H5N1, a subtype of avian flu, has been a major issue for American farmers since 2020. But it’s increasingly becoming a problem for people, too. Initially confined to poultry, the virus has spread to several species over the years, including foxes, minks, and alpacas, and, earlier this year, cows. This year, a large-scale infection in cows led to a spate of human cases. Most alarmingly, three recent infections in people — including one that left a teenager in British Columbia in critical condition — can’t be traced to contact with any animals, indicating the possibility of human-to-human transmission. There is still no evidence such transmission is taking place, and authorities say risk to the public is low. But this problem clearly isn’t going away, and the U.S. response to it has been sluggish and opaque. David O’Connor, a well-known professor of pathology and laboratory medicine at the University of Wisconsin, has been a notable critic of the government’s approach. I spoke with him and his colleague Tom Friedrich, a virology professor at Wisconsin, about how COVID fatigue, fragmented health-care oversight, and mistrust of the medical Establishment has combined to create a high-risk situation.
Professor O’Connor, you had a memorable quote in a recent New York Times op-ed by Tulio de Oliveira: “It seems that the United States is addicted to gambling with H5N1, but if you gamble long enough, the virus may hit a jackpot.” Can you explain exactly how we’re gambling here? What are we doing that’s so risky with regards to this virus?
David O’Connor: It’s been eight months since we first recognized that there were infections in dairy cattle. And while there has been a government response and a lot of conversations about it, the response has been what I would call too little and too late. It simply isn’t matching the scope of the potential risk, and we are gambling by not trying to take that more seriously. If you think about the dairy-cattle outbreak in particular, you have a bit of a gap. The USDA, CDC, and FDA can all credibly claim that they should be leading the response, but it doesn’t seem like any one organization has particularly taken up the mantle as the one in charge. USDA regulates agriculture; CDC is concerned about the spillover risk to human health; the FDA regulates milk. And they’ve all been involved, but I think you’d be hard-pressed to find one organization that would say they are running lead on this — that they are the H5 influenza czar overseeing and coordinating all aspects of the response.
When everyone’s in charge, no one’s in charge.
David O’Connor: So what you have are these policies by half-measure. Back in May, the USDA asked farms to voluntarily provide milk for bulk testing, but there was no particular incentive for them to do that, so very little bulk testing of raw milk was done. It took until November for the USDA to announce a rollout of bulk-tank testing, and that still appears to be a slow process. A couple of states took it upon themselves to do bulk-tank testing. And guess what? They found a bunch of previously unknown dairy herds that were affected. Colorado, California, and Massachusetts are the three I know.
So some states have done it, but that’s three out of 50. Where has the leadership been in terms of getting the bulk-tank testing so that we would simply know how widespread this virus is in dairy herds? The FDA, with some scientists, found that H5 genetic fragments could be found in pasteurized milk. So if you buy milk at the grocery store, you’d find that some of it has H5 genetic fragments. Milk is totally safe to drink — the virus is inactivated. But here we have the ability to understand how much H5 is in our dairy supply. And critically, we could use genetic epidemiology tools to understand the sequence of the virus. So how much H5 is out there? Where is it? What does it look like, how is it changing?
When the FDA has done surveys of these products, they tout the fact that they tested 167 products. Is that a lot of products to test, given the scale of the dairy industry around the United States? I would argue that they should be testing ten times, 50 times that number if we really want to understand how this virus is potentially moving through time and space. And for the CDC’s part, they recently released a study that found that among 115 dairy workers, 7 percent had antibodies to H5N1. That’s almost 10 percent — admittedly of a small number of people they reported on — who have apparently been exposed to this virus occupationally in the last year. So we have to be asking: Why have they only looked at 115 people? Why did it take a peer-reviewed scientific article for that information to come out? Why was it not broadcast more quickly and with greater urgency? This is the sort of thing that we are seeing from the response, and it’s just sclerotic and seemingly very, very conservative.
There was a lot of talk during COVID about our fragmented health-care system then, and how that was a real impediment. But this is more about the limits of our fragmented government surveillance system. Is there any workaround to that? Is that just the way the system is set up?
