Are We on the Cusp of a Major Bird Flu Outbreak? (2025)

A highly pathogenic subtype of H5N1 was first detected in birds in China in 1996, and the first human cases appeared soon afterwards. The virus diversified as it replicated in birds, and migratory bird species spread multiple subtypes around the planet. Avian influenza came to North America in 2014 and caused a severe outbreak in U.S. poultry, but the outbreak was controlled, and luckily the virus disappeared. Up until 2024, there have been almost 1,000 documented human cases of H5N1 in multiple countries, with a fatality rate of about 50 percent. In 2022, a new subtype of H5N1 clade 2.3.4.4b emerged in North America and spread to domestic poultry and many wild mammal species. Over the past year, this subtype of H5N1 has also been spreading in dairy cows.

Overall, then, H5N1 has a long history of transmission through multiple bird and mammal species worldwide, and the virus has undergone massive diversification into multiple genetic subtypes. The jump into dairy cows is new, and we don’t really know what the implications are yet. We’ve been monitoring the spread to humans, and most cases have been mild so far. There’s been no evidence of human-to-human transmission, but there has been human-to-cat spread, which suggests that humans can transmit the virus.

HMNews: How does H5N1 affect humans?

Stephenson: Historically, subtypes of H5N1 were quite severe, which is why everybody has been worried about the virus. Over the last year in North America, we’ve seen a milder and more self-limited presentation. In more than 90 percent of cases, patients develop conjunctivitis, or pink eye, which is not uncommon in influenza. Since 2022, there have been 70 human cases, and hospitalization has been very rare. Cases have been a mix of the two subtypes also infecting animals. In the last couple months, we have seen three patients with severe or critical illness and one death.

H5N1 testing is recommended for anyone with influenza-like illness and a high-risk exposure. We’re also doing some testing in the general population. Currently, we recommend that all patients hospitalized with influenza A get further testing for H5N1. The hope is that we will pick up cases we are missing in other ways.

HMNews: What is the status of treatments and vaccines for H5N1?

Stephenson: We recommend giving the antiviral drug oseltamivir, or Tamiflu, to anyone with a suspected or confirmed case or an exposure. Until recently, H5N1 strains did not show resistance to Tamiflu. However, a study from Canada documents the emergence of a resistant H5N1 strain in poultry. This strain has a rare mutation that has not yet been observed in humans. Still, the potential for resistance is why it’s so important to develop antivirals and monoclonal antibodies that could give us new treatments. We also need to continue genomic surveillance of H5N1 to understand the spread of resistance mutations.

The United States has stockpiled three licensed, pre-pandemic H5N1 vaccines. All three are safe and elicit antibodies against influenza. The problem is that these vaccines do not match the current strains. Research is mixed on how well they neutralize current H5N1 strains. It’s important to have these stockpiled vaccines, but we also need updated vaccines. Traditional vaccine companies are working on new H5N1 vaccines, but it’s going to take time to make them at scale. As a result, we’re looking at new technology, including mRNA, which was so successful for COVID-19 vaccines, and will be faster to update and manufacture.

I think we are living next to a volcano, and it may erupt or it may not. But we need to prepare for the possibility of a pandemic.

Jacob Lemieux

HMS assistant professor of medicine and infectious disease specialist at Mass General

Our ability to develop and test vaccines will require continued and long-term investment in our national clinical trial infrastructure. Nearly all of the COVID-19 vaccines were tested in National Institutes of Health-funded clinical trial groups around the country that had been building up for decades. Support for clinical research is not just funding for individual investigators but also funding for all aspects of clinical trials, including institutional review boards and clinical research centers. When we don’t know what our funding is going to be, we can’t prepare for the future — and since we’re working on pandemic preparedness, it paralyzes our work.

HMNews: What is happening in terms of surveillance in the United States?

Lemieux: Most human surveillance is being carried out by the Centers of Disease Control and Prevention, in coordination with state and local public health laboratories. There are two major surveillance programs: U.S. Influenza Surveillance, which tests and characterizes influenza specimens, and targeted surveillance of people after exposure to infected animals. The CDC is also doing some serological testing of dairy workers.

