Even if you don’t follow flu news closely and don’t live in China, by now you have probably encountered the new term “H7N9” once or twice. H7N9 is a subtype of bird flu that started jumping from birds to people last month in China. The Chinese government and the Chinese media (closely interconnected institutions) announced the first three cases to the world on March 31, eight days ago today. As of this writing, China says H7N9 influenza has been confirmed in 24 humans, and has killed seven of them.
I want to write about H7N9 risk communication in its first week, especially the H7N9 risk communication of two CDCs – the China Center for Disease Control and Prevention (China CDC) and the United States Centers for Disease Control and Prevention (U.S. CDC).
Based on what is known and suspected so far, China CDC has been unusually candid, greatly surpassing the very low bar China’s government set for itself a decade ago when it tried to cover up news of SARS. And the U.S. CDC has been understandably but perhaps unwisely low-key, not suppressing any H7N9 facts that I am aware of but certainly downplaying the possible significance of those facts.
Before I get to the risk communication, here are the technical basics:
- H7 is a subtype of bird flu virus that has spread from birds to humans infrequently in the past, usually causing only conjunctivitis in poultry workers. H7N9 in particular had never been found in humans until the first human cases were reported in eastern China in March 2013.
- So far there are only 24 known human cases. But an alarmingly high percentage of those cases (seven out of 24, or 29%) have died, and many of the rest are in critical condition. H7N9, in short, appears to be rare but deadly in humans. We’re not sure even of that; mild or asymptomatic human infections may be found as more people are tested in the areas where human cases have occurred.
- There are no known connections between any of the confirmed H7N9 patients. Patients’ friends and family who have been tested for H7N9 have tested uniformly negative. There is no laboratory evidence of any human-to-human transmission to date. However, two sons of the index patient, an 87-year-old man who died, also had pneumonia around the same time as their father. Laboratory tests on samples from the two sons were negative, but a lot of experts wonder if the sons might also have had H7N9. If they did, it could have passed between family members, or all three could have been exposed to a common avian source.
- Some of the patients had known contact with birds, but others didn’t. A search for H7N9 in birds in the parts of China with human cases has now found a genetically very similar virus in chickens, quails, and pigeons. These infected birds have not been symptomatic. The absence of avian symptoms, of course, makes it very hard to tell which birds are infected or how widespread H7N9 is already in birds.
- Although pigs are a frequent “mixing vessel” for bird flu viruses en route to humans, and although eastern China had an unexplained pig die-off last month, H7N9 hasn’t been found yet in pigs, or in any non-bird species except humans.
- The particular kind of H7N9 that is killing people in China shows genetic signs of adjusting to mammalian hosts. That could mean it is getting closer to being able to transmit human-to-human. Or not. This is the big question to which nobody has an answer: What are the chances that H7N9 will start spreading human-to-human like an ordinary human flu?
- Since it’s a new human strain, nobody has any immunity to H7N9. And there is no vaccine. Work started on developing one a few days ago, though there are no plans yet to mass-produce it and no reliable estimates yet of how long that would take. Even if H7N9 ends up endangering the entire world, most of the world’s population is unlikely ever to be offered an H7N9 vaccine.
- If H7N9 becomes capable of efficient human-to-human transmission, we’re in for another flu pandemic. And if pandemic H7N9 remains anywhere near as deadly to humans as H7N9 is so far, it could be the most horrific flu pandemic ever, a lot worse than the infamous 1918 pandemic. (The flu case fatality rate in 1918 was around 2.5%, compared to H7N9’s 29% so far.)
- The other significant threat is that H7N9 could acquire the ability to pass easily from birds to humans, though not from humans to humans. (This is not known to have occurred in the past with influenza – but influenza is notoriously full of surprises.) A flu strain that was widespread in birds, deadly in humans, and easily transmitted from birds to humans wouldn’t be a pandemic. It would be a “panzoonotic.” But given how many people in the world are exposed to poultry and other birds, it could still be horrific.