Tom Friedrich: As you guys both alluded to, there’s no one agency in charge. That’s not only true at the federal level, but as you were saying about the health-care system, which is also fragmented, there’s like 4,500 state and local public-health agencies in the country. And the CDC is not a national public-health agency. They have no authority to tell any jurisdiction what to do to take public-health measures. They can only make recommendations.
So that is a problem. And coming on the heels of debates about the balance between economic impacts and personal freedom and public-health response, I think we’ve chosen to err on the side of a low-key public-health response because we’re afraid of the pushback we’re going to get if things are mandated and not voluntary. People are pissed off about COVID, there’s no doubt about it. And some of the people that you’d most want to get into the fold, in terms of testing or potential vaccination — you’re talking about dairy farmers who are operating on very thin margins. And people who work on dairy farms who might be migrants, might be undocumented. They don’t necessarily have a really robust health-care plan or lots of time they could take off. So you’re talking about a very vulnerable population that’s being affected. The fear is that if I’m a dairy farmer and if H5N1 is found in my herd, and people decide that to control it, they’re going to cull my animals — I mean, it’s a gigantic economic loss.
With COVID, it seemed like people didn’t believe it could get so bad because a pandemic hadn’t happened in so long. Is the same thing happening with bird flu? The thinking being surely there can’t be two pandemics in four years, right? But with all the government/civilian conflict of COVID layered on top of that.
David O’Connor: I think it’s wishful thinking. People obviously don’t want it to happen again. And so it seems like it can’t possibly happen again. But it’s sort of like the once-in-1,000-year storms that seem to be happening with increasing regularity. You can actually have a couple somewhat back-to-back.
A lot of people in public health feel like they are being punched in the face repeatedly for their work. It’s important to bear in mind that we had a million people die of COVID in the United States, plus probably five to ten times that many who continue to live with long COVID. And it was the public-health actions that prevented the toll from being worse. Were there mistakes made? Absolutely. Were there places where we underestimated the mental-health impact and the amount of education that could be delivered virtually? Absolutely. But it’s a fog-of-war situation, and you’re going to get things wrong. And people want certainty that simply can’t be provided.
In this case, imagine if you’re in public health and you do what is perceived as a heavy-handed intervention to try to stamp out H5 in dairy cattle, and you make farmers incur these economic costs. And you’re successful, and you stamp it out. One of my favorite lines from The Simpsons was from an early season when Homer says something like, “Why are we giving Maggie vaccines for diseases she doesn’t even have yet?” And it’s the same idea, where if we stamped it out and there was no pandemic, people would say, “Look, they did all these things and it was for nothing. There wasn’t even a flu pandemic. And look at all the costs that were incurred.”
People in public health are people. They’re not just an amorphous blob of government, and they’re tired. Everyone tried to rise to the challenge in COVID. They sacrificed enormously to try to help galvanize the response and make sure that people were as safe as they could be. And I don’t think there’s an appetite right now. I think people are just sort of beaten down and, having seen what happened before, are just not as willing to go through it again.
There’s a real parallel with the exhaustion around Trump getting elected again. And now, if this outbreak gets any worse, we may be dealing with RFK Jr. as the head of HHS. Not necessarily who you want in the top spot if things spiral.
Tom Friedrich: Yeah, I think there’s a danger of having a vaccine skeptic in charge of HHS when we might want to be mounting a major vaccine.
I would say it’s not ideal.
Tom Friedrich: I’ll underscore two points Dave made. One is that anytime there’s a new situation, it’s a fog of war. People endlessly pick apart the decisions, in retrospect, of government officials. It’s second only to Monday-morning quarterbacking of the Packers.
Similar to weather forecasting, too.
Tom Friedrich: I think weather forecasting is a very good parallel. We have models, we can make some reasonable predictions based on a good understanding of the system, but each example that happens is a little bit different. And so we can only predict. And we’re going to get some educated guesses right and other ones wrong as we learn about a new virus, as it makes its way into the human population. And the things that people got wrong, or were perceived to have gotten wrong during COVID, are still so fresh in the minds of both the public and public-health officials, who got death threats over mask mandates and whatnot. I can understand why people would be a little reticent to go in and take some kind of bold action at this point.
