Why is human HPAI influenza screening so difficult, when the signals generated are potentially so critical for poultry health??

Maybe if we don't look for asymptomatic human infection, it won't TRANSMIT??

I’ve been off-line in California for about a week and just scanned a couple of pieces released last week that I want to tie together with some other thoughts on influenza and early zoonotic and reportable disease detection in general bouncing around in my head right now.

Let’s turn first to an NPR Report out earlier last week:

Bird flu surges among poultry amid a scaled back federal response : Shots - Health News : NPR

…“Absolutely we’re going to see human infections, just like we did a year ago,” says Seema Lakdawala, a virologist at Emory University who studies bird flu. “Nothing has changed substantially to suggest otherwise.”

And though no human cases have been reported since early this year, scientists involved in tracking the virus say they don’t have good visibility into what’s happening in animals or humans. That’s because the federal government has scaled back on surveillance and communication.

“We’re not in a great position for monitoring things,” says Wendy Puryear, a virologist at Tufts University. “I’ve been deep in the weeds since it arrived here in the U.S., and I’m finding myself in a very uncomfortable place.”

Scaled back federal response

Puryear says a network of influenza researchers used to be in constant contact with their counterparts at federal health agencies about H5N1, mapping out research priorities and discussing changes in the virus.

But now, she says, “much of that infrastructure has been either completely closed down or significantly hampered,” under the Trump administration, due to cutbacks in staffing, early retirements and other policies.

In fact, Puryear recalls a recent meeting that consisted entirely of her colleagues announcing they’d had to halt their studies related to H5N1 in humans because of the administration’s funding cuts.

“It was a pretty depressing meeting,” she says. “Just one person after another was closing their doors.”

Another pressing issue: Data coming from the federally-run lab that does genetic sequencing of virus collected from infected animals has slowed to a trickle, something that was true even before the government shutdown, says Dr. Keith Poulsen, who runs the Wisconsin Veterinary Diagnostic Laboratory.

He says loss of staff at the U.S. Department of Agriculture — and the fact that communication with scientists needs to be “pre-approved” — has created obstacles, at a time when the threat of spillover from wildlife is increasing.

The shutdown has only exacerbated the situation.

For example, the network of veterinary labs that’s responsible for testing ceased its regular meetings because of the shutdown, Poulsen says.

“The shutdown just kind of pours gas on a smoldering fire and makes us vulnerable,” he says. “It’s a national security issue.”

None of this is surprising to any of us who have been closely watching for updates and new information from our USDA and CDC partners. We continue to get H5N1 poultry outbreak counts, while the true dairy herd situation gets more and more mysterious with decreased bulk tank testing allowed with negative results reported under the National Milk Testing Strategy (NMTS). This drop in testing frequency is concurrent with a drop in within herd prevalence and intermittent infection and shedding, making negative bulk tank results from lowly positive herds more likely. With non-poultry H5N1 testing limited to occasional milk PCR sampling, we really have little idea of H5N1 viral infection incidence in mammalian livestock species.

A second piece I read was a research review and accompanying commentary from the Journal of the American Medical Association Network, which covered the potential for asymptomatic human H5N1 infections and related human to human transmission potential, via a review of existing literature by Dawood et al:

Asymptomatic Human Infections With Avian Influenza A(H5N1) Virus Confirmed by Molecular and Serologic Testing

Findings Of 1567 unique reports that underwent title or abstract screening, 42 were selected for full-text screening, of which 10 met inclusion criteria (3 reports about 2 MSC (molecularly and serologically confirmed) cases and 7 reports about 16 MC (molecularly confirmed) cases). The 2 MSC cases occurred among adults in Pakistan and Vietnam who were identified by investigations of household contacts of index A(H5N1) case patients; 1 case patient also had exposure to A(H5N1) virus–infected chickens as the possible infection source and 1 is thought to have been infected through human-to-human transmission. Neither MSC case patient used personal protective equipment. Of 16 reported MC cases (14 adults, 2 children), 11 were identified by enhanced surveillance of persons exposed to A(H5N1) virus–infected poultry (8 in Bangladesh, 2 in Spain, and 1 in the UK) and the remaining 5 MC cases (3 in Vietnam, 2 in Cambodia) were identified by investigations of household contacts of index A(H5N1) case patients.