The ifs surrounding the pandemic and panzoonotic threats of H7N9 are big ifs. Since 1997 a different bird flu subtype, H5N1, has similarly threatened to go pandemic (or conceivably panzoonotic). So far it hasn’t done so. It continues to transmit occasionally from bird to human, and very rarely from human to human. It is even deadlier than H7N9, killing about half of the small number of people who catch it. H5N1 accounted for 79 lab-confirmed human deaths in its deadliest year so far (2006) – 79, not millions.
Some experts think there is more reason to worry about H7N9 than H5N1. For one thing, we’ve known about H5N1 in humans since 1997 and it has been clearly here to stay since 2004, but it hasn’t yet launched a pandemic, giving rise to the argument that if it were going to do so it would already have done so. H7N9, by contrast, has been making people sick for only a few weeks, as far as we know.
A technically sounder basis for worrying more about H7N9 is this: H7N9 shows signs of mammalian adaptation that H5N1 doesn’t show. A World Health Organization H7N9 website page, for example, notes that “analysis of the genes of these viruses suggests that although they have evolved from avian (bird) viruses, they show signs of adaption to growth in mammalian species. These adaptations include an ability to bind to mammalian cells, and to grow at temperatures close to the normal body temperature of mammals (which is lower than that of birds).” Flu expert Richard Webby put it more succinctly in an interview with the Canadian Press. “I think that’s what’s concerning about this,” he said. “This thing doesn’t any longer look like a poultry virus.”
What kind of case can be made that H7N9 is less scary today than H5N1 was when it was new on the scene? No case at all, technically. But non-technically, two big things have changed since H5N1 made headlines in 2004–2007. First, no H5N1 pandemic has materialized, which leaves those who aggressively sounded the H5N1 alarm almost a decade ago (including me) looking a bit foolish to many people. And second, in 2009–2010 we actually had a flu pandemic, not bird flu but H1N1 swine flu, and it turned out milder than a typical flu season, so mild that public health agencies were widely accused (especially in Europe) of manufacturing a fake pandemic in order to sell vaccine. In the wake of the non-appearance of the H5N1 pandemic and the mildness of the H1N1 pandemic, it must be tough for officials and experts to summon the courage to issue strong warnings now – with a grand total of seven deaths so far – about the prospect of an H7N9 pandemic.
On the one hand, then, flu experts are feeling pretty nervous about H7N9. On the other hand, you can see why they wouldn’t want to risk being perceived as alarmists yet again – and why they wouldn’t expect journalists or the public to take to heart yet another flu pandemic warning. No wonder H7N9 has received fairly muted public attention from the experts, and fairly muted coverage in the media.
The wonder is that H7N9 has gotten as much attention as it has – that more than a few experts have said alarming things to reporters and more than a few reporters have used what they said. Helen Branswell of the Canadian Press – widely considered the Dean of flu reporters – tapped her extensive Rolodex for an April 4 story that began this way: “Three sick people. That’s all it took to make scientists who study influenza edge-of-seat nervous this week over a rapidly evolving public health situation in China. Though the case count is now up to nine infections and three deaths, those first three sick people and the genetic sequences of the flu viruses that infected them were enough to make the hairs on the backs of knowledgeable necks stand on end.” And commentator Laurie Garrett wrote a blog for the Foreign Policy website entitled “Is This a Pandemic Being Born?”
But most of the coverage has been a lot more reassuring than that – calm coverage for a calm and even skeptical public.
It’s a different story in China. China lived through SARS and H5N1 firsthand, whereas the U.S. experienced them mostly as faraway news stories and possible future threats that never materialized. And China is living through the first few H7N9 cases firsthand as well. If the public reaction in the U.S. ranges predictably from ridicule to apathy to casual interest, the reaction in China can be expected to be significantly more anxious.
That is the context in which we should look at how the two CDCs handled the first week of the story.
The Report from China
The first announcement of human H7N9 came on March 31: three cases, two dead in Shanghai and the third hospitalized in Anhui. The initial announcement seems to have come from China’s National Health and Family Planning Commission, but China CDC wasn’t far behind with a March 31 Q&A on its website.
The key question was number six:
6. Can A(H7N9) avian influenza virus [be] transmitted from person to person?
Given the limited number of confirmed cases to date, it is still uncertain how this virus is being transmitted from birds. Epidemiology links between the 3 confirmed cases haven’t been discovered. Investigation is ongoing, but so far no further cases were identified among the contacts of the case. Up to date, no subclinical infections have been determined from all the blood taken in contacts. But we cannot rule out the possibility of human to human transmission until we know more about the virus characteristics and results from ongoing investigations.
