. Diabetes, chronic lung, kidney, or heart disease, active cancer, corticosteroid use, and other causes of impaired cell-mediated immunity each push up both the odds of infection and the odds of dying from it. The 2015 South Bronx outbreak killed about 12% of confirmed cases, largely because so many were in these groups. Community-outbreak fatality now runs about 4 to 5% overall, but ICU cases carry roughly 26% mortality, and reviews place mortality in immunocompromised or hospital-acquired patients as high as 40%.
| Risk group | Effect on getting sick | Effect on dying if infected |
|---|---|---|
| Healthy adult under 50 | Low baseline risk | Very low; usually flu-like if symptomatic at all |
| Adult 50 or older | Rises with each decade of age | Each decade adds substantially to the odds of dying |
| Diabetes | Elevated | Trend toward higher mortality, not always statistically significant |
| Chronic lung, kidney, or heart disease | Substantially elevated | Each is an independent predictor of severe outcome |
| Immunosuppressed (steroids, transplant, active cancer) | Highest | About one in three fatal in transplant recipients; up to 40% in immunocompromised or hospital-acquired cases |
The two-week window that decides the outcome
Legionnaires' looks like any severe pneumonia, which is why it gets missed. Incubation is 2 to 14 days. Onset is often abrupt: high fever above 39°C, a cough that may or may not produce sputum, muscle aches, and headache. What tilts the picture toward Legionella specifically is diarrhea, confusion or altered mental status, a serum sodium under 130, and a very high CRP. None are specific alone; together they should trigger targeted testing.
The fastest way to confirm the diagnosis is the Legionella urinary antigen test, about 74% sensitive across studies. Two caveats. It detects only L. pneumophila serogroup 1, which accounts for 50 to 80% of cases. And its sensitivity drops sharply in milder illness: in one large Spanish outbreak it caught about 86% of severe pneumonias but under 40% of mild ones. A negative antigen alone doesn't rule this out. If your symptoms fit the window, ask for PCR or sputum culture alongside it.
Which antibiotic matters less than starting one fast. If a plausible case is unfolding, it's reasonable to ask for empiric coverage while confirmatory tests are pending.
What still isn't answered
With all 59 culture-positive towers cleaned and the last case dated to July 21, transmission from that source has stopped. New York State passed a law in August requiring twice-yearly certified cooling-tower inspections, faster reporting of serious hazards, and higher penalties. That prevents the next cluster, not this one. Which specific tower drove this outbreak, and what the molecular typing shows, are both still pending.
For someone who lives or works in the affected ZIPs today, the practical answer is narrower than the headlines suggest. You don't need to avoid your neighborhood, your neighbors, or your kitchen tap. What you should do, if fever and cough or unexplained confusion turn up in the next couple of weeks, is walk in with the exact question: could this be Legionella, and can we test for it before waiting for the pneumonia to declare itself? Between the incubation window and the limits of the rapid test, that question is what decides whether this ends as a scare or something worse.