Antibody Testing Picks Up More H5N1 Cases in Dairy Farm Workers

Some 7% of workers on Colorado and Michigan dairy farms with H5N1 bird flu-infected cows were seropositive for the virus, according to findings that prompted CDC to revise its guidance for protecting workers.

Eight of 115 workers (45 in Michigan, 70 in Colorado) were confirmed on two separate antibody tests to be seropositive for the 2.3.4.4b clade of H5N1, which is currently circulating among dairy cows, Alexandra Mellis, PhD, of CDC’s National Center for Immunization and Respiratory Diseases (NCIRD), and colleagues reported in Morbidity & Mortality Weekly Report (MMWR).

Among the eight workers who tested positive, four recalled having symptoms, mainly conjunctivitis. The other four did not recall having symptoms, Demetre Daskalakis, MD, MPH, director of NCIRD, said during a press briefing.

A major limitation, Daskalakis said, is that the surveys were conducted an average of 49 days after the first exposure to infected cows, and some were interviewed up to 90 days later.

“That makes recall of minor symptoms potentially difficult,” Daskalakis said. “Nonetheless, data help us better understand the potential for infection with mild or no symptoms.”

All eight workers reported milking cows or cleaning the milking parlor, “supporting these as higher risk activities,” Daskalakis said. None of the workers wore respiratory protection, and only three wore eye protection, “highlighting the need for better tools to support worker protection,” he said.

Nirav Shah, MD, JD, principal deputy director of the CDC, said during the press briefing that the findings led the agency to update guidance for protecting workers. He described it as an “intensification” of the recommendations “to focus on the highest exposure tasks in poultry and dairy operations, ultimately to reduce the risk of infection.”

First, the agency expanded the recommendation for who should be tested to include people who were exposed but asymptomatic. “We in public health need to cast a wider net in terms of who is offered a test so that we can identify, treat, and isolate those individuals,” Shah said.

Previously, testing guidance was focused on those who were exposed and had symptoms.

“This active case finding reduces the likelihood that a mild infection could turn into a severe infection, or that the infection spreads to anyone else,” Shah said. “Simply put, the less room we give this virus to run, the fewer chances it has to cause harm or to change.”

Next, the agency expanded oseltamivir (Tamiflu) access to those who had a high-risk exposure to the virus but remain symptom-free. Such an exposure could include “an unprotected splash in the face with raw cow milk, or in the poultry setting, something that might have happened during a depopulation or culling event where the individual was not wearing appropriate PPE [personal protective equipment],” Shah said.

Giving the antiviral sooner “reduces the likelihood of an asymptomatic case being symptomatic … and lowers the risk and the chances of onward transmission to close contacts,” he added.

Finally, new recommendations prioritize PPE based on workplace exposure, focusing on farm tasks that present the highest risk, such as culling efforts or working in the milking parlor with sick or infected animals.

A separate study published simultaneously in MMWR showed that use of PPE increased on Colorado farms once H5N1 was detected. Dairy workers reported a 28% increase in PPE use after detection, which included a 40% jump in use of eye protection during milking. Use of respirators and other masks remained low, however.

Overall, there have been 46 human cases of H5 influenza in the U.S. in 2024, Daskalakis said. Of these, 25 were related to the ongoing outbreak in dairy cows, 21 of which occurred in California. A total of 20 cases were tied to poultry depopulation efforts, including 11 recent cases from Washington state, he said.

There was also one case in Missouri with no known animal or animal product exposure. In that case, a household contact also tested positive for H5N1 infection on serology but was not included in the final case count (nor will the Colorado and Michigan workers who tested positive on serology be included in final counts, Daskalakis said).

When asked why serology testing took so long — workers were sampled and surveyed from June to August of this year — Daskalakis said it was a relatively standard timeline for this type of work, but researchers did have to develop additional protocols when it became clear that cross-reactivity with seasonal influenza needed to be examined.

Indeed, the study assessed seroprevalence for H1N1 seasonal flu, which ended up at 66%, suggesting previous vaccination or infection with seasonal flu among these workers.

He said the seroprevalence findings are in line with other seroprevalence studies investigating this clade of H5N1, including one conducted among Egyptian bird market workers in 2022-2023 that found a seropositivity prevalence of 4.6%, and a U.S. preprint from earlier this year that found about 14% of workers on two Texas dairy farms with infected herds had an asymptomatic infection.

Daskalakis emphasized that there’s been no evidence that the virus has changed to make it better able to infect humans and that the CDC’s assessment of the risk to the public health from H5N1 currently remains low.