What We Know About New York's Patient-Zero Pandemic Drill

A regional test of 73 health care facilities in New York, New Jersey, and the US Virgin Islands found that 40 percent failed to properly mask and isolate a patient presenting with symptoms of a potential pandemic virus, according to a study published in the Morbidity and Mortality Weekly Report (MMWR) and reported by Ars Technica.

What did the drill test, and who ran it?

Health officials in New York organized the exercise to measure how quickly front-line facilities could spot and contain a patient carrying a virus with pandemic potential, according to the MMWR report cited by Ars Technica. The drill, conducted between January and June 2026, sent professional patient actors into real clinical settings without advance warning of the exact date, though the facilities had agreed in advance to take part.

The scenario simulated H5 bird flu, a virus type public health officials consider to carry pandemic risk, Ars Technica reported. The test measured a single, binary outcome at each site: did staff mask the patient and move them to isolation.

How many facilities failed, and where?

Of the 73 facilities visited, 60 percent — 44 sites — both masked and isolated the actor, meaning 40 percent did not, per the MMWR data relayed by Ars Technica. That overall figure masks wide variation by setting. Emergency departments, which received the most visits (42 of 73), passed at roughly 55 percent. Hospital outpatient clinics, visited 19 times, passed at about 68 percent. Urgent care centers, visited 12 times, passed at about 67 percent — meaning emergency rooms, the facilities most likely to see a walk-in patient with an emerging infection, had the weakest results of the three settings tested.

Geographically, the drill concentrated on New York City, which hosted 52 of the 73 visits. The remaining sites were split among New York State outside the city (15), New Jersey (3), and the US Virgin Islands (3), which the federal government classifies within the same health region as New York, according to Ars Technica's account of the report.

What symptoms and story did the actors use?

To mimic an H5 bird flu case, the actors wore makeup to appear unwell, used eye drops and petroleum jelly to simulate conjunctivitis, and faked a cough, Ars Technica reported. They told clinical staff they felt weak and feverish and said they had taken Motrin before arriving.

When asked about travel, the actors said they had not left the country. When asked about sick contacts, the script called for the actor to ask, "Does a sick bird count?" and then volunteer, without prompting, an account of finding and trying to save an ailing duck in a park — a duck that, in the story, did not survive — while handling it without gloves, according to the MMWR details relayed by Ars Technica.

What counted as a pass or fail?

The pass threshold was deliberately simple: masking the patient and placing them in an isolation room, Ars Technica reported. Officials had an ideal timing target — masking within one minute of arrival and isolation within 10 minutes — but for scoring purposes in this drill, any masking and any isolation counted as a pass regardless of how long it took.

Even against that lenient bar, 18 percent of facilities never offered the coughing, feverish actor a mask at all, according to the report cited by Ars Technica. The source material does not specify the isolation-only failure rate or the breakdown of how many facilities met the faster, ideal timing benchmarks.

What do health officials say needs to change?

The officials who conducted the drill wrote that the results point to specific training gaps rather than systemic resource shortfalls, according to the MMWR language quoted by Ars Technica: "These drills identified needs for health care workers and health care facility staff member training to achieve earlier identification, masking, and isolation of potentially infectious febrile patients."

The report frames the exercise as a local readiness check rather than a national one — a single health region testing whether its own emergency departments, clinics, and urgent care centers can execute the most basic containment step before a suspected pandemic pathogen moves further into a building. Ars Technica's account does not include details on what corrective training, if any, has since been implemented at the facilities that failed, nor does it name the specific hospital systems tested.