Search Peter’s website:

Watching the COVID Booster
Sausage Being Made

This essay is a bit of a self-indulgence in several ways. It’s a lot longer than many readers will care to endure. And a lot of it is only tangentially about risk communication. It’s largely about the policymaking process, and what that process itself communicates to the public – especially what it communicates about the doubts, disagreements, and dilemmas of public health professionals.

At least some of this essay goes beyond my expertise – which is pretty cheeky, since it accuses public health professionals of going beyond theirs. For example, I offer summaries of the scientific evidence relevant to boosters and the legal prerogatives of the FDA and the CDC that are blithe and oversimplified – and possibly flat-our mistaken in spots. My wife and colleague Jody Lanard usually reads my drafts, and finding my oversimplifications and mistakes is one of her many contributions. She decided this one was longer than she wanted to read.

I started writing this back in August, when President Biden first proposed making COVID boosters available to all American adults, and the public health profession responded with considerable outrage. I kept revising it – and lengthening it – as the policymaking process continued. By late October I had a near-final draft that focused largely on what I considered a face-saving set of FDA and CDC decisions that actually allowed anyone who wanted a booster to get a booster, while seeming to impose complicated restrictions on booster eligibility. The public and the politicians had supported a boosters-for-everyone policy when the agencies and experts opposed it. They led the way and forced the issue. The agencies and experts resented that enormously, and found ways to save face even as they lost the fight.

I was still tinkering with my draft as the evidence kept getting stronger that fully vaccinated people might actually “need” boosters. The agencies’ policies and the experts’ opinions kept changing too. Now, effective November 19, Biden has the policy he sought in August: universal adult COVID booster eligibility. Boosters-for-everyone has emerged from the closet, explicitly endorsed at last by the FDA, the CDC, and most of the experts. So my blow-by-blow description of the COVID booster “sausage-making” process is arguably beside the point.

But I think it’s fascinating and instructive. I hope there will be at least a few readers who will think so too.

Note on out-of-datedness: I wrote most of this essay in real time, and I didn’t always revise as fully as I should have as events evolved. Two closely connected problems I need to acknowledge. One, when booster eligibility criteria changed, webpages on which those criteria were explained sometimes changed too – so some of my links no longer say what I said they said. If you want to check on me, use the Wayback Machine. Two, a lot of what I wrote in the present tense stopped being true as both the booster eligibility criteria and the booster attitudes of public health professionals morphed. Sometimes I went back and adjusted my verb tenses; sometimes I didn’t.

I finished what I thought was my final edit on November 23 and sent this long essay off to my webmaster. On November 25 we first learned about Omicron. I have resisted any further updating except for this paragraph. As I write this on December 2, we don’t know much about Omicron yet. But the CDC no longer simply permits everyone over 18 to get a booster. It now recommends with something like urgency that every eligible adult get a booster, in the hope that if our vaccines work less well against Omicron than they do against Delta, a booster might help keep us safe. Mainstream opinion among public health professionals is now firmly pro-booster.

For those who want a sampler, I have also put together a set of excerpts, with links back to the full essay.

---

In 1869, poet John Godfrey Saxe wrote: “Laws, like sausages, cease to inspire respect in proportion as we know how they are made.” (The quote is often misattributed to Otto von Bismarck.)

That’s how public health professionals feel about policies in their field: If the public gets to watch them hashing out their many uncertainties and disagreements, they believe, public confidence in the final product will be shaken.

They’re not entirely wrong. The public does prefer to imagine that public health experts are in omniscient agreement. The question is whether public confidence is a blessing or a problem when it is grounded in not realizing how uncertain the experts actually are and how much they actually disagree with each other.

That’s not a question in the minds of most public health professionals. They’re quite sure that uncritical public confidence in their wisdom and their unanimity is a blessing, not a problem. That’s a big piece of why they believe so fervently that they should “speak with one voice” – a mantra with which I strongly disagree unless they all truly have the same opinion.

So most public health professionals have been dismayed by the transparent process that began on August 18, 2021 when the White House proposed making boosters available to every adult COVID vaccinee, and that ended on November 19 when the Food and Drug Administration (FDA) and the Centers for Disease Control and Prevention (CDC) signed on to exactly that policy.

They were especially dismayed by the first couple of months of this three-month process – when it was embarrassingly clear that the agencies and experts were reluctantly following the lead of the public and the president, instead of the other way around. In the final month the agencies and experts were able to convince themselves that the evidence was now strong enough that they could fully and explicitly endorse universal boosters as a science-based decision, without looking like they were deferring to public opinion and political power.

I was delighted with the process from beginning to end. I think it’s a good thing for the public to watch the sausage being made. Trust in the experts’ faux-omniscience and manufactured consensus was never a sustainable goal. Better to aim for a public that understands and accepts how uncertain and divisive so many COVID policies are – and how much they entail value judgments and not just scientific judgments. Those uncomfortable truths have never been clearer than they were as the FDA and the CDC confronted the question of COVID vaccine boosters.

This very long essay will be mostly about the COVID booster sausage-making process – the process of deciding to approve boosters for the Pfizer, Moderna, and Johnson & Johnson (J&J) COVID vaccines. But first I need to clear away some underbrush.

---

The Importance of Boosters

1. Boosters are not the most important thing.

For the world as a whole, and for every community within it, the top COVID priority is getting more people their primary vaccine doses. Whether or not that can “end the pandemic,” it will surely save many lives and make all our lives less unpleasant. Vaccinating the unvaccinated is Job One – not boosters.

Boosters probably aren’t Job Two either. In the U.S., I suspect, Job Two is increased testing – especially quick, easy, inexpensive COVID tests. Or maybe better treatments for people who catch the disease. Not boosters. At least not yet – not as long as relatively few vaccinated people are getting severely ill with COVID. For a public health professional looking broadly the way public health professionals are trained to look, boosters are somewhere down on the priority list.

So many public health professionals have been understandably irritated to be forced to attend to the booster controversy – and devote time to actually giving people booster shots – when they have so much else on their plate that they consider more important.

2. But boosters are still pretty important.

But getting a booster shot is rationally a pretty high priority for an individual who was “fully vaccinated” more than six months ago, especially someone who is elderly or who has a comorbidity or two (or both). Is getting a booster shot more or less effective as an individual risk-reduction strategy than wearing a high-quality mask wherever you go, or giving up on working out in the gym and singing in the church choir?

Nobody knows. It is certainly easier, less disruptive, and less unpleasant.

Whatever precautions fully vaccinated people are continuing to take even though they are fully vaccinated, it is bound to feel irrational to the point of craziness to tell them that they must keep taking those difficult, disruptive, unpleasant precautions but should absolutely not get a booster shot yet as well.

Consider boosters versus masks, for example. Both are COVID precautions whose value is difficult to measure but clearly greater than zero. Boosters are momentarily painful. Masks are endlessly uncomfortable. I know very few vaccinated people who would rather wear a mask for the duration than get another shot. And if it comes to that, I know very few vaccinated people who would rather wear a mask forever than get a shot every six months forever.

Put aside the complexity of the debatable extent to which boosters and various sorts of masks reduce transmission and thereby protect others. Purely in terms of protecting the individual booster recipient / mask wearer, what attitude toward these two precautions makes sense? Of course the answer depends partly on how much protection each precaution provides. That’s debatable too. So we all have to make decisions about these two COVID precautions without good data on their effectiveness. Which choices make sense?

I realize they’re different sorts of precautions. You either get a booster shot or you don’t, whereas you can don and doff your mask at will, depending on who’s around, how much COVID is circulating, and other factors. And it’s not an either/or choice:

• It may make sense to do neither, to judge that being fully vaccinated is protection enough.
• It may make sense to do both, to maximize safety by taking all feasible precautions.
• It may make sense to prefer the booster because it has a smaller downside – to plan on ditching the mask (or at least wearing it less often) once your immunity has been boosted.
• It makes no sense whatever to don your mask time after time after time, but decline to roll up your sleeve one more time. Yet that’s the option public health professionals who opposed widespread boosters were urging throughout September and October – the option that makes no sense in terms of individual risk management.

Sure, some people hate getting shots and don’t find masks all that uncomfortable. Some people are convinced that, for them at least, boosters won’t add as much protection as a properly worn, close-fitting mask. Some people are focused mostly on trying to figure out which precaution protects other people more.

But here’s the bottom line: If I am high-risk enough to make me wear a mask much of the time, how could I possibly not be high-risk enough to let me get a booster?

On September 16, when much of the public health profession was actively agitating against widespread boosters, the Washington Free Beacon ran an article by Aaron Sibarium entitled “Critics of Boosters Still Want Lockdowns.” Writing about an influential anti-booster paper in The Lancet, Sibarium reported that many of the paper’s 18 authors, including top FDA and World Health Organization officials, “have consistently opposed lifting public health restrictions in the face of new variants, guidance that seems to contradict their argument about the mildness of breakthrough cases.” He added: “The Washington Free Beacon asked those scientists why, if boosters aren’t necessary, masks and social distancing are. They did not respond to a request for comment.”

3. Vaccinees who want boosters are public health’s home team.

Despite the above boosters-versus-masks reasoning, my strong hunch is that most booster enthusiasts don’t plan to ditch their masks any time soon. They want to add a booster shot to their armamentarium, not substitute it for more unpleasant precautions. My guess is that vaccinated people who have been boosted have a higher rate of mask-wearing than vaccinated people who decided not to bother.

This raises a point I think is crucial. The people who craved a booster shot as the agencies were debating whether to give their okay and the people who got a booster shot as soon as they were allowed to do so are the public health profession’s “home team.” They’re the people who surfed kludgy websites late into the night to get an early vaccination appointment, who never wear their masks below their noses, who can discourse knowledgeably on the implications of aerosols versus droplets.

Public health professionals should ideally avoid sounding contemptuous of anybody. But their expressions of contempt for antivax activists are at least understandable. It’s even understandable that sometimes that contempt gets generalized to people who are more vaccine-hesitant than antivax – especially if their hesitancy takes the form of suspicion about the integrity and competence of public health professionals. But it was incredibly self-defeating for public health professionals to voice contempt for booster enthusiasts, their strongest supporters. That struck me as nearly pathological. “Booster enthusiasts are what politicians call your base!” I wanted to shout. “Stop dissing your peeps!”

Now that boosters have the official okay and are going into arms, these expressions of contempt have quieted considerably. They were most apparent vis-à-vis people who sought boosters before they were officially okayed, or people the author believed didn’t fit the complicated eligibility specifications in force for a month in late October and early November. These people were variously described as “clamoring” for boosters, “scampering” to wherever boosters were available, or “abusing” the booster distribution honor system. Their determination to avoid illness – something public health professionals usually admire and encourage – was routinely dismissed as “selfish” and “irrational.”

Even back in August, September, and October, not all public health professionals were anti-booster. The experts and the profession were genuinely divided. And unlike some other COVID controversies where the experts and the profession were quietly, almost secretly divided, this time the disagreement was visible to all. There were highly respected leaders outspoken on both sides, pro-booster and anti-booster. I thought the transparent expert debate was great. But most experts and public health professionals found it horrifying. So they deserve credit for not pushing the debate underground and enforcing a fake consensus, as they have too often done in the past.

What wasn’t so great was the rancor – especially the expressions of contempt on the part of booster opponents for fully vaccinated Americans who wanted that third dose. They were and are the public health profession’s home team. My plea to booster opponents: Just as antivaxxers are your natural adversaries, booster enthusiasts are your natural allies. They’re your boosters, not just booster boosters. Disagree with them if you must, but don’t alienate them.

---

The Science

4. The science isn’t and wasn’t controversial.

Throughout September and October I kept hearing that the science of boosters was hotly controversial. It wasn’t. Essentially all the experts were reading the same papers and assessing the same evidence – and they were reaching the same scientific conclusions. They pretty much agreed on what was known about boosters, what looked likely but not yet settled, and what was anybody’s guess.

To see how true that is, just watch the videos of the September and October booster discussions of the two key U.S. advisory committees. Here are the first two, on the Pfizer booster:

• The September 17 meeting of the FDA’s Vaccines and Related Biological Products Advisory Committee (VRBPAC)
• The September 22–23 meeting of the CDC’s Advisory Committee on Immunization Practices (ACIP)

These meetings were substantially more contentious than the ones that followed. But you won’t see any shouting, or even much courteous disagreement about the science. The two main things the experts disagree about (usually courteously) are how clear the science needs to be before taking action and which scientific findings people ought to care most about. These are not scientific questions. They’re not even questions that scientists are especially qualified to answer.

I’ll come back to this key point later. But first I want to say some more about the science.

The science summaries in the sections that follow are oversimplified. I’m way outside my field here, and “oversimplified” is as detailed as I dare to get. But I’m pretty sure these summaries are basically correct. Importantly, I haven’t needed to modify these summaries much since I first drafted them in late September. There are more studies, but they have mostly just confirmed what was already pretty much known. And what wasn’t known in late September is for the most part still not known as I write this final draft in late November.

5. Boosters are safe-ish.

As a risk communication consultant, I taught my clients never to say anything is “safe.” Better to say the data you have so far suggest that it’s safer than some quantitative criterion, or safer than something else, or safer than it used to be, or safe enough that you feel okay letting people to do it while you continue to collect data. Safe-ish, in short.

There was some concern early on that boosters might have worse short-term side effects (like chills or muscle aches) than the primary doses had. So far that turns out not to be true. There is still some concern that more serious but rare side effects (like myocarditis) might be less rare for boosters than for primary doses; they’re rare enough that we don’t have definite answers yet, though the evidence so far is encouraging. These uncertainties are relevant mostly for young people whose risk if they catch COVID is lowest, especially young males whose risk of vaccine-induced myocarditis is highest. Most already vaccinated people considering whether or not they want a booster shot have more reason to fear COVID than another vaccine dose.

It is always conceivable that some totally unanticipated, weird, horrible delayed effect might emerge in years to come that will make us all wish we hadn’t gone down the COVID booster road (or even the COVID vaccine road). But I don’t know of any vaccine expert who argues that boosters pose a big risk. Not one. The anti-booster argument was never that the booster risk is big. It was always that the booster benefit might be so small for some people that even a small risk could still outweigh that tiny benefit.

As September gave way to October and then November, the safety data on boosters kept getting stronger. Real-world evidence from Israel (the first country to institute a massive booster campaign) kept accumulating, as did evidence from smaller but more controlled studies initiated by the three vaccine manufacturers. The September VRBPAC and ACIP meetings about the Pfizer booster didn’t show any serious safety concern, but participants did keep saying they wished they had more robust safety data. By the October meetings about the Moderna and J&J boosters, participants sounded more confident that there weren’t going to be any unpleasant booster safety surprises.

6. Boosters give you more antibodies.

All three manufacturers of COVID vaccines approved for U.S. use have published data showing that a booster shot produces a huge increase in antibodies. The antibody count declines over time after vaccination (or after infection). The booster drives the count back up; in some cases it drives the count up higher than its previous top. The increases are not equal. Pfizer and Moderna boosters increase the antibody count far more than a J&J booster does, at least initially. But any booster, coming after any primary vaccination, sizably increases the antibody count. That’s what everybody in the field knows.

What nobody in the field knows yet: whether it matters. The “correlates of protection” of antibodies – the relationship between antibody count and illness – is unknown for COVID (and for many other diseases).

There are at least three relevant unknowns:

• How long will the antibodies last – will booster antibodies be any more durable than the antibodies produced by the primary vaccination, or will people keep needing boosters every six months or so?
• What antibody count is enough – does the higher antibody count resulting from a booster add significant protection, or is it more or less superfluous and close to useless?
• How will boosters affect longer-term elements of immunity like B cells and T cells that are central to preventing severe illness – are they boosted too, and if so for how long, and with what impact on outcomes?

Nobody thinks more antibodies from a booster shot are a bad thing. Nobody knows yet how much of a good thing they will turn out to be.