Conclusions and Relevance Asymptomatic human infections with A(H5N1) virus have been infrequently reported, with most cases identified through enhanced surveillance or household contact investigations of persons with known exposure. Robust data collection is needed from persons with possible asymptomatic A(H5N1) virus infection to inform future public health responses.

This is not a robust dataset; moreover, it is based on earlier H5N1 genotypes prior to the 2024 U.S. dairy outbreak. However, in limited cases, asymptomatic infection was confirmed, as was likely human to human spread of infection.

A more detailed review of the data and other work supporting both asymptomatic human infection and human to human spread of H5N1 is found in the accompanying invited commentary in JAMA Open by Rick Bright and Nicole Lurie:

Asymptomatic Influenza A(H5N1) Infections and Sustained Surveillance—Sustaining Surveillance Beyond the Crisis | Infectious Diseases | JAMA Network Open | JAMA Network

…Human cases have been confirmed in multiple states, and as of mid-2025, approximately 70 cases have been officially reported, most with mild illness but including at least 1 death. The report from Dawood and colleagues thus challenges the notion that there has been no human-to-human transmission related to these events. It also highlights a critical gap in many national and global surveillance systems, which tend to focus on the detection of symptomatic illness rather than infection, and underscores the importance of carefully examining clusters where limited transmission may occur. The implications are substantial. If asymptomatic infections are occurring, transmission chains can go undetected, giving the virus opportunities to adapt and spread widely before an emerging pandemic is recognized. These silent events will not be captured by systems that rely primarily on symptom-based case finding. Dawood et al provide a strong evidence base for expanding the scope of influenza surveillance to include molecular and serologic testing of individuals at higher risk of exposure, whether or not they are experiencing symptoms, as well as testing of contacts of infected persons. This is particularly important for pathogens with pandemic potential, where early detection is essential for containing spread before it becomes sustained. With regard to the dairy cattle–associated outbreak, the cases reviewed by Dawood et al suggest that cases could be significantly underestimated. Several investigations have detected antibodies to influenza A(H5N1) among veterinarians and dairy workers who had contact with infected cattle. In some studies, up to 7% of those tested had serologic evidence of infection, and about half of those individuals reported no symptoms. In contrast to the cases in the report by Dawood et al these cases now arise from within the US. They point to a gap between what surveillance systems capture and the true scope of infection and highlight the need to strengthen, not curtail, surveillance and reporting.

While there’s plenty of evidence that increased surveillance for H5N1 would be prudent, a blog from Michael Coston of Avian Flu Diary yesterday brought home the message regarding the public’s lack of concern for the H5N1 threat to people:

Avian Flu Diary: Preprint: Biosecurity Uptake and Perceived Risk of Avian Influenza Among People in Contact with Birds

The findings of the British study cited in the blog state the following:

Our findings suggest that use of BMs (biosecurity measures) by people who have contact with birds in the UK are influenced by the number of birds a person is in contact with (as a proxy for farming operation size) and the perceived risk to bird health (high) but not perceived risk to human health (low).

Further work should be undertaken to identify the characteristics which lead to low BM uptake among different cohorts in contact with different bird species to help inform zoonotic AIV public health messaging and campaigns. Moreover, a more extensive investigation into the barriers and facilitators of BM uptake is vital for improving our understanding of nuanced behavior in the context of zoonotic influenza spillover.

As Michael states in concluding his piece:

While the strains of HPAI H5 currently circulating in the UK, the U.S. and in Europe have thus far proven less risky to human health than those found in Asia, there are no guarantees how long that will remain the case.