In an April 6 blog post, my wife and colleague Jody Lanard said she was “pretty impressed” by the last sentence of this answer: “This sounds so ‘ordinary,’ but is actually highly unusual for China to acknowledge uncertainty like this. It embeds some “anticipatory guidance” that things might get worse. It puts the reassuring information first – no known cases of H-2-H spread – but ends on a note of caution, as Peter Sandman and I always advise [in uncertain situations].”
Jody was also impressed, as I was, by how quickly China CDC and other Chinese authorities told the world about H7N9. “If these really are the first cases China has found,” she wrote, “they deserve kudos for early reporting and non-over-reassurance.”
China CDC wasn’t just transparent and timely in its H7N9 announcements. As Jody points out in her blog post, it kept emphasizing its intention to remain transparent and timely. That’s the kind of public promise that gives invaluable ammunition to the proponents of candor in whatever internal battles may arise: “Look,” they can say. “We said we’d be honest. Now we have to live up to what we said.”
Promises to be transparent and timely are most credible, and most binding, when they’re accompanied by acknowledgments of prior failures of transparency and timeliness. Making such promises (and thus increasing the chances of living up to them) seems to be part of a coordinated high-level government decision. That’s clearest in an April 3 editorial released by Xinhua, the official government press agency, stunningly entitled “Ten years after SARS, what has China learned?”
After reminding readers of the SARS cover-up and its role in making SARS even more terrifying to the Chinese public than it would otherwise have been, the editorial criticizes the Shanghai government “for not notifying the public about the two H7N9 deaths until nearly half a month after the deaths occurred.” It stresses that authorities “need to make persistent efforts to satisfy people who have become much more aware of their right to knowledge regarding public health issues.” And it concludes:
If there is anything that SARS has taught China and its government, it’s that one cannot be too careful or too honest when it comes to deadly pandemics. The last 10 years have taught the government a lot, but it is far from enough.
Jody’s comment: “I never expected to see such a bold critique from Xinhua. I assume this is being done with the blessing of the government. If that is true, it is a very good sign indeed.”
Not all the signs from China are as good.
On April 2, the head of the Shanghai Health Bureau said: “The health bureau will take effective and powerful measures to prevent and control the disease, to make sure the flu epidemic is effectively guarded against and to safeguard the health of the city’s residents.” Any high-ranking health bureaucrat surely knows this is an empty promise.
And an April 7 Reuters story was headlined “China confident it can control bird flu outbreak.” This time the empty promise was attributed to Li Bin, the head of China’s National Health and Family Planning Commission and, according to Reuters, “the most senior Chinese health official yet to comment publicly on the subject.”
Jody noted Shanghai’s over-reassurance in her original blog; Commissioner Li’s over-reassurance prompted her to add a less optimistic postscript.
Another dereliction Jody didn’t mention: On April 3, China’s Ministry of Agriculture told Xinhua that H7N9 had been found in wild birds around the world but never in poultry, strongly implying that wild birds rather than poultry were probably the source of the recent human infections. The next day H7N9 was found in Chinese poultry – a point flu blogger Michael Coston made in his brief commentary on Jody’s blog.
Coston concluded: “I suspect some well-ingrained bureaucratic habits are going to be very hard to break. But compared to the ‘the bad old days’, China does appear to have been far more forthcoming over this past week, and less cumbersome in their messaging.”
The Report from the U.S.
The risk communication challenge confronting China CDC was daunting but obvious: Tell as much of the truth as you know; keep acknowledging that there’s a great deal you don’t know; acknowledge your uncertainty and the possibility that things could get a lot worse; resist the temptation to over-reassure your public, which is going to become very worried as soon as it hears about these cases. By and large, so far, China CDC has done well – certainly by comparison with the Chinese government’s previous record.