7. Boosters reduce mild infections significantly.

Everyone agrees that boosters reduce the incidence of breakthrough COVID infections – that is, infections in people who have already had their primary vaccine doses. Everyone agrees that breakthrough infections have been getting more common. Everyone agrees that breakthrough infections are usually mild or even asymptomatic, but can occasionally be severe or even deadly.

One huge area of disagreement: How important are breakthrough infections that aren’t severe enough to require hospitalization? Booster opponents have tended to claim that vaccines are “designed” to prevent severe threats to life, not mild inconveniences. So they think we should reserve boosters for people for whom the likelihood of a severe breakthrough infection is significantly reduced by a booster shot. Booster supporters, on the other hand, often argue that mild infections are not negligible. They add to human misery; they keep people away from their jobs (including healthcare workers and other essential frontline workers); they pose some threat of long COVID; they increase transmission and thus the total community burden of COVID.

There is a third school of thought: that mild cases are themselves a kind of booster, increasing natural immunity, which benefits both the individual and the society; and that mild cases should therefore be fostered, or at least not prevented. This viewpoint was widely excoriated early in the pandemic, especially when advanced by proponents of the so-called Great Barrington Declaration, which opposed lockdowns and other measures the signatories considered too extreme. It has become less unacceptable as experts and the public have come to accept that COVID is going to be endemic and we need to learn to live with it.

Even so, booster opponents seldom go so far as to claim that mild COVID cases are good for you or good for society. They do question whether mild cases are harmful enough to justify using boosters against them.

There are scientific findings relevant to this debate, but it is fundamentally not about science. Everyone is entitled to an opinion about how useful it is to reduce the odds that a vaccinated person will get a mild breakthrough case of COVID. Whether that’s a good enough reason to get a booster shot is an individual judgment. Whether it’s a good enough reason to launch a booster rollout or even to authorize booster shots at all is a public health judgment and a political judgment. Neither is purely or even mostly a scientific judgment.

8. Boosters may or may not reduce severe infections significantly.

There is disagreement among the experts about the value of reducing mild breakthrough infections. (As I have already noted, it isn’t fundamentally a scientific disagreement, so their expertise shouldn’t count for much.) But the experts all agree that booster shots accomplish this, at least temporarily.

The state of expert knowledge vis-à-vis severe infections is exactly the opposite. The experts all agree that reducing severe breakthrough infections – those requiring hospitalization, for example – is important. But they disagree about the extent to which booster shots are likely to do the job.

The extent to which boosters reduce severe breakthrough infections is tough to measure. Precisely because breakthrough infections are seldom severe, the “n” for hospitalized cases in booster studies is small even in the placebo group – too small to draw reliable conclusions.

Since they don’t have useful results from randomized controlled trials, the experts try to interpret observational data instead. So far they’re relying mostly on observational data from Israel, which was the first country to administer millions of boosters and thus the first to have real-world results to report.

Part of the reluctance to rely on observational data from Israel is that Israel is different from the U.S. in so many ways. Two are especially salient. First, Israel vaccinated its people earlier, so it’s getting earlier evidence that vaccine efficacy may be waning. The vaccine primary doses are working better in the U.S. so far; we have proportionally fewer breakthrough infections or severe breakthrough infections. Second, Israel succeeded in vaccinating a much higher percentage of its people. Breakthrough infections of vaccinated people are a bigger deal in Israel because Israel has proportionally fewer unvaccinated people. Breakthrough infections are Israel’s main COVID problem. In the U.S., by contrast, the pandemic is still mostly a pandemic of the unvaccinated. Our big unvaccinated problem makes our smaller breakthrough problem feel comparatively unimportant.

To oversimplify: Israel had one problem, breakthroughs, and solved it with boosters. The U.S. has two problems, breakthroughs and hesitancy. Our breakthrough problem so far is smaller than Israel’s was. Our hesitancy problem is much, much bigger than our breakthrough problem. So American experts haven’t seen much reason to focus on breakthrough infections, even severe ones.

Another difference that matters: Israel has relied primarily on the Pfizer vaccine. So its booster findings have limited relevance to decisions about the value of boosting people who received Moderna or J&J vaccines, or the value of using Moderna or J&J boosters.

Why should we think boosters will do much to prevent severe breakthrough infections? There are three possibilities:

(a)

Maybe boosters have little or no effect on severe breakthrough infections. T cells and B cells are thought to be our main protection against severe outcomes. Even if our antibodies have waned, maybe our T cells and B cells are sufficient without boosters. And whether they’re sufficient or not, maybe boosters don’t make them work any better.

(b)

Maybe boosters do reduce the likelihood of severe breakthrough infections, but the effect is hard to see because severe breakthrough infections are rare anyway. And for the same reason – they’re rare anyway – maybe reducing the likelihood of severe breakthrough infections is just not terribly important from a public health standpoint. How many millions of people do we want to give booster shots to just to protect a handful of people who would otherwise have contracted severe cases of COVID?

(c)

Maybe severe breakthrough infections will get a lot less rare in the months ahead, as we get further and further from the dates of people’s first vaccinations. If so, it may soon be clear that boosters add public health value. So maybe we should start giving people boosters now.

Only (c) is a strong public health case for prioritizing a booster rollout. There is already some evidence of (c) from Israel. That’s where elderly people were vaccinated earlier than anywhere else in the world, So maybe the finding there that boosters reduce the incidence of COVID hospitalization among the elderly is a forerunner of what Israel will soon see in younger populations and we will soon see in the U.S. as well. In fact, the data Israel reported in October showed more severe breakthrough infections among non-elderly people than the September data, and more benefit from booster shots to the non-elderly.

As we shall see, the new data from Israel changed the tone of the FDA and CDC advisory committees between their September meetings to consider Pfizer boosters and their October meetings to consider Moderna and J&J boosters. Even though the new Israeli data weren’t directly relevant to Moderna and J&J, they convinced some previously skeptical committee members that boosters might be useful even against severe outcomes.

On the other hand, there are methodological objections to the Israeli evidence that suggest maybe it doesn’t mean what it seems to mean.

The Israeli data kept coming as October gave way to November. The evidence mounted that severe breakthrough infections were increasing among all age groups, not just the elderly; and that boosters significantly reduced the incidence of severe breakthroughs, not just mild ones, and that too among all age groups.

But that’s Israel. Even in late November it remains a lot less clear to what extent and in what age groups severe breakthroughs are increasing in the U.S. And there’s no real-world U.S. data yet showing how much boosters affect severe breakthroughs. The companies’ studies clearly prove that the boosters reduce breakthroughs. But severe breakthroughs aren’t frequent in the U.S., at least not yet. The studies are too small to yield meaningful results about whether boosters make them even less frequent.

From the beginning of the booster debate, most experts agreed that if (c) is true we should start giving boosters to elderly people vaccinated near the beginning of the year, probably to other high-risk people as well, and quite possibly to everybody vaccinated more than six months ago. Even now, in late November, most experts agree that we can’t be sure whether (c) is true or not, and that we will know more in a few months.

The disagreement all along has been whether to start boosting while the data are unclear or to wait for clearer data – once again, not really a scientific disagreement.

To put the point more carefully: The disagreement has been about how strong the evidence needs to be that boosters reduce severe breakthrough cases to justify how broad a booster rollout. The evidence got stronger in October than it was in September, and stronger in November than it was in October. As the evidence got stronger, many experts started to migrate from anti-booster to pro-booster, and many not-so-expert public health professionals followed their lead. Some have stayed skeptical – not because they doubt that the evidence has gotten stronger, but because they don’t yet consider it strong enough.

I like the formulation in this October 14 MedPage Today article: “The bottom line: boosters give extra protection to everyone, but that protection may be really meaningful only to those over 65 and other vulnerable populations.” I agree. The extra protection everyone gets from boosters may not be “really meaningful” yet to young, healthy vaccinees who aren’t especially worried about mild infections. It may never become “really meaningful” to young, healthy vaccinees who aren’t especially worried about mild infections. But boosters give extra protection to everyone.

Given that boosters have a pretty benign safety profile, I think that’s reason enough to offer them to everyone. It’s reason enough to tell the most vulnerable groups that they really should get boosted now. And it’s reason enough to tell everyone else that if they decide not to get boosted now they really should watch closely for new evidence that might change their minds in the months ahead.

By late November, I think, that was the majority view among the experts too. It was not the majority view among the experts back in September and October.

When public health officials want people to take or tolerate some precaution that might or might not turn out to add value, the catchphrase they often deploy is “abundance of caution.” (They used that catchphrase a lot, for example, in support of their decision to pause the J&J vaccine rollout because of a rare blood clot side effect.) The ordinary term for the same concept is “better safe than sorry.” Both terms are grounded in the commonsense understanding that if a precaution has very little downside and an uncertain but potentially sizable upside, the rational thing to do is to take the precaution.

Perversely, this time it was the booster-seeking public demonstrating and advocating an abundance of caution. And far too many public health professionals decried the booster precaution, especially for healthy young people, because the evidence was still uncertain how useful it would turn out to be.

And what did they say about the people whose pre-authorization desire for a booster demonstrated an abundance of caution? “Finagling.” “Booster-mania.” “Mass hysteria.” “Unfounded panic.” “Gimme, gimme, gimme.” This was how they described their pro-vaccine home team!

9. Breakthrough over-reassurance was unsustainable.

Back in September, public health officials and experts were endlessly reassuring the public that breakthrough infections (even mild ones) were rare. From the outset, that was unwise risk communication, as overconfident over-reassurance always is. The over-reassurance was unsustainable. We all started seeing news stories about breakthrough infections among celebrities, and most of us ran into family members, friends, and acquaintances with breakthrough infections as well. “Rare” wasn’t cutting it.

To what extent were public health officials and experts genuinely overoptimistic about the ability of COVID vaccines to prevent breakthrough infections, and to what extent were they dishonest, intentionally over-reassuring the public in order to encourage people to get their primary doses? I can’t say. Some of each, surely. It’s not irrelevant that the CDC decided on May 1, 2021 to stop counting breakthrough infections other than ones leading to hospitalization or death. Maybe CDC officials thought a national tabulation of how many vaccinated people had caught COVID despite their vaccinations wasn’t useful for public health planning. Maybe they thought it was better for vaccination promotional campaigns not to know the answer.

The primary motive for understating the frequency of breakthrough infections wasn’t to reassure the vaccinated that they wouldn’t get a breakthrough. It was to encourage the vaccine-hesitant to get vaccinated too – by encouraging them to believe that just one shot (J&J – “one and done” ) or at most two (Moderna or Pfizer) would pretty much end the pandemic as far as they were concerned. That’s still one of the key motives for discouraging boosters now: If we start giving boosters to fully vaccinated people, booster opponents reason, unvaccinated people will see that as evidence that the vaccines “don’t work.”

There is nothing “scientific” about a strategic communication decision to understate the risk of a breakthrough infection and the benefit of a booster because you want to deter unvaccinated people from underestimating the value of getting an initial shot. The two relevant fields of expertise are ethics and risk communication – not virology or epidemiology or immunology or public health.

It is now accepted – though not aggressively communicated – that breakthrough infections are common and getting more so. They’re still mostly mild or asymptomatic. Severe breakthrough infections are still uncommon. But “still” is a tell. Whenever you see that word, you know the speaker is aware that it’s possible what is “still” true now may not stay true for long.

There’s a genuine dilemma here. The COVID pandemic isn’t over. The data we have tell us only what seems to be true so far. Predictions about what may be true in the months to come are tentative; sometimes they’re little more than guesswork. Will breakthrough infections stabilize? Or will they continue to slowly get more common? Or will vaccine protectiveness fall off a cliff we didn’t know was there, making breakthrough infections – and severe breakthrough infections – suddenly a big, big problem?

And given that we don’t know which of these scenarios is our future, should we base our policy decisions on the data we have, or on the likeliest future according to the majority of the experts, or on the worst case scenario in an abundance of caution?

As a risk communication consultant, my advice to clients is always to share the dilemma. Telling people that breakthroughs are rare no longer accurately conveys current reality. It never did justice to the uncertain future, or to the policymaking uncertainties that result.

10. The waning debate doesn’t matter.

Why are breakthrough infections getting more common? There are three main hypotheses:

• Because vaccine effectiveness is waning.
• Because the Delta variant is more transmissible than prior variants, so vaccinated people are encountering infected people more often, with more opportunities for their vaccine protection to fail.
• Because vaccinated people are taking fewer precautions (wearing masks less often, joining others indoors more often, etc.).

All three are probably true. How much each of the three contributes to the increase in breakthrough infections is debatable, and hotly debated.

But as far as the booster debate is concerned, this question is profoundly unimportant. It doesn’t matter why breakthrough infections have been getting more common. What’s important is that breakthrough infections have been getting more common. So the case for boosters has been getting stronger.

How much stronger depends partly on whether you think mild breakthrough infections matter, and partly on whether you’re worried that severe breakthrough infections may be getting more common too. It doesn’t depend on what you think about waning.

11. The science isn’t dispositive.

So let’s summarize where we’ve got to in the last few sections.

• With the possible exception of young males, boosters are almost universally considered safe enough that safety isn’t really an issue. The issue is how much benefit boosters deliver.
• We know boosters increase antibodies at least temporarily, but we don’t know how important that is.
• We know boosters reduce the number of mild breakthrough infections, and we know the number of mild breakthrough infections has been increasing – but we differ in how much value we put on fewer people getting mild COVID cases.
• We agree that fewer people getting severe COVID cases would certainly be valuable, but its public health value depends on how many severe COVID cases we’re talking about. So far, without boosters, there aren’t a lot of severe breakthrough cases, at least not in the U.S. But that number, too, might (or might not) increase substantially in the months ahead, because of waning vaccine efficacy, new variants, and/or reduced precaution-taking.
• And anyway, even if severe breakthrough cases threaten to become more common, we don’t know how effective boosters will be in reducing their frequency, and for how long. There is some encouraging evidence from Israel, but it’s still preliminary and not necessarily dispositive.
• We know we will know more about many of these unknowns in a few months. We disagree on the merits of waiting for better data versus rolling out a national booster campaign now versus an intermediate strategy of selectively boosting the most vulnerable people while we wait for better data.

All the tough questions are about values, not science. They’re about the importance of certain outcomes, like protecting people from mild breakthrough infections. And they’re about the dilemma of how to deal with uncertainty – when to act on uncertain information and when to wait for better data.

I keep repeating this point because it’s so crucial. Virtually no one argues that boosters are dangerous (except maybe for young males). Nor does anyone argue that boosters are useless. Expert opinion on the value of boosters ranges from “only a little useful, except for the most vulnerable” to “very useful for pretty much everybody.” That’s the whole scientific debate: Are boosters a little useful or very useful, for a few people or a lot of people? And even on that question, the experts pretty much agree that they don’t know the answer. They all know they’re all guessing.

So the scientific debate is how useful will boosters turn out to be, and for whom. And the two core underlying meta-debates aren’t about science at all:

• How useful, in what ways, would boosters have to be before it was worthwhile to make them available to people who want them? In particular, should we care much about preventing mild cases?
• Given that we don’t know how useful boosters will turn out to be, should we make them available now or wait till we know more?

In a September 28 Washington Post column, former Baltimore health commissioner Leana Wen summarized the COVID booster controversy this way: “The data themselves weren’t in contention. Rather, nearly the entire debate centered on broader value judgments that should not fall under the sole purview of federal scientific agencies.”

---

The Anti-Booster Case

12. What about the unvaccinated?

As I acknowledged at the start of this column, getting boosters into fully vaccinated people is way less important from a public health standpoint than getting first doses into fully unvaccinated people.

Of course a public health standpoint is not the only standpoint. If you’ve already been vaccinated, the most useful dose to you is a booster dose in your own arm.

And of course even public health priorities are a moving target. If enough people get vaccinated and enough vaccinated people start having severe breakthrough infections, boosters may eventually become public health’s top priority. That’s what happened in Israel. But for now, in the U.S., primary doses are public health’s top priority, and rightly so.