Finding better ways to communicate the risks to the public won’t be easy, but this is one virus we really don’t want to learn to respect the hard way.

Despite all the admonitions to respect the zoonotic potential for H5N1 transfer from poultry (or cattle) to humans, I really don’t think the human risk will gain much traction unless or until we have multiple hospitalizations or fatalities from a mutated virus. The ag industry has been habituated to ignore zoonotic risks from years of “nothing has happened yet”.

However, zoonoses pass both ways! What if asymptomatic or subclinical human infections are more common than assumed in people that move between infected and uninfected populations of animals? We may shower in and out and use biosecurity protocols. What about nasal passages of all possible cross contacts, not just the “symptomatic” ones? At the swine influenza session at the Leman Conference, “asymptomatic” or “preclinical” human influenza in people was shown to be a risk factor in passing zoonotic influenza on to negative nursery pig populations (see my last column (Swine Influenza - More Complicated Than We Ever Imagined). It’s not an unreasonable leap to speculate that “asymptomatic” or “preclinical” human H5N1 influenza infections could pass H5N1 2.3.4.4b on to naive poultry or dairy cattle populations.

We easily default to wildlife, fomites, “poor biosecurity”, etc. in explaining new flock or herd breaks. Have we truly eliminated possible human spread of virus by asymptomatic carriers as a risk factor? How much (if any) swabbing of asymptomatic humans are we doing to determine if there could be potential carriers?

Indiana as a Current Hot Spot

I pulled up the following information from the current outbreak in Indiana only as a current HPAI hotspot, NOT to pick on its Board of Animal Health (BOAH), one of the finest state animal health units in the U.S. I only use this example as an area where a large outbreak in 24 concentrated flocks in a small 2-county area in a 3-week period provides an opportunity for some detailed epidemiological investigative work, which is likely underway right now.

This is a tragedy for the 24 producers currently involved, with more close neighbors likely under imminent threat. Now it’s possible that each outbreak was an independent spillover from wildlife. The standard “talking point” is that HPAI is a spillover from wildlife (birds) to poultry flocks; a wildlife spillover does start most or all outbreak cascades. However, once initiated, it’s likely that some subset of these may be spread from farm to farm via pre-clinical spread of infection. Phylogenetic analyses can likely provide some decent evidence of the relatedness of the various farm outbreak strains.

With this much virus floating around in poultry flocks in the 2 counties in this example, all poultry farms in the area suffer from multiple imminent risks for infection, given their proximity and common contact sources (feed, management, suppliers, service staff, etc.) within the immediate area. Every susceptible livestock farm in the area should be testing healthy birds AND people in direct contact with birds daily for molecular evidence of H5N1. If the virus is replicating either in “healthy” (preclinical) birds or in people, animal health and public health both need to be aware of it before the clinical outbreak in the affected flock. Additionally, any viruses isolated need to be rapidly sequenced and shared for pandemic preparedness. none of this testing is currently allowed under existing HPAI testing protocols.

(Note- currently testing is only authorized on birds showing early signs of clinical illness, or in flocks in outbreak or surveillance zones. People are not tested, unless showing clinical signs and associated with an infected premise. All testing is done with the official H5 PCR test at a NAHLN lab, except human testing is handled by state public health guidance)

I’d argue that it IS the poultry/livestock industry’s concern whether people are asymptomatically infected with H5N1 or not, because people’s noses are as much a viable a fomite for spread of virus as a truck tire or a non-disinfected boot! Evidence continues to accumulate (as described above) that we cannot ignore human carriage of zoonotic viruses and still maintain responsible stewardship for animal/bird health.

If we test thousands of asymptomatic noses over time in areas like northern Indiana (or ND/SD/MN last month) with 100% negative results, maybe we can begin to discount this as a valid risk factor; however, right now we are leaving a giant hole in our biosecurity assessments by ignoring H5N1-exposed afebrile people as potential carriers in naive bird or dairy populations. We’ve already documented that preclinically H1/H3-infected people infect naive swine populations!