The risk communication challenge confronting the U.S. CDC was less daunting, but also less obvious. Of course it needed to tell what it knew about events in China, and what it was doing or proposing to do to respond to those events. But the U.S. public wasn’t very worried; if anything the U.S. public and U.S. journalists were insufficiently worried, inured to the threat of flu pandemics by their experience with H5N1 and H1N1. So there should have been little if any temptation to over-reassure. The tough question: Should the CDC have gone along with the overall mood of mild, casual interest? Or should it have tried to pierce the apathy with warnings that “Yes, H5N1 wasn’t The Big One and H1N1 wasn’t The Big One and maybe H7N9 won’t be The Big One either. But maybe it will. Someday we will face The Big One, and that day just might be coming this time.”
The CDC chose to go along with the overall mood of mild, casual interest. I’m not sure this was the wrong decision. Piercing the apathy would have been difficult and perhaps premature. If H7N9 never amounts to much, the CDC might have done some damage to its credibility by trying to pierce the apathy. On the other hand, if H7N9 does start looking like The Big One, the CDC will have passed up an early teachable moment. But on a third hand, if the situation starts to look worse, more teachable moments will come along, and the CDC can prepare for them in advance: reminding its experts not to over-reassure and not to sound overconfident about managing a pandemic; figuring out what to ask of the public; etc.
In June 2007, expert concern about a possible H5N1 pandemic was high, but public concern was fairly low (though it was a lot higher than public concern about H7N9 right now). That’s when I wrote an article for employers entitled “Why talk now: The case for communicating with employees before the pandemic arrives.” The article begins: “Your company is preparing for a pandemic, but your employees aren’t. Can that possibly make sense?” It goes on to list ten reasons why I thought employers needed their employees to take pandemic risk as seriously as they were taking it. Not all my reasons apply to the U.S. government and H7N9 as clearly as they applied to employers and H5N1. But some of them do:
- You need employees [the public] to know that influenza pandemics are unpredictable.
- You need employees [the public] to prepare at home.
- You need employees [the public] to prepare emotionally.
- You need employees [the public] to know how your company [government] has prepared.
- You need employees [the public] to help your company [government] prepare.
- You need employees [the public] to see you as a useful source of pandemic information.
I am especially preoccupied with #3 on this list, preparing people emotionally. The theory of adjustment reactions predicts that people go into a kind of funk – the adjustment reaction stage – when they first become aware of a crisis threat. We must get through our adjustment reactions before we can start thinking clearly, planning sensibly, and acting collaboratively. One crucial goal of pre-crisis communication is to help people into and through their adjustment reactions, so they’re ready to cope when the crisis arrives.
Still, we have known about H7N9 for only a week and a day. It’s not crazy to think that the CDC could wait a while to see how the pandemic risk evolves before it starts trying to shake people by their collective lapels.
So I’m not sure how critical I am of the way the U.S. CDC handled its first and so far its only H7N9 press briefing on April 5. But I do want to document what it said and didn’t say in that briefing.
To start with, the P word – “pandemic,” that is – appears not once in the press briefing transcript. Neither CDC Director Tom Frieden nor CDC flu epidemiologist Joe Bresee (the two presenters) used the term, nor did any of the journalists who asked questions. Frieden and Bresee did make frequent reference to the issue of human-to-human transmission, emphasizing that there was no evidence of any so far but acknowledging that that could change. The possibility of h2h transmission would signal “possible pandemic” to a knowledgeable reporter – and most of the reporters participating in the briefing were knowledgeable. Still, it says something that they could all get through an entire briefing without naming the scenario that had to be uppermost in everybody’s mind.
In his introduction, Frieden quite properly stressed that “[t]here has been no person-to-person transmission and no epidemiologic link between any of the cases so far.” This had two implications, given that there had been 16 cases from four different Chinese provinces reported at the time of the briefing. First, it presumably meant that H7N9 wasn’t transmitting efficiently from person to person; one or two such cases might have been missed, but if h2h transmission were the main way the virus was spreading, we’d almost certainly have seen some sign of it already. The second implication is less reassuring: The virus was presumably widespread enough already in one or more species (probably bird species) to cause 16 human infections with no obvious connection in four not-very-close provinces. This pretty obviously wasn’t a case of everyone getting infected at one live bird market.
Frieden and Bresee focused on the first implication: no h2h transmission. They didn’t mention the second implication: too late to lock the barn door.