Booster opponents incessantly stress their view that the public health priority on first doses means it’s a bad idea to focus also on booster doses They have a point – a small one, I think – vis-à-vis the time and effort of local clinicians, pharmacists, etc. But multitasking has always been a public health reality. And they do not have a point vis-à-vis the doses themselves. The U.S. has access to enough vaccine to vaccinate as many holdouts as are willing to be vaccinated, and still roll out the robust booster program the White House envisioned.

Booster opponents also argue that a robust booster rollout might further discourage some of those holdouts – might even make the difference for some who were on the brink of finally rolling up their sleeves until the news about boosters changed their minds. There’s some survey data to support this: In several surveys a majority of unvaccinated Americans say boosters make them even less likely to get vaccinated. That’s a long way from saying they would probably get vaccinated if only they could be confident they ’d never need a booster. Even so, the possible impact of a booster rollout on vaccine hesitancy isn’t a stupid thing to worry about.

Many unvaccinated people are understandably inclined to say yes when asked whether they think anything is a good reason to stay unvaccinated. People will seize on whatever ammunition they can find to support a decision that’s widely derided by friends, family, doctors, the President of the United States, and late-night talk show hosts.

The obvious reason boosters are hypothesized to increase vaccine hesitancy is because some unvaccinated people are likely to see the groundswell of booster enthusiasm as evidence that the vaccines “don’t work.” On the other hand, the purpose of boosters is to reduce the incidence of breakthrough infections, especially severe ones. Wouldn’t unvaccinated people also be likely to see breakthrough infections as evidence that the vaccines “don’t work” – especially breakthrough infections that end in hospitalization or death? And sure enough, that’s what the survey data show. In one study, two-thirds of unvaccinated people said breakthrough infections are evidence that the vaccines aren’t working. A roughly equivalent 71% said the same thing about boosters.

Which one will turn out to be the ammunition people use to justify not getting vaccinated, breakthrough infections or boosters? Whichever one is happening. If the news is full of breakthrough infections, that will be a big reason for unvaccinated people to say the vaccines don’t work. If a successful booster campaign reduces the number of breakthrough infections, that will be a big reason for unvaccinated people to say the vaccines don’t work.

In principle, which is better antivax ammunition, breakthroughs or boosters? My guess is breakthroughs. Knowing people who got vaccinated and got the disease anyway is a vivid demonstration that the vaccine “didn’t work” for those people – much more vivid than knowing people who decided they’d better get another shot and so far they’re fine. Flu vaccination might be a useful precedent here. When people tell you why they’re not getting a flu shot, “I know somebody who got a flu shot and got the flu anyway” comes up a lot more often than “I know somebody who gets a flu shot every year.” To the extent that boosters reduce breakthrough infections, I’m betting they reduce vaccine hesitancy more than they increase it.

And how important a question should that be in the first place? What I find fascinating is how much anti-booster argumentation is devoted to the idea that boosters are bad because they will discourage unvaccinated people from rolling up their sleeves. Both the FDA’s VRBPAC and the CDC’s ACIP discussions made reference to this idea, and so do many anti-booster op-eds. I can’t think of another example where a sizable segment of the public health profession explicitly opposed a medical procedure that would benefit one group of people because some other group of people might misinterpret it in harmful ways.

For decades vaccine-hesitant parents have complained that the large number of recommended pediatric shots makes them skeptical about vaccination. Has the public health profession ever eliminated a pediatric shot it considered beneficial in order to diminish vaccine hesitancy?

The fact that most vaccinated people say they want boosters is all too readily dismissed as irrelevant. “We don’t just give people what they want. We decide what they need.” But the fact that most unvaccinated people say they would see boosters as yet another reason to stay unvaccinated is embraced as a persuasive rationale for not launching a broad-based booster campaign. “We don’t want to do anything to discourage them further.” Catering to the misperceptions of the unpersuaded while shrugging off the rational preferences of public health’s home team strikes me as more than a little weird.

Weird or not, many public health professionals mean it when they cite the unvaccinated as a reason not to launch a booster rollout. They don’t want to distract themselves from desperately trying to get unvaccinated people to change their minds. And they don’t want to do anything that might discourage unvaccinated people from changing their minds. For what it’s worth, that’s got nothing to do with “following the science.” It is one among many examples where experts who claim to be data-driven have diverged from the scientific question before them – which prospective booster recipients will benefit from a booster shot more than they are harmed – to focus instead on an extraneous question.

I have to add: an extraneous question about risk communication, way outside their area of expertise.

13. What about the rest of the world?

Another anti-booster argument I hear often is that it’s selfish to prioritize boosters shots for Americans when there are billions of people in poorer countries who have yet to be offered their first shot.

It is probably true that first shots in the third world will save more lives than booster shots here at home. Even this is debatable. Most likely because their populations skew young, developing countries have reported low COVID death rates despite inadequate healthcare systems. (In the U.S., similarly, an unvaccinated child is at much lower risk of dying of COVID than a fully vaccinated senior, so boosters for the elderly should arguably be a higher priority than pediatric vaccinations.) Even so, I’ll stipulate that on balance shots given as primary doses in poor countries probably save more lives than shots given as boosters in rich countries.

There’s also a self-interest argument for prioritizing vaccine philanthropy over domestic boosters: the risk of new variants emerging in developing countries where COVID is circulating widely. I’m not qualified to assess this argument, but I’m skeptical. Among the scariest new variants would be those that aren’t stopped by existing vaccines. Where is such a variant likeliest to emerge and flourish? Wouldn’t it be in a country with lots of transmission (so lots of chances to mutate) despite lots of vaccinated people (so selective pressure to mutate toward vaccine resistance) – a country like the United States?

In any case, the shots available for vaccine philanthropy are not the shots available for domestic boosters. When the U.S. orders vaccine doses from manufacturers, it signs contracts providing that it may not repurpose those doses for international use. Nor would most countries (even poor ones) accept delivery under those circumstances. The supply chain for U.S. vaccine philanthropy and the supply chain for U.S. domestic vaccination are separate from the start. Doses that have been ordered but not accepted might be repurposed, but doses that have already been accepted by the feds and shipped to the states almost certainly can’t be donated elsewhere. In most cases they can’t even be shipped to a different state. They will be used in situ or they will pass their expiration dates and be wasted.

Despite all that, boosters pose a genuine ethical dilemma for the U.S. government. If boosters become a big enough thing, sooner or later they will compete with vaccine philanthropy for new vaccine orders. For that reason the World Health Organization has inveighed against boosters. WHO Director-General Tedros Adhanom Ghebreyesus calls them an example of “vaccine injustice and vaccine nationalism.” Tedros has repeatedly urged rich countries to forgo boosters. I think it’s a political nonstarter in the countries he’s targeting, but it goes over well in the developing world, and maybe it succeeds in motivating rich-country governments to contribute more in other ways.

Vis-à-vis doses that are already in your doctor’s or your pharmacist’s hands, I can’t see any ethical dilemma when you’re considering whether to get a booster shot.

Even so, it may feel like an ethical dilemma. It obviously does to many public health professionals, some of whom have announced that they personally won’t get boosted until they are satisfied that international health equity has been achieved.

COVID vaccinations are not the only inequity in the world, or even the worst one. And unlike vaccine doses, many resources are fungible. Everyone who lives a comfortable first-world life has the option of selling everything to give to the poor – or simply living less high on the hog so we can increase our charitable contributions. We do not have the option of sending “our” booster dose to a developing country. We do, however, have a range of other, more realistic options, from lobbying on behalf of government vaccine philanthropy to increasing our own personal philanthropy.

When I was a child, my mother sometimes cajoled me into eating my dinner on the grounds that there were starving children in China. Many public health professionals now want me to forgo a third dose of vaccine on the grounds that there are unvaccinated people in Africa. In hindsight, my mother made more sense. Like my childhood dinners, U.S. vaccine booster doses will not be shipped overseas. It seems ethically better to consume them than to waste them.

In 1961, Kurt Vonnegut wrote a dystopian science fiction short story entitled “Harrison Bergeron,” in which no American is allowed to be smarter, stronger, more attractive, etc. than anyone else. Equality is enforced by a Handicapper General. The title character is a 14-year-old who refuses to be handicapped. I think fighting to get vaccine doses to developing countries is a worthwhile endeavor. But urging people to forgo a COVID booster until everyone in the world can be vaccinated strikes me as Harrison Bergeron public health.

14. Why is anyone anti-booster?

As should be clear by now, “the science” pretty clearly says boosters are likely to do more good than harm for anyone except possibly young males. The public and the White House are following the science when they champion universal booster availability. Scientists who oppose boosters are not.

To be sure, giving every fully vaccinated person a booster exactly six months after that person’s primary doses may not turn out to be optimal. Maybe nine months or twelve months or fifteen months is better, at least for young, healthy people with the most robust immune systems. On the other hand, maybe five or four months is better, especially for elderly people and others with less robust immune systems. Maybe we’ll all need multiple boosters, and we’ll figure out who can afford to space them out at longer intervals and who can’t.

In short, the time may come when we understand COVID dynamics enough to advise some people to wait longer than six months. But we’re not there yet. For now, the tentative state of scientific understanding calls for an abundance of caution: boosters sooner rather than later, for everybody or nearly everybody.

Aside from the question of why so many experts think preventing mild COVID isn’t worth a booster (but is worth a mask), the bigger question is why so many think we can afford to wait and see what happens to the rate of severe breakthrough cases. There may or may not be a future increase in severe breakthroughs, because of some mix of waning and other factors. Future waning in vaccine efficacy may grind to a halt or continue to be gradual – or it may speed up or fall off a cliff.

It takes time to roll out a successful booster campaign. The downside of boosting people who turn out not to need boosters looks like it will be very small, mostly a few additional myocarditis cases. The downside of not boosting people who turn out to need boosters might be huge – not just severe cases of COVID, but severe recriminations and a severe loss of trust in public health.

Public health professionals invariably urge vaccination against measles because if too many people are unvaccinated, measles might come roaring back. They don’t require evidence that measles is roaring back already (though they do sometimes falsely claim that it is). Similarly, public health professionals encourage flu shots without knowing whether the coming year will bring a bad flu season, a mild flu season, or indeed any flu season at all (though they do sometimes pretend to know). Suddenly with regard to COVID boosters they want evidence that getting boosted is important now. Suddenly the unknown but nontrivial possibility that it might suddenly become important X months from now is shrugged off.

If the science is really that clear, why are so many public health professionals anti-booster? I’ve already discussed some of the many reasons. Others will come up later. Here’s a quick list of what I consider the main ones:

• They’re trying to avoid sending a message to unvaccinated people that the vaccines “don’t work” (that is, don’t work so well that boosters aren’t needed).
• They’re trying to keep their limited resources (time, money, etc. – but not really the vaccine doses themselves) focused on promoting and delivering primary doses, not booster doses.
• They believe building a market for boosters will sooner or later diminish the vaccine supply donated to poorer countries, most of whose people have yet to get their first dose.
• They’re uncomfortable okaying boosters based on mushy, contradictory, incomplete data, preferring to wait till the evidentiary record is stronger. (Nixing boosters based on mushy, contradictory, incomplete data apparently feels less uncomfortable.)
• They’re uncomfortable okaying boosters based on data from real-world experience in other countries, as opposed to made-in-America randomized controlled trials. (Once again, nixing boosters on that basis feels less uncomfortable.)
• They’re not convinced that the pandemic crisis justifies departures from the agencies’ more slow-moving Standard Operating Procedure, at least not vis-à-vis boosters.
• They think preventing mild infections isn’t important enough to justify boosters, and they’re not convinced that boosters will significantly reduce the odds of a severe outcome for most people.
• They think it’s not good enough for probable booster benefits to exceed probable booster harms. They worry that even smallish harms might undermine vaccine confidence, so they want stronger evidence that the harms are negligible.
• They’re worried that if boosters get the official okay, affluent white people will avail themselves of the opportunity more enthusiastically or more successfully than others, exacerbating social inequity.
• They resent the Biden administration for leading instead of following the agencies and the experts, and for doing so publicly. They’re outraged that ordinary people not only watched the sausage getting made; ordinary people got to decide the recipe.

I have more sympathy for some of these rationales than I do for others.

I fully accept that everyone – even a scientist – is entitled to take nonscientific factors into account in reaching a judgment about what booster policy is best. What scientists are not entitled to do is to claim that those nonscientific factors add up to “The Science.” For the vast majority of the vaccinated public, the science says that boosters are likely to do more good than harm. Period.

---

The Sausage-Making Process

15. The ideal decision process as public health professionals see it.

As I mentioned at the outset, public health professionals like to hash out their dilemmas, doubts, and disagreements in private. They don’t like the public watching them make sausage.

But there’s something they like even less: when the public starts telling them what sort of sausage it wants them to make, and make right now!

Here’s the ideal sausage-making process as public health professionals see it:

Experts → Agencies → Politicians → Public

The experts figure out what public health policies they consider appropriate, hashing out their dilemmas, doubts, and disagreements and eventually reaching something approaching a consensus – at least enough of a consensus that the outliers are deterred from kicking up much of a fuss. Some of the key experts are officials at agencies like the FDA and the CDC, and at the various state and city public health agencies; others are on agency advisory committees or in less formal networks with agency staff. Thus the expert consensus gets communicated to the agencies, which launch the policymaking process needed to flesh out the details and implement what the experts have decided.

Political leaders are brought into the discussion, partly so they can contribute political expertise to the implementation plan, but mostly so they feel some buy-in and don’t belatedly dispute the expert consensus. Eventually the time comes to clue in the public, telling it just as much as it needs to know to encourage it to watch approvingly from the sidelines as the policy is implemented.

This process is often summarized as “Follow the science.” But I have stressed enough already that many of the key dilemmas, doubts, and disagreements aren’t really scientific. Many of the policymaking decisions, therefore, aren’t really scientific either. A better summary than “Follow the science” would be “Follow the scientists.”

Scientists are understandably but culpably prone to equate whatever scientists think or do with “the science.” Here’s an extreme example. President Biden’s chief medical advisor Anthony Fauci famously said in June that “people who then criticize me about that are actually criticizing science.” The “that” Fauci was talking about wasn’t his messaging on boosters. (He’s an early and strong booster supporter.) It was his messaging on masks. His point was that his mask messaging morphed suddenly from “you shouldn’t wear them” to “you must wear them” simply because the science changed. It would be an arguable position – except that Fauci had already acknowledged that his main reason for telling people not to wear masks in the early days of the pandemic wasn’t that he believed at the time that masks wouldn’t do them any good; rather, he was worried about a mask shortage and wanted to preserve the available supply for healthcare workers.

Fauci was being criticized for telling a “noble lie,” not for following the science. He admitted the noble lie. And then he insisted that anyone who criticized it was “actually criticizing science.” When scientists lie for a good reason, Fauci seems to believe, that’s part of the science.

16. The actual booster decision process (part one – the public).

From the perspective of most public health professionals, the sausage-making process for COVID boosters was exactly backwards from the ideal. Here’s how it went:

Public → Politicians → Agencies → Experts

The distinct possibility that COVID vaccines would require boosters sooner or later had been in the air from the outset. Many other vaccines require boosters. The COVID vaccines in particular were developed and tested under emergency conditions. There wasn’t time to find out how durable the vaccines were, and there certainly wasn’t time to test whether different doses or different intervals between doses might yield more durable immunity. There was simply no way to guess in advance whether and when boosters might be needed.

The public health profession’s dominant early messaging wasn’t about boosters, of course. The core message was that the vaccines are incredibly safe and even more incredibly effective, far more effective than anyone had dared hope. Even so, the experts knew and sometimes said that a protocol calling for just one shot or two closely spaced shots wasn’t likely to be sufficient, that we might need more doses over time as our immunity waned and/or the virus mutated. So the frame from the outset was: (a) incredibly safe; (b) incredibly effective; and (c) by the way, watch for breakthrough infections and a need for boosters.

Boosters didn’t start becoming a major public controversy until July 8, when Pfizer went public about its evidence of waning antibodies and its plan to apply to the FDA for a booster Emergency Use Authorization. The FDA and the CDC immediately pushed back with a rare joint statement asserting unequivocally that “Americans who have been fully vaccinated do not need a booster shot at this time” – as if that meant that boosters wouldn’t be useful (as opposed to “needed” ) soon (as opposed to right now). It was clear that the two agencies didn’t want to talk about boosters, and didn’t want Pfizer to talk about boosters either.

