Unannounced drills reveal gaps in healthcare facilities' infectious disease response

When trained actors walked into 69 healthcare facilities across New Jersey, New York, and the U.S. Virgin Islands without warning, coughing and reporting fever, they were meant to test something no protocol on paper can measure: what actually happens when a potentially infectious patient arrives during an ordinary shift.

The results, published in the Morbidity and Mortality Weekly Report, showed that American healthcare facilities are not ready. Between January and June, the New York University Standardized Patient Program sent 73 patient actors into emergency departments, urgent care centers, and outpatient clinics, each one presenting with avian flu-like symptoms and a history of contact with sick or dead birds. The actors were young—ages 20 to 28—and their stories were consistent. What varied wildly was how the facilities responded.

In just over 60 percent of the drills, patient actors were both masked and isolated. That sounds acceptable until you look at the timing. The target for masking is one minute. Only 43 percent of facilities hit it. The target for isolation is ten minutes. Only 48 percent achieved that. When a patient actor was masked, the median time was two minutes—not catastrophic, but a full minute beyond the standard. When isolation occurred, it took a median of eleven minutes, one minute past the goal. These are not large margins, but in infectious disease response, they matter.

The gaps revealed themselves in different ways across different settings. Nearly every facility—93 percent—screened patients for symptoms, and most clinicians took travel histories. But only 9.6 percent of facilities asked about exposure to birds specifically, the detail that would have flagged avian flu risk most clearly. When clinicians did recognize the risk, they notified infection prevention and control staff in just over half the cases. Emergency departments isolated patients in 57 percent of drills. Urgent care centers did so in 100 percent. The inconsistency itself is the problem.

Masks tell another story. Eighty percent of facilities had masks available in waiting areas, but that availability dropped to 58 percent in outpatient clinics and rose to 95 percent in emergency departments. In 18 percent of facilities, a patient actor who was visibly coughing and reported fever was never given a mask. In 19 percent of drills, clinicians did not wear a mask or respirator during clinical evaluation of a potentially infectious patient.

Nang Thu Thu Kyaw, who led the study from the New York City Department of Health and Mental Hygiene, noted that unannounced drills capture something that tabletop exercises and written protocols cannot: the real behavior of real staff under real conditions. "Sending trained patient actors into healthcare facilities without warning captures how frontline teams—including non-clinical staff—respond in real time," she said. The study found that receptionists, registrars, greeters, and security personnel performed symptom screening in more than half the drills and handed out masks in nearly two-thirds of them. These are the people patients encounter first, yet they are often not included in infection control training.

The researchers identified a structural problem: urgent care centers and outpatient clinics are common sites where acutely ill patients seek care, yet they operate with less robust screening and isolation infrastructure than emergency departments. Kyaw and colleagues called for all these settings to establish standardized protocols, keep staff updated on current threats, and use decision-support tools that prompt clinicians to ask the right questions. The New York City Department of Health and Mental Hygiene now offers a mystery patient drill toolkit online for other facilities to use.

The study had limits. The number of drills at outpatient and urgent care centers was relatively small, which may mean the findings do not fully represent how those settings perform nationally. Patient actors could not exhibit all avian flu symptoms—they could not actually run a fever or develop conjunctivitis—which may have made clinicians less likely to consider avian flu in their differential diagnoses. Still, the core finding stands: when an infectious disease threat arrives unannounced, most American healthcare facilities are not prepared to respond within the time windows that matter.