One thing the public-health authorities got dinged for was not being totally transparent and withholding some information from the public because they thought the public couldn’t handle — treating them with kid gloves.
Tom Friedrich: Communicating uncertainty is very difficult. As scientists, when we talk to each other, we use all sorts of hedge words, like “These data suggest that …” And we’re coached when we talk to the public not to do that because people don’t want to hear, “Well, it could be true, or maybe not.” But that does shoot you in the foot when you actually are uncertain.
David O’Connor: I think you’re right, though, about transparency. And I think this was the point of Tulio’s article, which is that our government’s response has not been a model of transparency in terms of sharing data on the genetic sequences of the viruses and providing geographic information about where the sequences are coming from. When the information is dripped out very slowly and piecemeal, it makes it more difficult for scientists to put that information together. And it also just undermines faith in the entire system. I think Tulio was a hundred percent right in that piece when he said that if this particular highly pathogenic avian flu were to spread, and does become a more global threat, people are going to look at the United States and the laggard response in the first year that we’ve had to control it. And they’re going to say the inaction and the lack of transparency is an unforced error.
And there’s been so much lead time, too. I’ve done two interviews about this topic with the epidemiologist Katelyn Jetelina. In the latest one, in May, I asked her what she thought the next real warning sign would be. And she said it would be if the virus started spreading among pigs. And now we have these human cases, including three where there’s no connection to poultry.
Tom Friedrich: Some important context: I don’t know that there’s spread in pigs, although we’ve seen infection in pigs. There’s not pig-to-pig transmission that we know of yet, although I wouldn’t be surprised if that comes. We’ve actually had, in a way, a really long lead time because H5N1 viruses that have this particular HA — this attachment protein that attaches the virus to cells to infect them — have been causing a panzootic, like pandemic in birds, for several years now. And viruses that have this type of HA entered the United States through the Atlantic Flyway and the Pacific Flyway from different directions, starting way back in 2022.
People like flu virologists were noticing then that this virus seems to be extremely deadly for birds, which is unusual. And then they noticed that it was infecting a lot of mammals who ate the dead birds, and that’s unusual. The viruses that infected the poultry workers in Washington and the pig in Oregon — and especially concerning to me is the teenager in British Columbia — those are not connected to the dairy-cattle outbreak at all. The dairy-cattle virus entered the U.S. from the Northeast, and these viruses came to North America from Asia. And so they’re not identical to each other, and these ones don’t have anything to do with dairy. But, yeah, this teenager has no known risk factor, no known infected contact.
And does that indicate to you the probability of some kind of human-to-human transmission going on under the surface here, or is that too much to say?
Tom Friedrich: We can’t really speculate about that. The virus in this teenager in British Columbia appears to have some mutations that we would be very concerned about making the virus more transmissible among humans. It sounds like Public Health Canada have done a really good job; they’ve tested like 40 people who had contact with this kid, and none of them have shown any sign of infection, which is kind of amazing to me, but that’s quite reassuring. However, we’re seeing increasing exposure of humans to these viruses from different directions, not just the dairy cattle. But the more those viruses replicate and go from mammal to mammal, whether it’s dairy cattle or pigs or whatever, especially if those are animals that humans have frequent contact with … that’s the gambling Dave was talking about. We’re just giving the virus more pulls on the evolutionary slot-machine handle.
David O’Connor: I think what we have is a situation where there may be nothing that comes of this for human health, but it is a potentially low-ish-probability, very high-consequence event that warrants the investment of resources to do an all-of-the-above approach. Contrast what Tom just described in the B.C. case and the close testing of contacts to when there was an enigmatic case in Missouri earlier in the year. At least based on public reporting, there were a number of close contacts who reported symptoms. And as best as I understand, underscoring what Tom said earlier, the CDC was not invited in to help lead that response. It was coordinated by the Missouri State Public Health Department, and they were unable to, or did not for whatever reason, test a lot of those close contacts. So we don’t know if those individuals got infected or not. And it’s those kinds of gaps in visibility that just make it impossible to know what’s going on.