That was shortly before Delta swept across the U.S., overwhelming hospitals just as governments were easing restrictions and people were dropping precautions. By mid-August the U.S. was averaging more than 100,000 new cases a day, roughly a tenfold increase from the first week of June.

Together with my wife and colleague Jody Lanard, I made four unsuccessful tries to interest various reporters in covering the booster risk communication story. We converted our final attempt into a July 23 website column, outlining our main objections to the way the agencies and the public health profession generally were addressing the COVID-19 booster issue. Predictably, the column didn’t attract much interest either.

Though there wasn’t a lot about boosters in the media yet, the public itself was beginning to pay attention. Despite official reassurances to the contrary, it was hard to miss the growing number of breakthrough infections of fully vaccinated people, whether celebrities or acquaintances.

The early messaging that breakthrough infections were incredibly rare wasn’t false. It became false as breakthrough infections became more common. Pretty soon we all started knowing people who had them, and seeing media stories about famous people who had them – way too many people for the “incredibly rare” message to be sustainable. So the message changed to breakthrough infections are almost always mild, so they’re no biggie.

Even as they reassured the public that breakthrough infections weren’t a serious threat, experts were quietly debating the pros and cons of boosters on Twitter. The debate wasn’t mostly about the pros and cons of boosters for all – more about the pros and cons of boosters for themselves. Many had been early vaccinees; many had high-risk occupational exposures to COVID patients. Some had already arranged unsanctioned boosters for themselves and more admitted to thinking about it.

But the public stance of the FDA, the CDC, and the public health profession remained firm: The vaccines are working incredibly well and right now nobody needs a booster.

There were occasional exceptions. One stunning June 30 exception was a podcast interview with CDC head Rochelle Walensky, in which she conceded that lots of people were finding ways to get boosters, even mix-and-match boosters, and that they weren’t necessaily unwise to do so. I totally missed this interview at the time, and wasn’t aware of it until David Leonhardt of the New York Times discussed it at length in his October 21 Morning Newsletter. Walensky was asked what she would advise a friend who had gotten the one-dose J&J vaccine and was now feeling tempted to get a Pfizer or Moderna booster. Would she say her friend was making a huge mistake? “Not with what I’ve seen so far,” Walensky replied. She described the current official CDC position that one J&J dose is protective enough, then went on to agree with her hypothetical friend that a booster, especially a Pfizer or Moderna booster, might add value and is unlikely to do any harm.

Leonhardt’s reaction is worth quoting:

I found her comments both jarring and refreshing at the time, because they reminded me of how officials and experts often speak when talking to friends and relatives (or when talking off-the-record with journalists).

Walensky was being transparent. She was treating her audience like adults capable of handling subtlety. She was … helping them think through the unavoidably uncertain options. Health officials are frequently unwilling to [do that] in public…. In many of their minds, an extremely cautious approach is the right one. Until they know something with near certainty, they don’t talk about it publicly. When asked about an issue involving uncertainty, they tend to duck the question and restate official policy.

But this approach creates its own problems. During a crisis, people have to make uncertain decisions. Pretending otherwise doesn’t eliminate those decisions. If anything, the refusal to acknowledge uncertainty can undermine officials’ credibility….

Imagine how things might be different if officials were more straightforward with the public.

That was on June 30, a week before Pfizer applied for a booster Emergency Use Authorization and six weeks before President Biden announced his support for universal booster access. But it was a one-off. It was distinctly not a signal that Walensky and the rest of the public health establishment were prepared to give boosters their official endorsement, or for that matter prepared to acknowledge officially that boosters might be especially important for recipients of the one-dose J&J vaccine.

The vaccinated public – especially the enthusiastically vaccinated public – had been successfully taught that COVID vaccines are safe and effective. They didn’t necessarily think “effective” meant permanent. Most were familiar with the booster phenomenon from their childhoods and parenthoods; lots of safe, effective vaccines need boosters. So when they started to see chinks in the COVID vaccines’ effectiveness, their minds naturally turned to boosters. Even before Pfizer’s July 8 announcement, some had started agitating for COVID boosters, while others simply went about getting themselves boosted.

By August 10, a confidential CDC estimate had been leaked to ABC News that 1.1 million U.S. residents had already received unauthorized boosters. Jody and I were two of them.

17. The actual booster decision process (part two – the White House).

President Biden and his White House advisors took notice of the growing public interest in boosters. They not only decided to integrate boosters into their pandemic battle plan; they decided to say so publicly.

On August 18 the White House announced the plan, under which “a booster would be administered, eight months after an individual’s second dose, beginning the week of September 20.” The goal of the announcement was partly to increase public support for boosters among those who admired the president; partly to increase public support for the president among those who wanted boosters; and partly to forewarn the organizations that would be charged with actually delivering booster doses to get ready.

The goal was also, obviously, partly to put some pressure on the FDA and the CDC to hurry up and authorize boosters. “You’ve got one month to push the booster plan through your agency’s sausage-making factory and out the other end” was the announcement’s perfectly clear implicit message to the two agencies.

In the days that followed the August 18 announcement, Biden flirted with the possibility of authorizing boosters as early as five months after people’s primary doses instead of eight months after. The quickly reached compromise was six months after.

From the outset, the White House was explicit that its plan was contingent on okays from the FDA and the CDC, leaving open (though not so explicitly) the possibility that the agencies might okay only a smaller or slower booster rollout than the one the plan called for. “A lot of people were saying, ‘We’re making a decision,’” Fauci told the Wall Street Journal on September 24. “We weren’t. We were making a plan.” Even so, the White House was being disingenuous when it claimed the plan was just a way to get states and pharmacies ready to roll as soon as the agencies gave the go-ahead. It was that. It was also a way to pressure the agencies to give that go-ahead, and to do so as quickly as they could.

Fauci and the rest of the top scientific and public health leadership of the federal government were all publicly onboard with this White House plan, including the heads of both the FDA and the CDC – no doubt with varying degrees of enthusiasm or perhaps even suppressed disapproval. Their public support was another unmissable sort of pressure on their underlings: “Your boss has already signed onto this. Now you get to decide – quickly – whether you’re going to agree with your boss or undermine your boss.”

The pressure had the desired effect. The FDA and CDC bureaucracies moved faster than they would otherwise have moved, and they moved in the direction the White House and the public wanted. The White House had promised action the week of September 20, contingent on the FDA’s authorization and the CDC’s recommendation. The agencies said such a short deadline was unrealistic and unachievable. Then they achieved it.

But only for Pfizer vaccinees. Before anybody could get a booster, the FDA had to issue a license or an Emergency Use Authorization, and then the CDC had to issue a recommendation. But first, before any of that could happen, the vaccine manufacturer had to submit a formal application and supporting data to the FDA. Pfizer was weeks ahead of Moderna and J&J in meeting this essential requirement. So the process did go forward on the accelerated timetable the White House had wanted. But it was only to say yea or nay to Pfizer’s proposal to give Pfizer boosters to Pfizer vaccinees. Decisions on Moderna and J&J would have to wait – though not for long.

The FDA and the CDC convened their respective advisory committees, and then issued their respective recommendations – all in quick succession in advance of the White House’s “deadline” of a rollout to start the week of September 20. I’ll have more to say about how that happened in the next section. Here’s the outcome, what they decided vis-à-vis Pfizer boosters for Pfizer vaccinees:

• If you were 65 or older, or if you lived in a nursing home, a booster was recommended.
• A booster was also recommended if you were 50 or older and had a high-risk medical condition, one that made you likelier than others to get severely ill if you were infected. There are various lists of which preexisting conditions might be considered “high-risk” vis-à-vis COVID, some very broad and some quite narrow. The CDC list deployed for booster eligibility was broad, and got broader over time. And it was explicitly illustrative, not all-inclusive. As a practical matter, if you thought you had a high-risk medical condition, then insofar as booster eligibility was concerned you had a high-risk medical condition.
• If you were 18 to 49 and had a high-risk medical condition, a booster was permitted but not necessarily recommended. It was up to you and your doctor (if you consulted a doctor). Whether you actually had a high-risk medical condition was also up to you and your doctor (if you consulted a doctor).
• A booster was also permitted but not necessarily recommended if you were 18 to 64 and your working or living situation made you more exposed to COVID than others, and therefore likelier to get infected. Again there was no list specified of which occupations were sufficiently high-exposure to be covered. There were also disagreements about what sorts of high-exposure living situations were covered. The obvious examples were institutional ones like prisons, nursing homes, and homeless shelters. Less obvious ones might include crowded apartments or even crowded neighborhoods … or just hanging out a lot in crowded bars. Basically, if you thought something about your life put you at high risk of getting infected, you were entitled to a booster.
• It was all on the honor system anyway. The eligibility criteria were broad enough and vague enough that almost anybody could honorably decide, “Yes, I’m covered.” But if you thought you weren’t covered, you could lie.

The September eligibility criteria for Pfizer boosters added up to agency face-saving. Instead of just saying “Pfizer boosters are approved for anyone who was vaccinated more than six months ago,” the FDA and CDC established criteria that with a little creativity would be open to anyone who was vaccinated (with Pfizer’s vaccine) more than six months ago, but still looked and felt like they were based on evidence about risk. They even looked and felt like the agencies weren’t rubber-stamping the White House plan but rather insisting on modifications.

Appearances aside, the White House got what it wanted on its proposed schedule for Pfizer boosters, and looked overwhelmingly likely to get what it wanted for Moderna and J&J boosters in a matter of weeks. Boosters started officially going into arms late the week of September 20. Nobody seriously claims that would have happened without White House pressure.

18. The actual booster decision process (part three – Pfizer).

But the White House paid a predictable price for its decision to “put the cart well out in front of the horse” (as one online medical news site phrased it, precisely reflecting how most public health professionals felt about it). The decision irritated some agency senior officials and no doubt some rank-and-file as well. It seems to have irritated experts and public health professionals outside the agencies even more, or maybe they just felt freer to express their irritation. The angry op-eds that greeted the August 18 White House announcement were mild compared to the firestorm of even angrier tweets.

A fundamental principle in my approach to risk communication is that outrage influences hazard perception more than hazard perception influences outrage. The more upset we get about a risk, the more dangerous we believe it to be. A broader version of this principle provides that our opinions about a substantive disagreement tend to get more extreme as we get angrier at the people we disagree with.

So even as the agencies moved toward authorizing and recommending boosters (more precipitously than they wanted to, though more sluggishly than the White House wanted them to), outside expert opposition to the emerging booster policy became more vociferous. Before the president and the White House came out publicly for boosters, the experts’ booster debate was fairly low-key. Once it was fueled by outrage at the effrontery of presidential pressure, substantive booster opposition got shriller.

The process was both symbolized and aggravated by the August 31 announcement that two top FDA vaccine officials were going to resign, reportedly in protest against the White House pressure campaign. Dr. Marion Gruber, the director of the FDA’s Office of Vaccines Research and Review, and her deputy, Dr. Philip Krause, said they planned to leave the FDA in the fall. Shortly after their resignations were announced, but while they were still on the job, the two were among the coauthors of a September 13 article in The Lancet, a top medical journal, arguing the case against boosters. Their bottom line: “[C]urrently available evidence does not show the need for widespread use of booster vaccination in populations that have received an effective primary vaccination regimen.”

But the handwriting was on the wall.

Pfizer’s application asked the FDA to approve boosters for everyone 16 and older. On September 17, the FDA’s advisory committee (VRBPAC) turned that down flat. It voted instead to authorize boosters only for elderly Pfizer vaccinees and those with high-risk medical conditions. That would have been seen as a substantial rebuff to the Biden White House. It would have put the FDA in an impossibly awkward position, torn between the boosters-for-all stance of the president and its own top brass and the boosters-for-only-the-most-vulnerable position of its advisory committee.

VRBPAC was about to adjourn when a top FDA official, Peter Marks (Gruber and Krause’s boss and a booster proponent), warned the committee that the agency might well decide to “tweak” the authorization to include also people whose occupational or institutional settings made them likelier than most to get infected – for example, healthcare workers, prisoners, grocery clerks, maybe even teachers. At Marks’s urging, the committee took what amounted to a straw poll. It belatedly, quickly, almost casually – and unanimously – endorsed this huge increase in who could get a booster.

So now no tweaks were needed, and the FDA could give the White House a lot more of what it wanted and still accurately say it was following the advice of its advisory committee.

What might be termed the “Marks Afterthought” was so much a postscript that many news stories missed it, mistakenly describing the FDA’s booster authorization as going way beyond VRBPAC’s recommendations – which would have been true if not for Marks’s last-minute intervention.

Two days later CDC director Rochelle Walensky urged that agency’s advisory committee (ACIP) to follow suit – to recommend Pfizer boosters for everyone for whom the FDA had authorized Pfizer boosters. It did so unanimously vis-à-vis the elderly. But that was the ACIP’s only unanimous vote that day. Two of the 15 panelists voted against boosters for people 50 and over with high-risk medical conditions. The vote in favor of boosters for people 18–49 with high-risk medical conditions was a squeaker, 9–6. And after a long discussion the ACIP voted 9–6 against recommending boosters for people whose working or living settings exposed them to high risk of infection – with some ACIP members noting accurately that that was tantamount to everybody.

In the wee hours of the following morning, Walensky sided with the minority in that final ACIP vote, thereby aligning the CDC’s recommendations with the FDA’s authorizations … and with the White House’s and the public’s preferences.

Once the FDA and CDC had given the White House most of what it wanted, at least vis-à-vis Pfizer, non-agency experts and public health professionals began to fall in line. Some who had been pro-booster from the outset voiced their agreement with the outcome despite their disapproval of the White House pressure – as if the outcome would have been achievable without White House pressure. Others who had staked out a position that boosters weren’t needed yet swallowed their pride and went grudgingly along, grateful for the agencies’ face-savingly detailed eligibility specs. Still others remained disapproving, though I suspect their disapproval of the process (the White House pressure) was more deeply felt than their disapproval of the outcome.

Still, there was plenty of disapproval of the outcome. In a revealing September 24 op-ed in the New York Times, Megan L. Ranney and Jeremy Samuel Faust complained that the eligibility criteria were “premature and too vague.” Ranney and Faust were okay with the CDC recommending boosters for people over 65 and those over 50 with comorbidities. But they were highly critical of the CDC for “expanding booster eligibility to a huge swath of the population” – younger adults with comorbidities and anyone with a high-risk working or living situation.

Their most fervent complaint: The CDC didn’t recommend for or against boosters for these latter groups; it merely said they “may” get boosters “based on their individual benefits and risks.” Those permissive eligibility categories, said Ranney and Faust, “go well beyond the data and foist the decision of appropriateness onto individuals and their doctors.” How can we expect the public to follow the science, they seemed to be asking, when the scientific agencies default on their obligation to lead?

19. The actual booster decision process (part four – Moderna).

Doling out Pfizer boosters while stiffing Moderna and J&J vaccinees was obviously untenable. Even booster critics saw that. Moderna and J&J quickly submitted their applications and supporting data. The FDA and CDC advisory committees (VRBPAC and the ACIP) quickly scheduled October meetings to consider the evidence. Decisions were more or less promised before Halloween.

The Pfizer booster authorization and recommendation decisions set the pattern for Moderna. The two vaccines are similar, with similar efficacy and similar side effect profiles. Both rely on an mRNA platform. Moderna had already submitted an application to the FDA, requesting a booster Emergency Use Authorization for everybody whose primary Moderna doses were more than six months before. After the Pfizer decisions, it amended its application to ask for what Pfizer got.

On October 14, VRBPAC voted unanimously to give Moderna what Pfizer got.

