How is NYC's Health Department handling the Legionnaires' outbreak?

In the face of a new outbreak of Legionnaires’ disease, New York City’s Department of Health and Mental Hygiene quickly pivoted to a strategy of aggressive remediation and significant transparency, with more information promised shortly. But what about describing to the public the common course of Legionnaires’ in more detail?

Jul. 9, 2026 – If I had written this story on Monday afternoon, it would read very differently from the way it reads now. After the New York City Department of Health and Mental Hygiene (DOHMH) announced late on July 2 that it was investigating a likely community cluster involving parts of the Upper East Side of Manhattan, apparently arising from cooling towers, there was not very much specific information released over the course of the Independence Day Weekend.

DOHMH emphasized the importance of early diagnosis and treatment and assured the public that it was working as hard as it could, but did not give any details about what kind of testing it was doing (known), how many of the relevant cooling towers had been inspected (evolving), or what it would do in respect to towers that had tested positive (not necessarily decided).

By Monday evening, that had changed substantially.

Legionnaires’ disease can indeed be a nightmare

Legionnaires’ disease refers to a set of bacterial diseases, with the genus Legionella encompassing 60 or more species. The species of interest is Legionella pneumophila. There are 15 subgroups or “serogroups” within the species, and serogroup 1 (LP1) is the one that is found in water cooling towers that wind up causing disease.

LP1, like other types of Legionella, can be an especially nasty pneumonia, especially if not diagnosed and treated earlier, and especially in those with preexisting medical vulnerabilities. The Centers for Disease Control’s (CDC’s) webpage on legionnaires’ states that the case-fatality rate (CFR) for Legionnaires’ is approximately 10 percent. A 2015 water tower (LP1) outbreak in the Bronx had a CFR of nearly 12 percent;1a 2025 water tower (LP1) outbreak in Harlem had a CFR of about 6 percent.2

The facts about hospitalizations as commonly cited are also grim. The same CDC page describes hospitalization as “common”; a 2021 paper states that “as high as 90% of patients diagnosed will be hospitalized.”3 (The hospitalization rate in the Bronx outbreak was 93 percent4; in the Harlem outbreak, the rate was 78 percent.5) It is critical to understand that these rates are based on the subset of patients who are diagnosed. For an explanation of why the rates for all people who contract the disease – while still quite significant – are materially lower, please see the box at the bottom of this page.

The CDC page on Legionnaires’ also states that the disease is a type of “severe” pneumonia. I spoke with Janet E. Stout, a leading expert on Legionnaires’ disease who has been working in the field for more than 40 years. She confirmed that Legionella pneumonia is “much more severe” than is staph or strep pneumonia.

Stout, like DOHMH, emphasized the importance of early diagnosis and treatment with the right antibiotic (levofloxacin or azithromycin), which would not ordinarily be given if the physician does not think he or she is treating Legionnaires’. That prompt treatment is highly effective and patients treated early will generally do “very well.”

On the other hand, even modest delays in diagnosis and treatment, Stout says, begin to ramp up the morbidity associated with Legionnaires’. As the disease progresses, “you get something called acute respiratory distress syndrome. Your lungs are severely inflamed. They’re treating with steroids to reduce the inflammation. They’re treating with antibiotics to control the infection. Now you’ve got kidney involvement. So, it cascades rather quickly.”

Stout continued with a description of what happens when fluid accumulates in the lungs because of inflammation: that type of fluid build-up makes it “really difficult to breathe, and so then [the clinicians] intubate to assist in breathing.” Many patients, she concluded, “end up in the ICU because of the severity and the need for assisted ventilation.”

The current outbreak in NYC

According to DOHMH data updated as of 10pm on July 8, there have been 36 cases, with 22 hospitalizations (61 percent). So far, there have been no deaths associated with the outbreak.

When I interviewed William Fowler, the spokesperson for DOHMH, on July 8, he said that testing of the approximately 160 buildings in the portions of the Upper East Side under investigation was expected to be completed by the end of the day. In a follow-up email on July 9, Fowler confirmed that nearly all of the cooling towers in the affected area have now been sampled, by which he meant to include the caveat that more samples may be taken throughout the course of the investigation, as needed. 

There are two ways to test water towers for Legionnaires’. One is PCR testing, which you may be familiar with from Covid-19 testing. PCR testing is molecular testing (a DNA test). It provides fast (one- to two-day) turnaround for results. It has high sensitivity — 94 percent in one article that pools 16 studies from a variety of contexts6 — which means that it is quite good at identifying towers that genuinely do have the relevant Legionnaires’ pneumonia serotype (here, LP1). Put another way, there are few false negatives. It is highly likely that a tower that is PCR-negative really is negative for LP1. Stout and Fowler broadly agree.

On the other hand, there are towers that test PCR-positive where it turns out that what was detected was not live LP1, but rather dead LP1. This, as Fowler pointed out, is not a “false” positive: the PCR testing correctly identified the presence of LP1. The issue is that a sample that reflects dead LP1 and no or only minor live LP1 is unlikely to have any health consequences.

Running cultures of samples taken from towers generally distinguish live from dead bacteria. Those samples, according to DOHMH and Stout, take about two weeks before they have grown, been analyzed, and had a final result reported. 

Regardless of whether a PCR sample is positive or negative, DOHMH will also have cultures run, Fowler explained. 