Tom Friedrich: My understanding is that they were tested subsequently for antibodies, some of them, but not all agreed to be tested. But they should have been followed in real time to detect the virus, if we’re concerned about person-to-person transmission. I suspected it did not cross the minds of folks in Missouri that this could be avian influenza perhaps until it was too late.
So our public-health infrastructure has wasted a lot of time not doing the right kind of surveillance on this virus. What is a realistic best path forward now, especially considering we’ll have a new presidential administration coming in?
David O’Connor: At the risk of saying something maybe a touch controversial, Tom and I were both commenting earlier that we’d listened to this week’s Ezra Klein podcast where the author of Recoding America was talking about how processes often become a substitute for judgment in terms of what the role of government officials should be. And I think we definitely see this — that a lot of public health is policy-driven. Tom and I see this in our own day-to-day lives. In our state of Wisconsin, farmers produce 30 billion pounds of raw milk a year, and yet it’s taken us five months to get permission to bring several ounces of that raw milk into our secure labs because of biosafety concerns that there might be something in this milk that would pose a risk. Even though it’s the same thing that’s being produced in the tens of billions of pounds per year and is being consumed legally and drunk by people in a dozen states.
So I think that one place where there may be an opportunity for some common cause in the new administration — where there is a pointed idea that there should be less regulation, that maybe there should be less indexing on process and more on outcomes — would be to look at this with fresh eyes and say, “What are the things that we are really trying to accomplish here? What are the goals?” And then ask the question, “Are the approaches that we’re using being driven by the best science and the best public health? Or is it being driven by other considerations, like we don’t want to step on the feet of another agency that may also have a stake in this response?”
Public health is something we all need to do. And I’d like to think that maybe we can move things faster if there is a little bit more of an emphasis on outcomes. A lot of things that public health needs to do may be unpopular at the individual level. It may be difficult for individuals, but it’s needed for community health, for literal public health. And maybe in a new administration, an optimistic take is that a reduction in regulation would be one potentially positive outcome that could lead to a more effective response.
There’s the regulation part, and then there’s the tension between the individual and the communal here in the U.S.
Tom Friedrich: I think that’s true. I would add that there should be incentives, and maybe a new administration would find some will to do this. Because ultimately, we need to incentivize cooperation from farmers and their workers to be able to go into farms and do testing. And something that we’re not really talking about here, but underlies all of this, is just a reduction in trust in governments and institutions generally. If we want to avert a pandemic, if a pandemic is about to happen, then people are going to have to act decisively, and very quickly, it’s going to get out of control. So you have a narrow window of opportunity where by definition you have to be acting with incomplete information. If you’re going to be very deliberative about the whole thing, you’re just not going to be able to contain it.
It’s reminiscent of the early days of COVID. All the tests had to go to the CDC in Atlanta or whatever, which clearly wasn’t going to work.
Tom Friedrich: Yeah, you’re literally closing the barn door after the virus has gotten out. So it’s about doing whatever brings those constituencies onboard, including gaining the trust of the farmers and workers who need to be part of the solution by incentivizing testing, whether that means providing support or actual funds to offset losses that they might incur when the virus is found. Because basically, there’s very little upside to a farm to find the virus on it, and a lot of downsides.
David O’Connor: One thing building on that point is that viruses thrive in the shadows. So for everything I just said about things that could be good about reducing regulation, recognizing that a huge fraction of our agricultural sector is supported by undocumented workers means that a lot of what is being proposed for the new administration threatens to make things much, much worse. All the things Tom just said are completely correct, but they’re unlikely to occur if the people who work on the farms are driven further underground. It makes them much less likely to show up in an emergency room if they’re sick. It makes them much less likely to agree to being tested. So that’s going to be a tension that could potentially allow this virus, or other H5 influenzas — as Tom mentioned, there are multiple genotypes — to spread more quickly under the new administration.
This interview has been edited for length and clarity.
More From This Series
- They Shocked Assad and the World. What Will They Do Next?
- How Trump Could Attack Trans Rights
- How Luigi Mangione Probably Gave Himself Away