The case for Moderna boosters was a bit weaker than the case for Pfizer boosters in several ways. The Moderna primary doses worked a bit better than the Pfizer primary doses, and seemed to wane a bit less. Pfizer had presented some data suggesting that its vaccine might be becoming less effective against severe outcomes. Moderna had data showing diminishing effectiveness against infection, but not against severe outcomes. And Pfizer also had real-world evidence from Israel going for it; Israel hadn’t used the Moderna vaccine, so the Israeli results were less demonstrably relevant.

Nonetheless, the tone of the October 14 VRBPAC Moderna meeting was a lot less conflicted about boosters than at its September 17 meeting to consider Pfizer boosters. Partly that’s because additional data from Israel had emerged in the intervening month, more clearly indicating that vaccine efficacy was declining and boosters were reversing the decline. Even though the new evidence concerned Pfizer doses, not Moderna doses, some previously booster-skeptical VRBPAC members clearly took it to heart. Partly too, I think, the additional month gave some VRBPAC members time to get over their outrage at the Biden administration for launching a cart-before-the-horse process.

But mostly what changed is that Pfizer boosters had already been through the sausage-making process and were now going into arms with the FDA’s and the CDC’s official blessing. Several VRBPAC members commented that they really had no choice but to do for Moderna what they had done for Pfizer. VRBPAC didn’t even consider recommending a booster policy for Moderna that was more restrictive than the Pfizer booster policy already in effect.

But it wasn’t about to get any looser either. The committee discussed the possibility of lowering the age at which all Moderna vaccinees (and presumably also all Pfizer vaccinees) could be authorized to get boosters without going through the “high risk” rigmarole. Instead of 65+, how about 50+, or 40+, or even all the way down to 18+? Senior FDA official Peter Marks sounded pretty enthusiastic about such a possibility – which would have more explicitly given the White House more of what it wanted. So did VRBPAC’s chair, Arnold Monto, who kept pointing out how much less confusing an age-based standard was than a risk-based standard.

Marks and Monto explained several times each (as did the CDC’s Amanda Cohn) that the CDC would be free to make narrower booster recommendations than what the FDA authorized, but it couldn’t go any broader than what the FDA authorized – implicitly an argument for broadening the FDA authorization and letting the CDC do the dirty work of promulgating narrower recommendations.

Several other VRBPAC members said things that sounded like they might be on Marks’s and Monto’s side. Listening to the discussion in real time, I thought for a while that the tide had turned – that VRBPAC would recommend and the FDA would decide that boosters were authorized for all Pfizer and Moderna vaccinees over 40. Not 18, I thought, but 40 looked doable.

Then it became clear that other VRBPAC members, probably a majority, would have voted no. They were okay sticking with the broad, porous, easily gamed risk-based eligibility criteria they had previously okayed for Pfizer boosters – more comfortably okay with them now than they had been a month earlier when they voted for them. But they didn’t want to go any further down the slippery slope toward an explicit “boosters for everybody” policy. Marks conceded that he had “heard pretty loud and clearly that there was not a lot of appetite for moving down the age range very significantly, if at all.” He wisely chose not to ask for a vote on the question.

Still, booster policy was obviously a moving target. VRBPAC sounded significantly less skeptical about boosters at its October 14 Moderna meeting than it had at its September 17 Pfizer meeting. Maybe that’s partly because VRBPAC members were getting over their irritation at the White House for preempting expert prerogatives. Surely it’s partly because the evidence of waning, of breakthrough infections, and of booster effectiveness had improved.

I wrote this sentence a few days after the October 14 VRBPAC meeting: “A couple of months from now, if the irritation has faded further and the evidence has improved further, it wouldn’t surprise me to see the FDA authorize boosters for everyone over 40, quite possibly with unanimous VRBPAC approval.” As it turned out, just over one month later the FDA authorized boosters for everyone over 18, and it did so without even bothering to convene a VRBPAC meeting it knew would vote yes. But I’m getting ahead of my story.

The FDA had set two technical bars it said boosters had to clear to be considered effective. The data that Moderna presented to VRBPAC showed it met only one of the two. As the FDA’s Moderna booster briefing detailed, “the difference in seroresponse rates … did not meet the pre-specified immunobridging success criterion.” Moderna’s presenter explained why that didn’t matter, the FDA urged VRBPAC to vote yes anyway, and VRBPAC did. If asked, I guess the FDA could explain that there’s a difference between following its own pre-established standard and following the science.

The headline on an article in MedPage Today perfectly summarized the October 14 VRBPAC Moderna booster meeting: “Moderna Booster for Certain Populations Sails Through FDA Panel – All thumbs up, despite small sample size, data quality issues.”

20. The actual booster decision process (part five – J&J).

The next day, October 15, VRBPAC considered the J&J booster application. That one got complicated.

Remember, the J&J primary vaccine protocol requires only one shot, compared to two for Pfizer and Moderna. “One-and-done” was expected to be a major selling point for J&J, so the company was in no hurry to complete a two-dose trial and apply for a two-dose Emergency Use Authorization. But for a variety of reasons (manufacturing problems, a blood clotting side effect, etc.), uptake of the J&J vaccine was disappointing. Worse, the J&J vaccine showed significantly lower efficacy than the Pfizer and Moderna vaccines – probably not because it was a less effective vaccine but simply because vaccinees were getting only one shot instead of two.

So J&J asked the FDA to consider two options. The first option: Authorize a second dose, to be given at least two months after the first. It was debatable whether that ought to be called a booster at all. Was it really a second primary dose, finally turning the J&J vaccine into a two-dose vaccine like its competitors? J&J wanted it called a booster, so the company wouldn’t have to abandon its one-and-done marketing strategy altogether.

The second option: Authorize a booster just like the Pfizer and Moderna boosters, to be given at least six months after a vaccinee’s primary dose.

J&J submitted plenty of data showing that a second dose at least two months after the first was a good idea – and probably would have been a good idea all along. It offered a lot less data supporting its booster-after-six-months proposal; one of the studies it submitted looked at only 17 people.

But the elephant in the room was the comparatively low efficacy of the J&J vaccine (or, rather, of the one-primary-dose J&J protocol). It was clear to the VRBPAC members that all J&J vaccinees needed a second dose ASAP.

Yes, “needed.” The VRBPAC discussion returned again and again to the fact that J&J vaccinees’ one-dose vaccine was providing significantly less protection than their neighbors’ two-dose vaccines.

And yes, “all.” VRBPAC voted unanimously that the FDA should authorize a second dose for every J&J vaccinee. This time there was no mention of age limitations or preexisting medical conditions or working/living situations – no eligibility criteria referencing high risk. If all you had going for you was a single J&J dose, VRBPAC implied, you were already high-risk, not really “fully vaccinated,” and you needed that second dose.

And yes, “ASAP.” Pfizer and Moderna vaccinees had to wait at least six months after their primary doses before they were eligible for boosters. For J&J vaccinees, VRBPAC supported a minimum interval of just two months. Since the vast majority of J&J vaccinees had been vaccinated far longer ago than that, the two-month minimum amounted to right now.

VRBPAC members were comfortable commenting at length about the inferiority of one J&J dose to two Pfizer or Moderna doses. But it was awkward for the FDA to state or even imply that there was something wrong with the J&J vaccine or its one-dose protocol – exactly what the FDA had authorized. That made it hard for the FDA to explain why it was so much more enthusiastic about J&J second shots than it was about Pfizer and Moderna third shots. Mostly it didn’t try. It just said all J&J vaccinees really ought to get another dose ASAP, and let the media explain why with quotes from outside experts, never from the FDA itself.

The way J&J and the FDA had framed the question, the second J&J dose at least two months after the first was called a booster. So it was called a booster in the unanimous VRBPAC vote. But a number of VRBPAC members said they thought it was really a much-needed second primary dose, and urged the FDA to call it an “additional dose,” not a “booster dose” in the Emergency Use Authorization. Marks said the agency would take that under advisement. In the end, “booster” language prevailed.

A question I haven’t seen answered is whether a J&J vaccinee who gets a second shot after at least two months will be eligible for a third six months after that. Especially if your second shot is another J&J dose, maybe it should be considered part of your primary series even if they’re calling it a booster, and you should be eligible for a real booster six months later. Given that J&J vaccinees weren’t authorized for second shots until October, the agencies have till April 2022 to figure that one out.

As for the “other” J&J booster, the one to be given at least six months after vaccinees’ primary doses, VRBPAC declined to vote on it at all. J&J didn’t make a compelling case that waiting six months worked significantly better than waiting two months – and most J&J vaccinees had had their primary doses more than six months ago anyway.

After the J&J discussion, the VRBPAC meeting segued to a different topic: mix-and-match boosters. It reviewed data from the National Institutes of Health that showed equal or better outcomes in terms of antibody counts when people got a mix-and-match booster than when they got a booster with the same vaccine as their primary doses. Every possible combination of primary and booster doses worked well – but overall the mix-and-match combinations worked better. The biggest difference by far: People whose primary dose was J&J got a hugely greater antibody boost from a Pfizer or Moderna booster than from a J&J booster.

The result was a lot of VRBPAC support for total mix-and-match freedom: authorizing any of the three boosters regardless of what primary doses vaccinees had received. Marks said the FDA would take that under advisement too.

21. The actual booster decision process (part six – mix-and-match).

After the J&J/mix-and-match VRBPAC meeting ended, I was pretty sure the FDA would quickly go along with the unanimous VRBPAC vote to authorize a second J&J dose for every J&J vaccinee.

But I doubted there would be quick action on mix-and-match boosters. The NIH mix-and-match study wasn’t a randomized controlled trial; the samples were tiny; the only outcome variable was short-term antibody counts, not actual illnesses or even infections. It was possible that longer-term studies of more than just antibody counts might yield very different conclusions. NIH emphasized that the data were preliminary. Also, the NIH study had tested full-dose boosters of the Moderna vaccine; Moderna had proposed half-dose boosters instead to reduce short-term side effects, and that was what the FDA was about to authorize.

I knew that in all probability most people would eventually get COVID vaccine primary doses and COVID vaccine boosters without giving a thought to which brand they were getting – as is already the case for flu vaccines and most other vaccines. But I thought the FDA would postpone any decision on mix-and-match till it had better data, rather than follow such preliminary science.

I was wrong. The J&J/mix-and-match meeting was on a Friday. By Monday the New York Times had an anonymously sourced story that the FDA would authorize mix-and-match within a day or two, at more or less the same time as it would authorize Moderna boosters and J&J second shots. Pretty soon everybody else had the story too.

Certainly one main reason for this unaccustomed FDA speed was the evidence that J&J vaccinees would be significantly better off with a Pfizer or Moderna booster than with a second J&J dose. That evidence was publicly available, much discussed in the media, and obviously especially salient to J&J vaccinees. For the FDA to authorize the second J&J shot and not authorize mix-and-match boosters for J&J vaccinees would infuriate J&J vaccinees. It would invite thoroughly justified criticism that FDA stodginess was endangering people’s lives. On the other hand, for the FDA to authorize mix-and-match boosters for J&J vaccinees but not anybody else would underline far too clearly the inferiority of the J&J vaccine. FDA officials really didn’t want to say outright what everybody knew and had known for months: that at least so far the J&J vaccine with its one-primary-dose protocol was looking significantly less protective than its two-primary-dose mRNA competitors.

And in fairness there were other good reasons to authorize all three boosters for vaccinees with all three primary doses. It would arguably make the messaging simpler: Everybody can get any booster. (Some did advance a contrary argument that “the wishy-washiness of Pick whatever is confusing as hell.” ) It would certainly make implementation easier. A pharmacy or other vaccination venue wouldn’t necessarily have to stock all three vaccines, and when people came in for a booster it wouldn’t have to check which vaccine their primary doses were. If people had had a bad reaction to a primary dose – or had (or felt they had) some other reason to want to switch, they would be free to switch. Or if they didn’t remember which vaccine they got previously, they would be free not to care.

The FDA authorized mix-and-match, but it didn’t take any position on who should probably switch vaccines and who would be better off sticking with the one they started with. If anything, the agency seemed to lean slightly toward sticking with the one you started with. Acting FDA head Janet Woodcock said on a call with reporters that the agency “would expect many people will continue to get the same series that they already received.”

That’s an admirably simple message. But did it follow the science? Did it even explain the science? Yes, the science on mix-and-match is still preliminary, but as public health professionals endlessly proclaim in other contexts, following the science means following what the science says so far. And the science so far surely says that J&J vaccinees would be wise to get Pfizer or Moderna boosters, not J&J boosters. The best I can say for the FDA’s mix-and-match messaging is that it allowed vaccinees and vaccinators to follow the science. It didn’t actually tell them what the science says.

Most VRBPAC members pretty clearly thought J&J vaccinees really should switch. Not saying so was in keeping with the tentativeness of the preliminary NIH data. It was in keeping with the FDA’s conventional role of deciding what’s permitted without actually recommending anything. And it was certainly better than advising everybody not to switch. But it was probably way short of optimum for J&J vaccinees. I have to wonder if FDA leadership was actually just fine with J&J vaccinees getting J&J second shots, or if it was quietly hoping they would learn from other sources to choose a different booster.

As I have throughout the pandemic, I wish U.S. COVID policymakers would accept the risk communication principle to acknowledge uncertainty candidly – even to proclaim it unmistakably. “The science is still preliminary and uncertain, but so far it looks like people who got the J&J vaccine are the ones who most need a booster, and it looks like the most effective booster for them is Pfizer or Moderna, not another J&J.” Why couldn’t the FDA just flat-out say that?

Still, I was delighted to see mix-and-match get the official okay, and get it quickly.

I was less delighted with the FDA’s decisions regarding who should get boosters. It had no choice but to authorize boosters (or second shots) for all J&J vaccinees. I hoped that might lead it to say that all Pfizer and Moderna vaccinees could also get boosters. That would have meant reconsidering its September decision on Pfizer eligibility, and contravening VRBPAC’s advice regarding Moderna eligibility. So I knew it was unlikely, and I was right.

Instead, the FDA stuck with its Pfizer eligibility criteria and replicated them for Moderna.

22. The actual booster decision process (part seven – the ACIP).

So here’s where things stood as of October 20, as the focus shifted from VRBPAC and the FDA to the ACIP and the CDC:

• All J&J vaccinees can get a “booster” (that is, a second shot).
• Pfizer and Moderna vaccinees can get a booster if they’re 65+; or if their medical condition poses a high risk of severe outcomes; or if their working or living situation poses a high risk of infection – which in my judgment is everybody 18+ but doesn’t sound like it.
• If you’re eligible for a booster at all, you can get any of the three boosters, with no recommendation of which vaccinees should get which boosters.

On October 21, the ACIP met and cast unanimous yes votes on exactly what’s in these bullet points. The only thing it added is a distinction between Pfizer and Moderna vaccinees who should get a booster (65+; or 50+ with a high-risk medical condition) and those who can get a booster (everybody else the FDA said was eligible).

Though the ACIP votes were unanimous, there was some concern expressed about the signal those votes were sending to the “worried well” that maybe they should get a booster. (The people likeliest to notice that the eligibility criteria are a lot broader than they look are of course the people who wish they weren’t so broad.) As ACIP member Sarah Long put it: “I can’t say that I am comfortable that anybody under 50 – an otherwise healthy individual – needs a booster vaccine at this time with either Moderna or Pfizer.”

But the ACIP was bowing to reality. In an article in Medscape Medical News, infectious disease physician Jay Varkey (not involved in the ACIP deliberations) got to the nub of the matter: “The decision made by the FDA and the ACIP recommendations, I think, reflect the real world. The public is going to do what they feel driven to do. This at least adds a scientific review of the currently available data.” In other words, the agencies and their expert advisors had little choice but to follow the lead of the public and the White House.

Or as a STAT article concluded: “Some members of the committee also made it clear they are not convinced that all the people who meet the eligibility criteria actually need to be boosted at this time. But hanging over them was the reality that Walensky had overruled the committee in September when it decided not to recommend boosters for people 18 to 64 whose jobs or living arrangements place them at higher risk of catching Covid. Having different criteria for Pfizer and Moderna vaccine recipients was clearly not on the table, especially given that the Biden administration had originally pushed for boosters for all vaccinated Americans.”