The key action item

In a break with past practice, the agency is requiring building owners to fully remediate any tower that has a positive PCR result, without waiting for the result of the culture. Fully remediate means draining, cleaning, and disinfecting the tower. This compares to the practice in the 2025 outbreak, where a positive PCR result triggered an obligation only to engage in chemical treatment (or “boosting”) with a biocide (bacteria-killing) agent, with full remediation reserved for those towers that ultimately were found to have a positive culture.7

Unlike for cultures, for which local regulations establish different thresholds or concentrations of the bacteria that require different actions,[7] there are no local regulations regarding concentrations found in PCR testing. I asked the DOHMH spokesperson whether there was a PCR threshold that had to be reached for the remediation requirement to kick in. He demurred and noted that the agency scientists who could answer that question were not available.

At a July 6 Town Hall, however, the DOHMH Commissioner, Dr. Alister F. Martin said, “We are going to make sure that we treat anything and everything that could potentially be associated with a positive.”8 That statement certainly implies remediating whenever the PCR testing finds the presence of any LP1, but the proposition remains unconfirmed.

When I asked Fowler why the shift was made, he said it was a function of the Mamdani administration deciding to act aggressively, and that he was not familiar with the various scientific rationales. In fact, those rationales exist (whether or not they in fact were part of the basis for the DOHMH shift in policy). As background, note that local regulations do require full cleaning of cooling towers at least twice a year in the absence of any finding whatever (just as routine maintenance);9 full cleaning is not some newly invented procedure. 

First, a significant percentage of the PCR-positive towers will also come back positive on culture. Stout estimated 50 percent as a ballpark. The data specific to the 2015 Bronx outbreak show that two-thirds of all PCR positives were also culture-positive.10 In this category, the advantage is expediting a protocol that needs to be carried out anyway.

Second, bacteria can “hide” in the biofilm (colloquial translation: slimy layer) that exists in the basin and other water-contact parts inside a cooling tower (like the sump and spray nozzles). One study (in the context of potable water pipes) found that, after a “shock” level of chemical treatment only, the bacteria-containing biofilms were able to “survive this treatment and to continue to grow, ultimately exceeding [one million colony-forming units (CFUs)] per disc.” Thus, a follow-on sample taken on a cooling tower that had only received chemical treatment might come back negative, even though Legionnaires’ bacteria were continuing to grow in the biofilm.

I asked Fowler for a snapshot of the number of towers that have tested PCR-positive and negative, as well as those with PCR results pending. I also asked for the number of Commissioner’s Orders for remediation that have been issued to date. Fowler declined to provide that data (and declined to explain the agency’s reasoning) but said flatly that the agency intends to release the numbers and identify buildings with PCR-positive towers shortly, either by the end of this week or the early part of next week.

The importance of transparency

The first three of CDC’s principles of crisis and emergency risk communication,11 last updated in 2018, say:

Be First:
Crises are time-sensitive. Communicating information quickly is crucial. For members of the public, the first source of information often becomes the preferred source.

Be Right:
Accuracy establishes credibility. Information can include what is known, what is not known, and what is being done to fill in the gaps.

Be Credible:
Honesty and truthfulness should not be compromised during crises.

I spoke with Matthew Seeger, a Distinguished University Professor in the Communications Department at Wayne State University in Detroit and a co-director of the Center for Emerging and Infectious Diseases there. He specializes in crisis and risk communication, including in the public health context. He sees those principles as among the “ethical obligations or fiduciary obligations” of public health officials.

In the case of New York City’s current Legionnaires’ outbreak, DOHMH has exhibited or promised substantial transparency. It has made known its strategy, the timeframe for collection of samples, the number of cases, hospitalizations, and deaths (updated daily), and has promised to release shortly the results of all PCR testing. It has also engaged in outreach to the public through a variety of channels, including broadcast and digital news outlets, social media, and two town halls (the July 6 one, held virtually, had, according to Fowler, a peak number of viewers in excess of 750; an in-person town hall on July 7 had about 200 attendees). 

Relevant to the unwillingness of DOHMH to provide PCR results before the results of all towers are available is Seeger’s view that it is appropriate to release still evolving information (like partial results) to help the public understand what is going on. At the same time, it is “important for the public health agency not to make overly reassuring statements, to not say, ‘Everything’s okay, we have all the answers,’ but to reflect the level of uncertainty that exists.”

Seeger said – not referring to DOHMH, of which he has a high opinion – that the last few years have seen a level of self-censorship from public health authorities worried about backlash from the segments of the public.

In my interview with Fowler, I pointed out that the July 6 town hall, DOHMH’s July 7 press release about action steps, and the agency’s Legionnaires’ webpage are all characterized by only limited discussion of what the course of Legionnaires’ is apt to look like (certainly, nothing like this article’s “Legionnaires’ disease can indeed be a nightmare” can be found). 

Fowler acknowledged that including more information about what Legionnaires’ pneumonia does could be helpful in getting people to understand the gravity of the situation.

I asked Stout in a follow-up email whether DOHMH should be more direct or explicit in explaining that two things are true at the same time: (1) the overwhelming percentage of people in the affected area won’t contract Legionnaires’; but (2) those who do are likely to be hospitalized, at least if diagnosis and treatment is at all delayed. She agreed that “both messages are important to convey.”

This article will be updated as further information becomes available.