The last two paragraphs of Frieden’s introduction are most revealing of his decision to go along with the overall mood of mild, casual interest:
Before I turn the call over to Dr. Bresee to answer your questions, I wanted to put this into perspective. Two questions that CDC is frequently asked are, how concerned are we and how concerned should the public be about the situation? And I’ll say this. We at CDC work 24/7 to protect people’s health, including from threats such as emerging influenza viruses. We work to have the public’s back. It’s our job to be concerned and to move quickly whenever there’s a potential problem, such as the one in China. And that’s what we're doing. As far as should the public be concerned, there is no evidence at this point that the virus is being transmitted from one person to another or that the virus has caused any infections in the U.S.
There are no specific steps that people in this country need to take to protect themselves. People can go about their daily lives. Information can be a powerful antidote for people who are worried or concerned….
This is as clear as clear could be. “It’s our job to be concerned” so you the public don’t have to be. We’re paying close attention to H7N9 so everyone else can shrug it off. Everyone else can shrug it off because there’s not yet any evidence of h2h transmission and because it hasn’t yet infected any Americans. We're giving you this information as an “antidote” for “people who are worried or concerned.” The goal of the information is thus to keep people from being worried or concerned. Maybe he should have said “anodyne” instead.
The CDC took much the same tack at the start of the swine flu pandemic, prompting the title of my first swine flu column in April 2009: “The Swine Flu Crisis: The Government Is Preparing for the Worst While Hoping for the Best – It Needs to Tell the Public to Do the Same Thing!” My point here is a double-edged sword, of course. Once again the CDC is passing up the opportunity to help people into and through their pandemic adjustment reactions. The CDC was arguably right (post hoc) not to push for a more robust initial public reaction to swine flu, given how mild it turned out in the end. But warnings shouldn't be judged by how an uncertain situation actually turned out, when in real time the potential for a dire outcome was real.
I’ll give one more indication of the reassuring thrust of the April 5 press briefing. Probably the most alarming fact about H7N9 is the suggestion that it is adapting to mammalian hosts. Every time Frieden and Bresee pointed out that there has been no documented h2h transmission yet – which they pointed out numerous times – they had an opportunity to mention that the evidence of adaptation was reason to worry that h2h transmission might be coming. They passed up every such opportunity. They passed it up even when Miriam Falco of CNN asked, “Is there something about this virus that you think might make it more eligible to … quickly change to become more easily transmissible?”
Finally Richard Knox of National Public Radio raised the matter directly: “[Senior CDC scientist] Nancy Cox was telling me that genetic analysis indicates this virus might infect mammalian cells more easily than H5N1 does and I wonder … how heavily the genetic analysis of that weighs in your current threat assessment.” Notice how Frieden’s answer briefly acknowledges the mammalian adaptation, then segues to more reassuring information:
So we always have to analyze the genetic sequence in concert with the epidemiologic pattern. It is true that some aspects of the genes of this organism are partially adapted to mammals, suggesting that they may have spent time in swine or other populations. But ultimately the fact that we haven’t seen cases in contacts, that we haven’t seen widespread cases in children, that we haven’t seen a spike are all reassuring.
The reassuring information in Frieden's answer is all true. But nothing in this answer, or any answer, comes near to suggesting that the news could get a lot worse in the weeks ahead and that people may want to start preparing (at least emotionally) now.
Frieden’s concluding comment:
Sunlight is the best disinfectant and there are concerns when we hear about people becoming severely ill…. [T]he information we have so far suggests that all of the cases have been isolated, that they are mostly associated with spread of this virus in animals, likely poultry, chickens, ducks, geese, pigeons, perhaps. And we’re in close collaboration with authorities in other countries, including China, so that we can rapidly learn more about where it’s coming from and how – whether it has spread from person-to-person. This is an example of why we need to keep our guard up and make sure we're prepared to detect and respond to threats as they emerge whether in this country or around the world. So thank you very much for joining.
What Frieden seems to be trying to “disinfect” is public concern. It’s just isolated cases spread by some animal. We’re working with China and if any human-to-human transmission turns up we’ll be prepared to respond and protect everybody. Good night. Sleep tight.