The article also quoted vaccine expert Paul Offit, a longtime opponent of widespread use of COVID boosters: “We have just sent the message,” Offit said disapprovingly, “that you’re not fully immune with two doses, that it’s a three-dose vaccine.”

It wasn’t just the ACIP that was bowing to reality. The public health profession at large was doing so as well. That same STAT article noted: “The sense of the inevitability of Thursday’s outcome was evident in reaction to it, which was effectively a shrug.”

But bowing to reality isn’t the same thing as liking it. The most stunning evidence that many public health professionals still hated the booster rollout came in an October 25 New York Times article entitled “Are Vaccine Boosters Widely Needed? Some Federal Advisers Have Misgivings.” Actually, all the people quoted in the article expressed nothing but misgivings. Nearly all were VRBPAC or ACIP members who had just voted for wider booster eligibility criteria than they personally believed were appropriate – at least that’s how they came across in the Times article by Apoorva Mandavilli.

The article is chock full of revealing quotations, but I’ll confine myself to two.

• ACIP member Matthew Daley told Mandavilli: “I don’t think that we have evidence that everybody in those groups needs a booster today.” (I quail at the thought of generalizing Daley’s criterion to all vaccines or even all medical treatments. It’s not enough that you might need it today or that you will benefit from it today or even that you will surely need it soon. Absent evidence that you definitely need it today, Daley seems to think you shouldn’t get it.)
• VRBPAC member Paul Offit told Mandavilli: “You can see the hesitancy in all this…. It’s because in our hearts, I think people don’t quite agree with this notion of a booster dose. The door just got bigger and bigger and bigger, it got wider and wider with each step…. The companies got what they wanted, the administration got what they wanted.” (Offit’s resentment at what he sees as an inappropriate cart-before-the-horse process requires no elaboration.)

A few hours after the ACIP voted yes on everything the FDA had authorized, CDC Director Walensky signed off as well. By Friday morning October 22, the process was over. Temporarily over, that is.

23. The actual booster decision process (part eight – overview).

It’s worth remembering that the bulk of the Pfizer booster data came not from the company’s studies but from real-world experience in Israel, the first country to roll out a major booster campaign. Israel relied almost exclusively on the Pfizer vaccine. So when it came time to consider Moderna and J&J boosters, the agencies had no real-world evidence to look at, just the companies’ studies. There was a lot of criticism of the adequacy of those studies.

Back in September, many experts had thought there was too little evidence to justify going ahead with Pfizer boosters, even with the Israeli data. But once the Pfizer okay had been granted and booster doses were officially going into the arms of Pfizer vaccinees, there was no way the agencies were going to keep Moderna and J&J vaccinees waiting for more data. Considerably less data than the agencies had re Pfizer suddenly sufficed – albeit with repeated complaints – re Moderna and J&J.

It’s hard to see this as following the science. It seems more like following precedent.

Not that I’m criticizing. In fact, one of the most impressive things about the Moderna and J&J booster discussions, compared to the Pfizer discussions, was the growing realization that FDA and CDC Standard Operating Procedure might be the wrong model for the worst public health crisis in a century.

This was especially stunning coming from the FDA, which has a long record of reluctance to “cut corners” no matter how severe the emergency. Before the VRBPAC meetings to consider Moderna and J&J boosters, the FDA sent out (and posted) its reviews of the two companies’ applications. In both cases the FDA emphasized how sketchy the companies’ data submissions were. In the case of J&J, the FDA also emphasized that it hadn’t had time to reanalyze the company’s raw data the way it likes to do. (Slide after slide carried the disconcerting warning: “Not verified by FDA.” ) But the public health need to decide what to do about J&J boosters was too urgent, agency leaders said, to justify delaying the process for additional months while it completed its usual reanalysis. So it was asking VRBPAC to render its advice without this additional rigor. And without this additional rigor, VRBPAC recommended granting the J&J booster Emergency Use Authorization.

I need to stress that I’m not criticizing the FDA and its advisory committee for “cutting corners” in their assessment of the Moderna and especially the J&J booster data – more corners than they had to cut when they had assessed the Pfizer booster data a month earlier. I think the worst public health crisis in a century is more than enough reason for government agencies to deviate in minor ways from their Standard Operating Procedure. One VRBPAC member said she had initially been inclined to vote against J&J boosters because of data inadequacies, but changed her mind because of the “public health imperative.” I’m totally onboard with that.

Was President Biden and the White House wise or unwise to put public pressure on the FDA and the CDC? The public pressure certainly worked. If the pressure had been applied only behind the scenes, I doubt anybody in the U.S. except maybe the immunocompromised would have received booster doses before Thanksgiving. On the other hand, the public pressure irritated a lot of public health professionals, prompting more than a few to compare Biden angrily with his predecessor. The pressure was not without costs.

I think the pressure did a lot more good than harm. But I’m putting more on the credit side of the balance than simply the fact that booster doses started going into arms sooner than would have happened otherwise. In my accounting, it was a good thing for scientific agencies and experts to be forced to respond to public opinion and government policy – instead of the experts → agencies → politicians → public flow of influence that public health professionals cherish. And in my accounting, it was also a good thing that the public got to watch it happen; got to watch the experts’ dilemmas, doubts, and disagreements getting aired; got to watch the COVID booster policymaking sausage being made.

That’s not how most public health professionals calculate the upsides and downsides – not even the ones who have come to agree that it was a good idea to start getting booster doses into arms.

---

Deep in the Weeds

The next few sections are deep in the sausage-making weeds. Exhausted readers may want to skip to “The Rest of the Story.”

24. Booster opponents lost.

To really understand the COVID booster sausage-making process, you need to understand why I think the September Pfizer and October Moderna eligibility criteria – especially the inclusion of people at high risk of getting infected – represented a huge concession by the public health mainstream, which had steadfastly maintained that COVID vaccines are “designed” to prevent severe illnesses, not mild infections.

Arguably, “concession” is too weak. It was a defeat.

Healthcare workers have frequent occupational exposures to infected people, so they’re likelier than the rest of us to get infected. People who spend time in congregate settings (prisoners, for example) are similarly more exposed and therefore likelier to get infected. But there’s no reason to think that their infections are likelier to be severe. In fact, they may keep getting boosted by recurrent exposures, lowering their risk of serious outcomes. In some cases, even the claim of high occupational risk of infection is debatable. Healthcare workers and others in some high-exposure occupations tend to wear personal protective equipment and take other precautions; some studies find that their infections more often result from community exposures than from their occupational exposures. So why should they be prioritized for boosters?

Some formulations of the working/living situation eligibility criterion suggested that it’s not enough to be at high risk of getting infected. According to these formulations, to be eligible your infection had to be likely to lead to serious complications. But nobody I’m aware of proposed that any young, healthy nurse or orderly should be denied a booster. Most of the formulations didn’t even mention complications; high risk of getting infected was sufficient.

I mention nurses and orderlies because that’s where the “high risk of getting infected” criterion came from. I can understand why even public health professionals who think vaccines are meant only to prevent severe outcomes might want to make an exception for healthcare workers – because they’re so badly needed on the frontlines, because they’re heroes of the pandemic, maybe just because public health professionals know lots of healthcare workers. But that exception let the camel’s nose in under the tent, leading to reasoning like the following:

• What about other occupations where the risk of infection is high or thought to be high or at least claimed to be high – meatpackers, say?
• What about other pandemic heroes – starting with cops, firefighters, and other first responders, and working our way to grocery store clerks and delivery people?
• But we can’t include cops and leave out prison guards. And we can’t include prison guards and leave out prisoners.
• Teachers say they’re especially vulnerable too, because many of their students can’t be vaccinated; and they’re politically powerful; and we really, really want to keep the schools open.
• But parents of young children also have to spend time with the unvaccinated – their own kids! And some of them are terribly anxious about it. What about them?
• Blacks and Hispanics have suffered disproportionally throughout the pandemic, and so have poor people no matter what their ethnicity. And they’re likelier than white, middle-class people to have comorbidities they’re unaware of. We need to find a way to include them too.

Better not to make a list, or at least not a list that purports to be all-inclusive. Just say you qualify for a booster if something about your living or working situation makes you high-risk for getting infected. Give a few examples and make sure there’s an etcetera. And put it all on the honor system; instruct vaccinators not to check.

And that’s what they did.

25. How they “hid” the loss.

There was never a battle over boosters for immunocompromised people or J&J vaccinees. Whether we call it a booster or not, they obviously needed another shot – all of them. The battle was over which Pfizer and Moderna vaccinees ought to get boosters.

The outcome of that battle: Boosters were recommended for all elderly people, plus all middle-aged people whose medical histories make it especially dangerous for them to get infected. That’s a reasonable operationalization of the evidence vis-à-vis vulnerability to severe outcomes.

In addition, boosters were permitted for all adults who claim they are high-risk, as long as they think they have a medical condition that makes them especially likely to get very sick or a working/living situation that makes them especially likely to get infected. That’s a reasonable operationalization of everybody who wants a booster.

But if you don’t especially think of yourself as high-risk, you might think the FDA and the CDC have decided you’re not eligible. “Everyone who’s high-risk” doesn’t sound like it means the same thing as “everyone.”

The experts were polarized about boosters, and so were the members of the FDA and CDC advisory committees (especially in September – by October their anti-booster sentiment had substantially diminished). One side insisted that there was simply no evidence that boosters are “needed” yet, except maybe for people who are especially likely to get very sick if they get infected – the elderly and those with high-risk medical conditions. The other side pointed out that boosters are already at least a little beneficial to younger and healthier people too, and lots of them already want boosters, and the boosters are low-risk, and we may soon start to see data that it’s important to give everybody boosters – so why not just okay them for everybody now?

If forced to choose between the two extremes, boosters for everybody versus boosters only for a narrow range of highly vulnerable people, I don’t know which option would have won a vote of the FDA and CDC advisory committees (or for that matter a vote of the FDA and CDC staff experts). Back in September and even October, I suspect the narrow, only-the-most-vulnerable approach would have been the majority view. But the White House had clearly already adopted the boosters-for-everybody view, and top agency officials had signed onto the White House booster plan. As I’ve said before, the handwriting was on the wall. Or to put it the way booster opponents saw it, the fix was in.

Based on the September Pfizer and October Moderna decisions, boosters-for-everybody won. Not officially, but in effect.

Here’s how the FDA and the CDC concealed it. Instead of just saying “boosters are okay for everybody,” they specified complicated, broad, porous categories of “okayness,” thereby giving the impression that the more restrictive view won. Booster opponents weren’t fooled; they knew they lost. Booster supporters – including the White House – weren’t fooled either; they knew they won. But the agencies preserved their brand. They looked like they were standing tall for Science, not just bowing to White House pressure. And they helped booster opponents save face.

The downside from the perspective of booster supporters: Some people who were eligible for Pfizer or Moderna boosters didn’t realize it. They weren’t sure if they were eligible or not, or they were mistakenly sure they weren’t. I have talked with a number of healthy young and middle-aged people who definitely wanted to get boosted but were dutifully waiting till the agencies said they could – not realizing that the agencies had already tacitly said they could.

Some vaccinator organizations were in the same pickle, trying to decide who’s entitled to a booster. Even if they knew it’s on the honor system, they wanted to be able to tell prospective vaccinees what the eligibility criteria were. A lot of media coverage exacerbated the problem by “explaining” the eligibility criteria in ways that implied they were limited to people who are elderly or “high-risk,” without explaining that the definition of “high-risk” was intentionally vague; you’re high-risk if you think you are.

The CDC’s October 21 summary of who’s booster-eligible was the template for hundreds if not thousands of summaries produced by health departments, hospital systems, pharmacies, and journalists. Here’s what it says:

For individuals who received a Pfizer-BioNTech or Moderna COVID-19 vaccine, the following groups are eligible for a booster shot at 6 months or more after their initial series:

• 65 years and older
• Age 18+ who live in long-term care settings
• Age 18+ who have underlying medical conditions
• Age 18+ who work or live in high-risk settings

(Interestingly, the CDC’s own distinction between “recommended” and “permitted” has disappeared from this CDC summary. I like that distinction, and I have regretted seeing the CDC deemphasize it if not quite abandon it, focusing instead simply on who’s “eligible.” )

The links in the bullet points are to CDC webpages. Those webpages list (or listed – everything changed on November 19) examples of groups that the CDC considers to be covered. Importantly, each of the three lists begins with an explicit statement that the list is not complete.

26. You could get a booster if you wanted one.

Do you have an underlying medical condition? Do you work or live in a high-risk setting? I think anyone who was determined to get a booster shot could honestly answer “yes” to at least one of these two questions. I also think plenty of less determined people might easily get the impression that they don’t qualify if they have just ordinary sorts of medical conditions and the places where they work and live pose just ordinary sorts of COVID risk.

Again and again in VRBPAC and ACIP meetings, committee members asserted their strong aversion to authorizing or recommending universal boosters for all vaccinees. Again and again they complained that the eligibility criteria they were considering were broad enough and ambiguous enough to encompass all vaccinees. When presented with explicitly all-encompassing authorizations or recommendations, they voted against them or found a way to avoid voting on them at all. But when presented with authorization or recommendation policies that sounded restrictive but weren’t, they voted yes. (As we have seen, there was one exception, one time the ACIP voted no on a restrictive-sounding but porous eligibility criterion. The CDC director “overruled” that vote.)

Many of the same VRBPAC and ACIP members insisted that a universal booster policy would be bad science, complained that the eligibility criteria they were considering added up to a universal booster policy, and voted yes on those criteria. Here for example are two comments from pediatrician and vaccine expert Paul Offit at the Moderna booster VRBPAC meeting. Asked when if ever it might be a good idea to expand the eligibility of boosters to everyone over 18, Offit said “maybe never.” He also said: “I’m uncomfortable with how we’ve sort of tripped down the line for the thought of universal booster dosing, which I just think is wrong.” Offit did not dissent from the unanimous yes vote.

At the October 14–15 meeting of the FDA’s VRBPAC advisory committee to come up with booster recommendations for Moderna and J&J vaccinees, CDC official Amanda Cohn told the group that 60% of all Pfizer vaccinees between the ages of 18 and 65 were covered by the booster eligibility criteria that had been promulgated the month before. Cohn apparently thought 40% of Pfizer vaccinees under 65 weren’t eligible for boosters – that is, had no qualifying medical condition or working or living situation.

To reach this determination, or any determination, Cohn would have needed to consult inclusive lists of which medical conditions count as high-risk of severe outcomes and which working and living situations count as high-risk of infection. I have no idea what lists she was using. I can’t find any lists that claim to be inclusive, only illustrative like this one from CDC. [Now that the CDC supports boosters for everyone over 18, this link no longer has an eligibility list.] I wrote Cohn to ask what she based her 60% figure on, but didn’t get a response.

On October 28, the New York Times reported that the CDC had added several mental health conditions to its (illustrative) list of conditions that make sufferers vulnerable to severe outcomes – and therefore eligible for boosters at any age 18 and older. “The door just keeps getting wider and wider,” complained Offit, who estimated that about 85 percent of the adult U.S. population was now eligible.

I’ll stand by my 100% figure.

Even so, I have read endless media articles about who’s covered and who isn’t, and only one or two that explicitly agreed with me that the criteria were so vague and so porous that they actually covered everyone. Certainly if you casually perused FDA and CDC webpages you didn’t get the impression that everyone was covered. So there have got to be millions of people who thought they weren’t covered, and millions more who weren’t sure whether they were covered or not. On the other hand, if you scrutinized those same FDA and CDC webpages looking for an eligibility criterion you could squeeze yourself into, you would almost certainly end up deciding you’re covered if you want to be.

Arguably that was an okay outcome: If you believed strongly that you ought to be covered, you would find evidence that you’re entitled to the booster you’re seeking. If you were only mildly interested, you might get the impression that boosters aren’t for you.