There are a few factors about H5N1 that point to an increasingly risky and uncertain situation. Increased activity has been sustained for over a year, so the virus does not seem to be going away. The infection is spreading across domestic and wild animals and spilling over into humans. Human cases have escalated since last year, when the first outbreak occurred in cows. Symptoms in humans are relatively mild and common, making H5N1 hard to identify. Fortunately, we haven’t yet seen evidence of human-to-human transmission, but there is a need for urgent, escalated, and coordinated national, state, and municipal surveillance across sectors.

Goldstein: Here in Massachusetts, surveillance is an ongoing collaboration with our agricultural partners. Last summer, the Massachusetts Department of Agriculture Resources and Department of Public Health worked together with MassCPR partners to institute dairy farm testing across the state. We were the first state to do so, and we remain the only state to do it at scale. We are testing all farms in Massachusetts monthly for H5N1, and to date, no dairy farm has tested positive. We need continued collaboration with agriculture, public health, and academia to make sure that our surveillance network is as broad as possible.

HMNews: How can we improve surveillance?

Lemieux: In humans, I think we need to look more closely at respiratory syndromes linked to conjunctivitis. We also need to develop better diagnostics for subtyping influenza viruses. Finally, we need resources to support collaborations that enhance surveillance. With pauses in federal funding, there’s a lot of uncertainty right now about whether surveillance efforts can continue, but investment in surveillance is necessary to keep us safe.

Runstadler: On the animal side, we’re experiencing similar uncertainty in surveillance. In the Northeast, we have great partners in the Department of Public Health, MassWildlife, and many other organizations. That’s not the case in other parts of the country or the world, which makes it very difficult to track the virus. One area of concern is infections in marine mammals. In South America, thousands of seals and sea lions perished after being infected by H5N1. In the Northeast, we recorded a seal outbreak in 2022 and a few infections in 2023 and 2024. Spring is when we do most of our seal sampling, but one of our big partners has been hamstrung by federal funding cuts, so our sampling has gone way down this year.

The reality is that we can’t prepare for what we don’t know is there. A critical aspect of surveillance is understanding what is circulating and what is a potential threat for spreading to other species. H5N1 is causing a lot of death in marine mammals, but we don’t understand the basic science well enough to know whether something analogous could happen in humans. We need to study the dynamics of the virus in marine mammals to learn more about it — but we can’t do that if we don’t have funding and support for surveillance.

HMNews: On a scale from one to ten, how worried are you about H5N1 becoming a major pandemic?

Goldstein: As infectious disease doctors and public health professionals, we worry all the time, but I think we’re in the four to five range right now. I do think we need to have a healthy worry about H5N1, but the risk to humans at this moment remains low.

Luban: I would increase those numbers a little bit. Over the past year I’ve polled influenza experts, and there’s an enormous range of responses. Some said that H5N1 has been circulating since the 1990s, and so they don’t think a pandemic is going to happen. Others are extremely concerned about the fact that the virus has changed a lot and is spreading to new species, which allows it to acquire new properties.

Stephenson: I’m a vaccine person, so that's the lens I use to think about things. I’d put it at a five. We do have stockpiled vaccines that will probably help, but we’re in a very bad funding environment, and I’m not sure we have the infrastructure to actually test vaccines.

Runstadler: I’m probably at a six, but I’ve come up from a two or three in a fairly short time period. I’m also at a nine or ten for an influenza virus becoming a pandemic at some point in the future.

Lemieux: I would put it at a six or a seven. In infectious diseases, we have a saying that resistance is a function of time and titer, meaning that the ability for a pathogen to evolve depends on time under pressure and size of the reservoir. The virus has been with us for several years in multiple species, and the reservoir is large. I think we are living next to a volcano, and it may erupt or it may not. But we need to prepare for the possibility of a pandemic.

This conversation was edited for length and clarity. The story was updated on Monday, March 10.