Health officials in New York staged an unannounced outbreak drill this year and came away with a failing grade: only 60% of the 73 healthcare facilities visited correctly masked and isolated a walking 'patient zero'.

The exercise, published in the CDC's Morbidity and Mortality Weekly Report and covered by Ars Technica, tested the first local steps of an outbreak with pandemic potential. Professional patient actors staggered into emergency departments, hospital outpatient clinics and urgent-care centres wearing makeup to look unwell, with eye drops and petroleum jelly standing in for conjunctivitis and a fake cough completing the picture. The scenario was H5 bird flu: they complained of weakness and fever, mentioned taking an over-the-counter fever reducer before arriving, denied foreign travel and — if asked about sick contacts — would answer 'Does a sick bird count?' before telling a sad tale about trying to rescue a sick duck found in a park, handling it without gloves. The duck did not make it. The actor told the story unprompted.

To pass, a facility had to mask the patient and move them to an isolation room; timing was recorded but not required. The results, from drills run between January and June 2026 across New York City (52 sites), the rest of New York state (15), New Jersey (3) and the US Virgin Islands (3):

- 60% (44 of 73) both masked and isolated the patient. - Emergency departments performed worst: 55% passed, against 68% of outpatient clinics and 67% of urgent-care centres. - 18% never offered a mask to a coughing patient who said they had a fever. - Of those that did both, only 43% masked the patient within the one-minute goal and 48% isolated within the ten-minute goal. - Median time to mask was 2 minutes; median time to isolation was 11 minutes, and at least one outpatient clinic took nearly an hour. - 93% screened the actor for symptoms, but staff at only 7 of 73 sites asked anything specific to bird-flu exposure — despite the dead-duck story. - Personal protective equipment use was inconsistent: 25% of clinicians wore all four recommended items (gloves, mask, gown, eye protection); 6% wore none, and 19% skipped a mask or respirator. - Only 55% notified or planned to notify internal infection prevention and control staff as recommended.

The authors' conclusion is that healthcare workers and facility staff need more training to achieve earlier identification, masking and isolation of potentially infectious febrile patients, and they single out waiting areas as high-risk settings for respiratory transmission when infection-control measures are skipped.

Why it matters: the study is a reminder that pandemic preparedness is tested at the front desk, not in a national stockpile. The authors also caution that the results probably flatter the system — facilities that declined to take part, and may have scored worse, were never measured, and some sites may have had advance warning that a drill was coming.