At least not yet. As soon as the eligibility criteria were promulgated, there was considerable discussion about revising them to include more people (or as I would put it, to explicitly include more people). The data on waning, breakthrough infections, and severe breakthrough infections are a moving target that FDA and CDC officials were watching closely. An October 20 Washington Post story, for example, reported that the FDA was considering that it might soon want to lower the age at which everybody is authorized to get a booster from 65 to 40. As Peter Marks put it during the VRBPAC mix-and-match discussion: “There is evidence to suggest that lowering the age of those eligible for boosters may make sense in the future.”

And remember, all these eligibility criteria were on the honor system. Washington Post columnist and former Baltimore health commissioner Leana Wen summarized it this way in a September 30 newsletter: “If it were up to me, I would have said everyone 18 and older should be able to get a booster six months after their last shot…. Still, I think the ultimate decision was fine, because all that’s needed to get the booster is a self-attestation – no documentation required. Essentially, that means all adults who want a booster can get one.”

27. The “mild infections don’t matter” inconsistency.

Now I want to get even deeper into the weeds, looking at three inconsistencies embodied in the Pfizer and Moderna eligibility criteria as of late October.

Theoretical inconsistency isn’t an especially serious public policy sin. In fact, a fair amount of “smoosh” – compromising the uncompromisable by pretending the inconsistencies aren’t there – is certainly useful and arguably necessary to get anything significant through the policy sausage-making process. Nonetheless, I think understanding these three inconsistencies will shed light on how COVID booster policy sausage got made.

First, consider the claim that vaccines are mostly to prevent severe outcomes, so if severe outcomes are still being prevented, then boosters aren’t needed.

Leave aside the weirdness of pretending that prevention is dichotomous – so even if a vaccine’s efficacy against severe outcomes declines from around 90% to around 70% it still “prevents” severe outcomes; and even if one vaccine’s efficacy is 90% and another’s is 70% they both still “prevent” severe outcomes. And leave aside the weirdness of framing the question in terms of whether boosters are “needed” rather than whether they’re beneficial. (I’ll get to that one later.)

Focus instead on the pretense that mild COVID infections are negligible – a pretense that is inconsistent with everything else that public health professionals have been telling the public, and particularly the fully vaccinated public. It is inconsistent, for example, with the Phase 3 trials of the three COVID vaccines, which reported vaccine efficacy against laboratory-confirmed infection or against mild-to-moderate symptoms. The trials reported no statistically significant data on vaccine efficacy against severe illness – which didn’t keep the agencies from concluding that the vaccines work and deserve their Emergency Use Authorizations.

And it is inconsistent with the furor when the CDC announced in May that vaccinated people didn’t need to wear masks. Many public health professionals vociferously objected to that announcement. Even though the breakthrough infections of unmasked vaccinated people would probably be mild, they said, those infections still might be transmitted to unvaccinated people or immunocompromised people. The rampant criticism plus the emergence of the Delta variant forced the CDC to recant.

Experts and officials continue to assert with considerable fervor that fully vaccinated people should wear masks and take a range of other non-pharmaceutical precautions, despite the infrequency of severe breakthrough infections. They do so even though this insistence contributes to the hesitancy of the unvaccinated, undermining the appeal that you can get your life back if only you get vaccinated. They do so, presumably, on the grounds that mild breakthrough infections matter. Their contrary assertion that mild breakthrough infections do not matter is deployed only for discussions of boosters.

I have done a lot of work on flu vaccination controversies. I don’t recall public health professionals insisting that the flu vaccine is “designed” only to prevent flu hospitalizations and flu deaths. To the contrary, the argument that you can lessen your chances of an unpleasant week home in bed is pretty frequently deployed on behalf of getting a flu shot. Young, healthy adults whose flu cases are highly unlikely to require hospitalization are nonetheless urged to get vaccinated every year. Prior to the debate over COVID boosters, I seldom saw anyone in public health claim that vaccines are only to prevent severe outcomes and shouldn’t be used to prevent mild cases.

28. The occupational/institutional inconsistency.

The second inconsistency arises directly from the first.

Having committed themselves to the position that mild breakthrough COVID infections are a negligible problem, public health professionals faced the dilemma of how to justify boosters for healthcare workers. As we have seen, VRBPAC initially stayed consistent, confining its support for Pfizer boosters to the elderly and those with comorbidities – people with a high risk of severe outcomes if they’re infected. But at the last minute the FDA’s Peter Marks got the committee to give its informal approval to expanding from “high risk” of severe outcomes to “high risk” of infection – opening the door not just to healthcare workers but to all kinds of people whose working or living situations put them at high risk of infection.

Marks and VRBPAC might have made an exception for healthcare workers on the grounds that healthcare worker absenteeism is a big health problem in the middle of a pandemic surge. But there was no rationale for adding teachers and grocery clerks and people in prisons. To make sense of that, they had to abandon their claim that protection against infection doesn’t matter much.

They didn’t so much abandon it as tacitly undermine it.

Remember what we know about the science: Vaccine efficacy is already demonstrably waning against infection (so there are more breakthrough infections), but so far it isn’t waning much against severe outcomes (so very few breakthrough infections end in hospitalization). I italicized “so far” because it’s entirely possible that in the months ahead we will see declines in vaccine efficacy against severe outcomes too. Booster supporters tend to think reducing the number of mild infections is itself a worthwhile goal. And they (we) tend to think boosters also make sense as a precaution against the possibility of more severe breakthroughs to come.

But if you really think boosters should be forbidden except for people who have a proven high risk of severe outcomes, you’ve got no business allowing young, healthy people to get boosted, no matter what occupational or institutional exposure they face.

Pretty clearly, VRBPAC acceded to Marks’s request because Marks was signaling pretty clearly that if it didn’t the FDA might have do what Marks called “tweaking” its recommendations but the media would surely have called “overruling” them. VRBPAC didn’t want the FDA to overrule it, so it went along. A few days later CDC Director Walensky urged the ACIP to do likewise. The ACIP voted the other way, 9-6, so Walensky had to tweak/overrule (reject) its advice.

In order to support boosters for healthcare workers, prison employees, grocery clerks, and residents of congregate living situations (and maybe crowded apartments too), you have to decide that mild infections matter after all. Give boosters to hospital workers because hospitals are already desperately understaffed and can’t afford additional absenteeism when employees are mildly infected with COVID. Give boosters to nursing home workers because their patients are very vulnerable and can’t afford to get infected by a staff member. Give boosters to poor people who live in crowded conditions and are in frontline, outside-the-home jobs because they have already suffered more from the pandemic than anyone else. But in each case, your rationale for giving them boosters relies on believing that their mild infections matter.

For the first month or so, the FDA, the CDC, their advisory committees, and many public health professionals made periodic efforts to cover over this second inconsistency. They were a little reluctant to say simply that they’re authorizing/recommending boosters for everybody whose working or living situation puts them at high risk of getting infected. That sounds too much like everybody. So they sometimes deployed language like this, from the FDA’s September 22 announcement of its criteria for Pfizer boosters: “individuals 18 through 64 years of age whose frequent institutional or occupational exposure to SARS-CoV-2 puts them at high risk of serious complications of COVID-19 including severe COVID-19.”

The question is which term was operational, “frequent … exposure” (which sounds like everybody) or “high risk of serious complications” (which sounds like just elderly people and those with comorbidities). The answer is “frequent … exposure.” There was no effort to suggest that some people might have frequent exposure and nonetheless not be at high risk of serious complications. After all, every exposure poses some risk of serious complications. So no matter who you are, the more often you are exposed, the likelier it is that one of those exposures will lead to serious complications. Bottom line: Everybody with frequent institutional or occupational exposure qualified for a booster.

It’s not quite clear what “institutional” means here. Nursing homes for sure. But also prisons and dormitories, with all those healthy young people. Is a crowded apartment also an institution? Is a crowded bar?

By October, as the agencies were expanding their booster policies from Pfizer vaccinees to all vaccinees, the use of this “high risk of serious complications” framing had waned. Pretty much everybody had come to frame booster eligibility in terms of whether your working or living arrangements made you “high risk” – without specifying high risk of what. The claim that boosters shouldn’t be used to prevent mild cases was thoroughly undermined, but never explicitly abandoned.

29. The “only if you qualify” inconsistency.

The third inconsistency is more fundamental. The Biden administration made it clear from the outset that it wanted to offer boosters to everyone. The majority of public health professionals made it clear that they didn’t. The agencies and their advisory committees were caught in the middle. Their solution was to articulate eligibility criteria that sounded restrictive but weren’t.

I just explained how that works for people “whose frequent institutional or occupational exposure to SARS-CoV-2 puts them at high risk of serious complications.” Except for hermits, everyone’s frequent institutional or occupational exposure puts them at high risk of serious complications. It’s a pandemic! We’re all at high risk! That’s how you sound like you’re narrowing eligibility without narrowing eligibility.

As we shall see when we get to “The Rest of the Story,” by mid-November state governments were unilaterally declaring universal booster eligibility within their borders, simply by taking note that everybody within their borders faced some kind of high COVID risk and was therefore eligible under the FDA and CDC criteria.

The other major eligibility criterion (other than being 65+) is “individuals 18 through 64 years of age at high risk of severe COVID-19.” This too is broader than it sounds. There is no “complete list” of which specific comorbidities qualify, and certainly no list of which ones don’t. High body mass index (being overweight), for example, is correlated with severe outcomes. So are depression and anxiety. So is being a smoker or an ex-smoker. (Just one cigarette? Who knows.) So are some racial and ethnic categories. Men are substantially likelier than women to die of COVID, so is a Y chromosome a comorbidity? All of us have some condition that makes us likelier to experience a severe outcome than we’d be if we didn’t have it.

As I’ve already pointed out, booster administration was on the honor system. But it wouldn’t have to be. All of us have some condition that we can honorably judge makes us high-risk for a severe outcome. And all of us have some aspect of our working or living situation that we can honorably judge makes us high-exposure, and therefore high-risk for infection, and therefore high-risk for serious complications.

That doesn’t necessarily ensure that everybody who might like a booster would understand that the FDA and the CDC have tacitly said okay. The face-saving language fooled more than a few journalists, who reported the agencies’ decisions as if they restricted booster access. A lot of prospective booster candidates got the wrong impression too. I have talked with several acquaintances, friends, and family members who wanted boosters, considered themselves ineligible, and were dutifully waiting their turn, not realizing that their turn had come.

And as I noted earlier, maybe this is just fine. Prospective booster candidates who wanted it badly enough figured out that they could get it. They didn’t have any reason to raise a fuss at the authorities. Prospective booster candidates who preferred not to get it yet felt okay not getting it yet. Nobody would have any reason to raise a fuss at them.

Of course this analysis is grounded in the assumption that boosters are probably beneficial to the individual but boosting everybody we possibly can isn’t really important for the society. So let the ardent people get boosted and the half-hearted ones fall by the wayside. That seemed right to most experts in September, even in October. But as the evidence gets stronger that maybe breakthrough infections are a big deal after all, we may be moving to a different analysis: that all vaccinated people “need” their boosters and public health officials must do all they can to urge booster compliance.

Perhaps the strongest evidence that this reassessment is gaining strength among public health professionals is the increasing frequency with which I’m hearing that COVID vaccination requires a three-dose primary series. In other words, the booster isn’t a booster. You’re not “fully vaccinated” without it. That may very well be where we’re headed.

On November 19 universal booster eligibility became the order of the day, and the three inconsistencies disappeared. But we haven’t gotten to “The Rest of the Story” yet.

Like many other realities of booster policy sausage-making, the three inconsistencies may have disillusioned a lot of observers. But I think disillusionment with the sausage-making process is a feature, not a bug. The public’s illusions about expert omniscience, expert unanimity, and the pristineness of science are harms, not benefits.

---

Confusion

30. Some public confusion is inevitable.

Virtually every news story written about COVID boosters has pointed out that the decision process left the public “confused.”

I agree that people who tried to follow every twist and turn between August 18 when the White House announced the plan to start giving people boosters and September 24 when CDC Director Walensky outlined her agency’s booster recommendations for Pfizer vaccinees might well have been confused about exactly how the sausage got made. Further twists and turns vis-à-vis Moderna and J&J boosters in October only exacerbated the confusion.

Developments in November should have ameliorated the confusion – but I’m saving those for “The Rest of the Story.”

I also agree that those who tried to follow the twists and turns probably ended up believing that the decision process itself was confused, at least insofar as “confused” means that there was a lot of expert disagreement, a lot of backing-and-forthing, and a fair amount of dishonesty. They saw the White House pressuring the agencies while still claiming it was merely following the science and the scientists. They saw the experts disagreeing with each other, mostly about nonscientific issues, while still claiming they were standing tall for indisputable scientific truths.

And whether or not they followed the decision process closely, no doubt many prospective booster recipients were left less than certain exactly who the agencies had decided could get boosted. There was genuine confusion about how porous the agencies’ “high-risk” eligibility criteria actually were. Not everyone realized that they were not just porous but explicitly on the honor system.

This confusion was greatly exacerbated by the quite different ways some journalists and some state and local health departments reported – or interpreted – the eligibility criteria. There were overlapping but significantly varying bullet point lists of who’s eligible. The California government vaccine website, to cite just one example, reasoned from the CDC’s criterion re working/living situations that boosters are recommended, not just permitted, for various “communities of color” that are victims of “social inequity.” Some journalists and some state and local health departments explained or at least implied that the criteria were broad and vague, and that their lists were illustrative rather than inclusive. Others offered websites with a series of check-the-box options, and if you checked the wrong boxes up popped a bold-type dogmatic assertion that “You Are Not Eligible At This Time.”

In short, the FDA and the CDC genuinely created public confusion when they opted for eligibility criteria that were vague and porous but sounded restrictive. It would have been a lot less confusing to sign onto the August White House boosters-for-everybody plan.

That’s what infectious disease expert Celine Gounder told The Hill in a September 26 story on the Pfizer eligibility criteria: “There’s going to be confusion. If we are going to create guidelines that are essentially making the vaccine available to almost everyone, the simplest solution is, make it available to everyone.” Gounder was an opponent of boosters for everyone. But she was right that it added confusion without adding value for the agencies to allow boosters for everyone or nearly everyone, as Biden wanted, while pretending not to. Why promulgate loosely defined criteria that sound restrictive but cover pretty much everyone who makes an effort? Why not just say vaccinees who want a booster can get a booster?

I’m not sure that’s what cardiologist (and Medscape editor-in-chief) Eric Topol had in mind when he tweeted on September 23, “You’d have to work really hard to mess this up this badly and engender this much confusion.” Certainly most public health professionals didn’t support an explicit boosters-for-everyone policy. They would have preferred a different way to reduce public confusion: If only the White House had just butted out, letting the experts and agencies work toward a behind-the-scenes consensus on boosters, however long that might have taken.

Putting process confusions aside, the outcomes are confusing enough. J&J vaccinees can get another shot after two months, but Pfizer and Moderna vaccinees have to wait six months, except that all immunocompromised people can get another shot after just one month. There’s a full Moderna dose for people’s first two shots, and for immunocompromised people’s third shot, and for J&J vaccinees’ second shot, but a half-dose for everybody else’s third shot. There’s a Pfizer one-third dose for children 5–11, but still the Pfizer full dose for everybody else’s primary doses or booster doses. Etcetera. It’s sufficiently confusing that the FDA and the CDC had to put together charts to help vaccinators get it right.

But more fundamentally, what’s confusing is simply that COVID is confusing. There’s a lot nobody knows yet about how long vaccine efficacy lasts, or how important antibodies are versus memory cells, or how much good boosters will do. And a lot of what we know will turn out wrong, either because we got it wrong or because it actually changed. Just imagine what may change if a new COVID variant replaces Delta as the dominant circulating variant in the U.S.

In the face of this intrinsic and unavoidable confusion, I think trying not to confuse the public can easily do more harm than good – for example, if it leads the agencies to stick to prior recommendations because updating those recommendations based on new science could confuse the public. (I’ve seen that story before.) I think the public health profession has to trust the public to be grownups (and help teach the public to be grownups) – to tolerate uncertainty and expert disagreement and changing recommendations; and to make our own decisions knowing that those decisions may not always turn out the way we hoped.