Here is Indiana’s official guidance, found within Indiana HPAI Situation Update #65 regarding human infections and testing guidance, which in general reflects USDA-CDC-National Guidance:

Human Health
Recent detections of highly pathogenic avian influenza A H5N1 viruses (H5N1 bird flu) in wild birds and poultry pose a low risk to the health of the general public. Human infections are most likely to be rare and happen in people with recreational or occupational exposures involving prolonged, unprotected close contact with infected birds or contaminated environments. The Indiana Department of Health (IDOH) and local health departments (LHD) monitor people who have been exposed to test-positive animals for signs of influenza-like illness. To date, none have had evidence of infection or illness associated with AI virus.

IDOH offers the following guidance documents for poultry growers and emergency responders to protect their health:

• Indiana Department of Health Avian Flu Producer Information

I’d make the following points about current state-federal human protection and sampling guidance in general:

1. Respiratory protection (masks) is a “punt”, likely for reasons of comfort and practicality. Workers are not masking when animals are shedding maximal levels of aerosol virus during the (undiagnosed) preclinical infectious period in both dairy cattle herds and poultry flocks.

2. Human testing is not recommended until workers show clinical signs following exposure to known diagnosed animals - “The Indiana Department of Health (IDOH) and local health departments (LHD) monitor people who have been exposed to test-positive animals”. Further, testing is limited to workers exhibiting flu compatible clinical illness.

We probably can’t afford daily 100% medical PCR testing on all contacts at risk of infection; we may have to settle on Matrix PCR testing for the bulk of the people with N95 masking as a secondary barrier? We at least need to begin conversations on effective high-risk non-clinical signs-based testing of both people and bird populations.

Perhaps the first step is an animal health-public health-industry research partnership for daily PCR swabbing at a subset of high-risk “naive” negative poultry farms in extremely high-risk areas. Workers, birds, and possibly environmental (air inlet?) samples could be collected across multiple at-risk farms prior to diagnosed outbreaks. We need to begin looking earlier if we hope to narrow down likely initial sources of infection. It will take a wider net of testing in negative farms to find a subset of farms with preclinical infection for analysis.

I don’t raise this just to be difficult. The public is getting NO feedback from USDA regarding epidemiological analyses of HPAI poultry or dairy outbreaks. The early simplistic results from the dairy outbreak analyses turned out to be incomplete and over-hyped, as shown by subsequent independent on-the-farm work in California. Now all poultry outbreaks are blamed by USDA on lapses in biosecurity in the current narrative. While that may be technically true, not all lapses are preventable through USDA-approved methodology against already known routes of infection. We have a lot yet to learn about outbreaks that good, published and peer-reviewed epidemiological analyses would greatly assist in elucidating. Possible human spillover of H5N1 to birds is one of many possible farm-to-farm modes of spread that should be included in fresh reviews of existing case histories and phylogenetic data. We depend on the states and USDA to accomplish this with partners because they monopolize this data.

Possible human to bird transmission of H5N1 is only one of many current epidemiological unknowns that CAN also be addressed in the field if we choose as curious people to seek out the answers! We have existing politics, processes, policies and protocols standing in the way of gathering that data. I’m suggesting that avoiding robust human H5N1 field diagnostics is no longer smart politics for the animal health industry; it is possibly costing us in not attributing new poultry and possibly dairy herd infections to zoonotic spillover. At a minimum, we need to prove that not to be the case for the health of our livestock. Anything we can learn about zoonotic risk of human spillover viruses from our public health researchers is a bonus!

In the last column I discussed how the swine industry discovered that “One Health” work with H1/H3 provided invaluable lessons of benefit to the swine industry, in part because people are the biggest contributors of new influenza viral genomic material in the swine influenza gene pool. This week I’m suggesting via fairly robust research evidence that people’s noses may be a contributor to H5N1 infection in poultry operations. We at least need to seriously investigate the risks. When in doubt with flu, follow the pigs!

John