I accept that, statistically, the CDC’s posture of reassurance is likelier than not to turn out accurate – that H7N9 is likelier to fizzle than to launch the Great Pandemic of 2013 or 2014 or 2015. Most potential pandemics don't happen. And I realize how hard it is to warn people that things might get bad without sounding like you’re predicting that things will get bad and then getting ridiculed or worse when they don’t. But sooner or later things will get bad. Isn’t it part of the CDC’s job to keep reminding us of that, especially in moments of high technical uncertainty about how bad things might be about to get?
I sympathize with the temptation to over-reassure when people are likely to become super-anxious. That’s why I’m so impressed that China CDC has mostly resisted that temptation so far. But when people are apathetic or close to it, and likely to remain so, the rationale for over-reassurance isn’t to keep them calm. The rationale, if there is one, is to avoid getting accused of hype. That strikes me as a pretty weak reason for neglecting the duty to warn, for over-reassuring the U.S. public about a possibly pandemic flu strain that has just made its appearance in China.
The NBC website story about the CDC press briefing is headlined: “Don’t panic over new bird flu outbreak, CDC cautions.” The CDC isn’t responsible for that headline, and “panic” is another P word that didn’t get any use at the briefing. Nonetheless, the headline and the story that follows capture the tenor of the briefing correctly: It aimed to tell a public that already wasn’t alarmed that there was no reason for alarm.
A fundamental goal of risk communication is to replicate in your audience the level of concern that you yourself are experiencing. If you’re less concerned than your audience is, it’s appropriate to try to reassure. If you’re more concerned than your audience is, you should try to warn. I’d lay odds that right now the experts at the U.S. CDC are more concerned about H7N9 than the U.S. public is. They nonetheless decided to conduct a reassuring press briefing. Maybe – maybe – that’s okay in Week One. If the U.S. CDC is still concerned a few weeks from now, I hope it won’t still be hiding its concern.
Added April 12, 2013:
The foregoing column was written and posted on April 8, 2013. On April 11, the CDC came at least partway out of the closet about its H7N9 worries, in a “Perspective” article in the New England Journal of Medicine by CDC flu experts Timothy M. Uyeki and Nancy J. Cox, and in an impromptu CDC news conference to answer reporters’ questions about the article.
Candidly entitled “Global Concerns Regarding Novel Influenza A(H7N9) Virus Infections,” the article begins by acknowledging that H7N9 “might herald sporadic human infections from an animal source … or it might signal the start of an influenza pandemic.” Then it focuses on the reasons for worrying about the pandemic possibility, despite the absence so far of any evidence of human-to-human transmission. Among the reasons stressed are the genetic signs of adaptation to mammalian hosts and the challenges to developing an effective vaccine.
I can find things to complain about in the article. It focuses more on whether an H7N9 pandemic might occur than on how severe such a pandemic might be (though it does discuss the severity of the disease in some patients). And its claim that progressing “from vaccine development to availability will probably take many months” may sound like a warning, but it’s actually awfully optimistic, at least if “availability” is taken to mean worldwide availability to ordinary people and if “many months” is taken to mean less than a couple of years.
Still, the April 11 article can’t be called over-reassuring, as the CDC’s April 5 press briefing was. And the fact that the CDC scheduled a last-minute news conference about the article suggests that it intends to share its concerns with a broader public than New England Journal readers.
Media coverage picked up the new, more alarming focus. “H7N9 flu ‘may pose a serious human health risk,’ scientists write” was the Los Angeles Times headline. The Canadian Press story emphasized that “Making vaccine for new flu H7N9 flu virus could be a challenge, experts say.” The Wall Street Journal similarly headlined that “Vaccine to Fight New Bird Flu Strain Could Be Elusive.”
NBC’s website came closer still to what I think the focus should be. Its story on the New England Journal article (and other developments) was headlined “Deaths from new bird flu underscore grim fears, reports show.”
How much of a change is this? The same NBC reporter, JoNel Aleccia, had covered the April 5 CDC press briefing in the story headlined: “Don’t panic over new bird flu outbreak, CDC cautions.” From “don’t panic” to “grim fears” in less than a week.
My wife and colleague Jody Lanard contributed to this column.