There’s a natural OMG reaction when people learn how messy the process of scientific discovery is – and how much messier, even, the policymaking process is when the science isn’t clear and isn’t necessarily dispositive anyway. It’s natural that experts and policymakers prefer to shield the public from that reaction. But when they try and fail, they set themselves up for mistrust. And when they try and succeed, the success is temporary and the mistrust is that much greater when the public belatedly realizes it has been hoodwinked with a sanitized, overconfident myth.

Better to come out of the closet – to teach the public that uncertainty and expert disagreement are normal, and that when uncertainty and expert disagreement become publicly visible (as they have with regard to boosters), that’s a good thing. Normalize the sausage-making process and help the public bear its OMG reaction.

Experts and policymakers should validate public skepticism about their own pronouncements. Don’t just acknowledge uncertainty and expert disagreement sotto voce; proclaim them loudly:

We don’t know what we’re doing, we’re uncertain, we disagree. We’re building our boat and sailing it at the same time. We have made some mistakes, and we’re going to make more. We’ll keep changing our minds as we fight out our disagreements and take new evidence onboard. If you’re not sure we’re right about something, join the club: We’re not sure we’re right either. This is what a pandemic is like, a crisis with a new pathogen.

(Just in case it’s not totally clear, the above is a quote from nobody. It’s what I wish public health professionals would say about the COVID pandemic.)

Despite the confusion, I suspect that most people who were serious about wanting a booster figured out the basics. They knew that if they got their primary doses more than six months ago they could get a booster now as long as they were willing to assert that they’re “high-risk.” They knew that determining whether they’re actually “high-risk” is a guessing game, whether they choose to rely on the agencies’ guess or their doctor’s guess or their favorite expert’s guess or the internet’s guess or their own guess. And they knew that there’s no appreciable downside to getting a booster, but the magnitude of the upside is uncertain and debatable.

A great deal of what they knew had to do with how much the experts don’t know. Perhaps that’s partly why the experts kept calling them confused.

31. People aren’t confused about whether they want a booster.

One other thing most people know: They know whether they want a booster. And the majority of them do.

Long before boosters were authorized, the media were running vaguely hostile stories about fully vaccinated people who had managed to get boosted anyway. By August 10, before the White House had even announced Biden’s plan to start giving boosters in late September, the CDC was estimating 1.1 million unauthorized boosters given so far – and counting.

Most people were willing to wait for the official okay. In one late August survey, 81% of fully vaccinated Americans said that if a booster were available to them today they would get it; 10% said they wouldn’t and 8% didn’t know. In another survey around the same time, 77% said yes, 5% said no, and 12% weren’t sure. Data from mid-to-late September showed the numbers were still high: 81% said they would definitely (55%) or probably (26%) get a booster if it is recommended for people like them; only 13% said probably not (8%) or definitely not (5%).

Unlike vaccine efficacy, enthusiasm for boosters hasn’t waned, as it well might if the public were confused about their advisability. Starting with Biden’s August announcement, or even earlier, very few people were confused about whether they wanted a booster. More were “confused” about whether/when the government would let them get the booster they knew they wanted.

Dozens of October and November news stories reported frenetic booster enthusiasm, compared to a moribund market for first shots – often suggesting that the former was damaging the latter. Consider this October 11 New York Times article, headlined “Boosters Are Complicating Efforts to Persuade the Unvaccinated to Get Shots.” The lede: “Vaccinated people have been burning up the phone lines at the community health center in rural Franklin, La., clamoring for the newly authorized Covid booster shot. But only a trickle of people have been coming in for their initial doses, even though the rate of full vaccination in the area is still scarcely 39 percent.”

One consistent finding: Booster enthusiasm is highest in parts of the country where vaccine uptake is lowest. This is entirely rational. If you’re a vaccinated person surrounded by unvaccinated neighbors, a booster is a no-brainer.

Even back when only Pfizer boosters had been authorized, booster doses outnumbered first doses day after day, nationwide.

As of October 9, the CDC’s COVID Data Tracker said about 7.8 million fully vaccinated people had received an additional dose. As of October 16, one week later, the number was 10.1 million. The running tally includes immunocompromised people who were authorized to get an extra shot on August 13, well before anybody else; but it doesn’t include those who got unauthorized boosters before the CDC gave the go-ahead for Pfizer recipients on September 24. Figuring those two roughly balance each other out, let’s guesstimate that ten million Pfizer recipients (the only ones eligible at the time) got boosters between September 24 and October 16. That’s roughly 10% of the country’s 103 million fully vaccinated Pfizer recipients.

Now, of course, boosters are available to Moderna and J&J vaccinees as well. As of November 20, the CDC running tally shows that 17.6% of the 196 million fully vaccinated Americans have received a booster. I haven’t tried to calculate how many of those 196 million aren’t eligible yet because they got their primary Pfizer or Moderna doses less than six months ago. And there’s no way to calculate how many of the 196 million looked at the September/October eligibility criteria and reluctantly (and I think mistakenly) concluded that they weren’t eligible. Still, 17.6% is a lower number than I’d have expected. A more encouraging November 20 number: 40% of fully vaccinated people 65 or older have received their boosters.

Considering that around 80% of fully vaccinated people keep saying in polls that they want a booster shot, I’d have expected more than 17.6% to have been boosted by now – despite the drag of restrictive-sounding eligibility criteria.

The majority of vaccinees don’t seem to be confused about whether they want a booster. Maybe they were confused until very, very recently about whether they were eligible. Maybe they’re just not in a hurry. Or maybe a booster isn’t as high a priority for them as the survey data seem to suggest.

Whatever the reasons, the pattern seems to have shifted. In September and October, determined booster-seekers had to brave the skepticism, even the contempt, of public health professionals who thought they were being selfish or foolish or both. But by November, as we shall see, more and more public health professionals were turning into booster enthusiasts. With the barriers down at last, the next few weeks will tell us how enthusiastic the vaccinated public is turning out to be.

32. Some vaccinators have been confused.

Vaccinators know the people telephoning or showing up asking for a booster shot aren’t confused about what they want. But under the September/October eligibility criteria, some vaccinators were confused about whether to give it to them.

This was particularly an issue for doctors, especially doctors who know their patients – so they know which patient has what high-risk medical condition and maybe even which patient has what high-risk working or living situation. Even though the CDC was clear from the outset that booster qualifications are on the honor system, it’s understandably hard for some doctors to accept the patient’s word for something the doctor knows isn’t true. And even though the patient might judge that his or her medical condition or working/living situation is indeed high-risk, the doctor might feel entitled to a different opinion.

Still, this confusion wasn’t likely to interfere much, because most boosters are being dispensed in pharmacies, not doctors’ offices. Pharmacy managements have uniformly said they have two relevant policies: Don’t give people boosters unless they qualify, and take people’s word for it when they say they qualify. No pharmacy I know of has been checking that you really have a medical condition or working/living situation that the pharmacist considers high-risk. Many aren’t even checking that you really got your primary Pfizer or Moderna doses more than six months ago.

What if even you didn’t think you qualified under the eligibility criteria, but you thought you should be entitled to a booster anyway? If you were scrupulously honest with your doctor or pharmacist, it’s likely that your doctor or pharmacist turned you down. The FDA, the CDC, and a number of professional societies issued scary-sounding declarations about the awful things that might happen to doctors or pharmacists who knowingly gave someone an unauthorized booster. They might not be reimbursed for the cost of the dose. They might not be included in future vaccine distributions. They might not be protected from liability if the patient winds up with a serious side effect. And so on. The message was loud and clear: Don’t do it!

It’s legally significant that the Pfizer vaccine is already a licensed vaccine. Before August 23 the FDA had granted Pfizer only an Emergency Use Authorization (which is all the other two vaccines have so far). Medications with an EUA can be legally prescribed and dispensed only for the uses authorized. But fully licensed medications can be legally prescribed and dispensed “off-label” – that is, for uses that aren’t mentioned in their licenses. So maybe doctors were legally entitled to prescribe Pfizer boosters at will, even if the doctor thought the patient didn’t meet the eligibility criteria. And maybe pharmacists were legally entitled to fill these off-label prescriptions.

I say “maybe” because COVID vaccine doses in the U.S. are all owned by the U.S. government. That gives the government unusual power to decide who can legally prescribe and dispense them. As best I can tell, the legality of off-label prescriptions for the Pfizer COVID vaccine is unsettled. Off-label prescriptions for the Moderna and J&J vaccines are clearly illegal until those vaccines are licensed.

The likeliest off-label booster prospect was probably a young, healthy person who didn’t meet (and didn’t claim to meet) any of the high-risk categories, but who nonetheless sought the added protection of an additional vaccine dose. Nobody seriously claims that giving such a person a booster shot endangered the patient. But it might endanger the career of the doctor, pharmacist, or other vaccinator.

So I was intrigued to read this paraphrase of ACIP member William Schaffner in a November 8 Washington Post article – that “if a patient told him she wanted to visit her grandparents and do everything possible to reduce the risk of spreading the virus, he would give the go-ahead for a booster, as long as the patient was six months past the last shot.” Unless you think the eligibility criteria were as infinitely porous as I think they were, wanting to give your grandparents an extra margin of safety when you visit them did not meet the criteria. It certainly wasn’t on any list I’ve seen.

The other likely off-label booster prospect is someone who wants a booster less than six months after his or her most recent dose – for example, someone who expects to be out of the country and unavailable for a shot by the time he or she reaches the six-month anniversary date. Once again, nobody seriously claims such a “premature” booster is going to harm the patient.

A booster after just four months, say, probably yields a weaker immune response than one after six months; but then again a booster after six months probably yields a weaker immune response than one after eight months. It’s an open question how long to wait, balancing the risk of a breakthrough infection in the meantime against the value of a longer interval. After the White House flirted briefly with eight months and five months, the FDA and the CDC settled on six months as the optimal interval.

They may eventually change their minds if they decide the evidence supports a longer or shorter interval. In the meantime they feel very, very strongly that the interval they picked is the only interval any doctor or pharmacist or prospective booster recipient should pick.

---

Some Key Distinctions

33. Recommended versus permitted.

I really like the distinction between people the CDC says should get boosters and people it says may get boosters depending on their own and their doctors’ risk assessments.

This distinction hasn’t gotten proper media attention. A lot of news stories cut to the bottom line of who can get boosters – and they often got even that wrong, regurgitating the agencies’ complicated and vague criteria without noticing that they added up to just about everybody who got their primary doses more than six months ago. The stories that did take note of the should-versus-may distinction often pooh-poohed it as yet another source of public confusion.

It may be a bit confusing, but I think it’s wonderful. As I have already pointed out, the CDC hasn’t put much stress on its distinction between who’s recommended to get a booster and who’s merely permitted. Often it has conflated the two into a single list of everybody who’s “eligible” for a booster. I think that’s too bad. But at least “eligible” is a middle ground between required and forbidden. Occasionally the CDC has listed only the groups who are recommended to get boosters, leaving unmentioned those who are merely permitted to do so. That’s seriously misleading, but it’s still a middle ground between required and forbidden.

Much too often, public health agencies try to leave as little space as possible between behaviors that are required and those that are forbidden. That is too often their goal: Make people do what’s good for them; don’t let them do what’s bad for them. In the too-often-excluded middle is voluntary behavior. Officials have terminology for this middle ground. The CDC, for example, has long had a category of “permissive recommendations,” which are less strongly recommended than ordinary recommendations. But they don’t use such intermediate categories nearly enough.

Ideally, even the middle should be subdivided. Among the things you’re allowed to decide for yourself, there are some the experts think you should do and some they think you shouldn’t do, and in between are some they have no opinion on, no advice to give you. Put the pieces together and you get a coherent scale with required on one end and forbidden on the other: Must / Should / May / Shouldn’t / Can’t.

At least so far, boosters aren’t widely mandated. Odds are they will be sooner rather than later; they already are in Israel. At a White House COVID press briefing on October 22, CDC Director Walensky said: “Right now we don’t have booster eligibility for all people currently. So we have not yet changed the definition of fully vaccinated. We will continue to look at this. We may need to update our definition of fully vaccinated in the future.” Less than a week after the CDC okayed boosters for Moderna and J&J vaccinees, a New Jersey hospital announced that it would require employees who got the one-dose J&J vaccine to get a booster before Christmas. So far I haven’t seen any U.S. booster mandates for Pfizer or Moderna vaccinees, but they’re coming.

Meanwhile, three of the other four categories have been in use since September. Under the CDC’s October recommendations for Pfizer and Moderna vaccinees, you should get a booster if you’re 65+, or if you’re 50+ and have a medical condition that puts you at high risk for severe COVID. You may get a booster if your working or living situation makes you high-risk for getting infected, or if you’re younger than 50 and have a medical condition that makes you high-risk for severe COVID – and whether you fit into these categories is up to you. You can’t get a booster unless you had two doses more than six months ago.

We can argue about whether the categories are right. I’d especially like to change some can’ts to shouldn’ts. During the several weeks when Pfizer boosters had already gotten the go-ahead but Moderna and J&J boosters had not, I would have warned Moderna and J&J vaccinees that the agencies are still reviewing the evidence on Moderna boosters and J&J boosters and mix-and-match boosters, so they really should wait a few weeks until the agencies reach their decisions. But I’d have told them that if they choose not to wait, it’s their call and they can get a J&J booster now.

What I love is simply that the intermediate categories are getting used. The middle ground between required and forbidden isn’t just in use, it has been subdivided between recommended and permitted. I just wish the CDC would emphasize that subdivision more.

34. Rollout versus market.

The middle ground between required and forbidden would be more obviously appropriate if COVID vaccine doses were an ordinary product that people were free to buy or not buy as they thought best.

In the U.S., most medical products and procedures aren’t really free-market commodities. They need to be licensed (or at least granted an Emergency Use Authorization) before anyone can legally buy them. And for the most part, Americans don’t buy them anyway; their insurance companies or government insurers buy them. Vaccines are even less free-market commodities, because there are government programs that subsidize them, promote them, deliver them, require them, or even all of the above.

And COVID vaccines are not free-market commodities at all. At least so far, the U.S. government is the only U.S. buyer of any COVID vaccine. Every vaccine dose in the country is therefore U.S. government property. And at least so far, the government is giving away all its doses without charge. If the government decides to give you a dose, it’s free. If the government decides not to give you a dose, you can’t buy it on the free market. There is no such market in the United States.

So when the FDA decides to authorize boosters or the CDC decides to recommend them, they are authorizing or recommending a U.S. government rollout at U.S. government expense.

Suppose the FDA and CDC believe that a booster will do a particular group of fully vaccinated people only a little good but practically no harm. The booster’s expected benefit exceeds its expected harm, so in principle people ought to be able to get it if they choose to – but its expected benefit isn’t compelling enough to justify rolling out a government-funded campaign to give away millions of government-owned doses. As things stand right now, there is simply no way for the FDA and CDC to allow people (or their insurance companies) to decide whether or not they want to pay for a booster shot.

I hope that changes. Just as I believe there ought to be a middle ground between required and forbidden, I also believe there ought to be a middle ground between free and unavailable. But at the moment there is no such middle ground for COVID boosters, or indeed for COVID primary doses either.

Every argument about whether COVID boosters are “good enough” therefore raises an underlying question: good enough for what? Good enough for an individual to want if they’re free? Good enough for an individual to pay for? Good enough for the government to pay for? Good enough for the FDA to authorize? For the CDC to recommend? For a vaccine mandate to require? These questions ought to have different answers.

And the wisest answers might well be different for different groups of prospective booster recipients. For example, the government might ideally want to fund a campaign to convince elderly and vulnerable people to roll up their sleeves for free boosters, while allowing healthy young people the option of paying out-of-pocket to get boosted at a local drugstore. That option is off the table unless and until there’s a market in booster doses.

35. FDA versus CDC.

In theory, the FDA and the CDC have different tasks.

The FDA decides whether or not a medical product like a COVID vaccine should be granted a license (or in some cases an Emergency Use Authorization instead) – and if so for what purposes and what groups of people. It grounds the decision in data on the new product’s safety and effectiveness, with comparisons where appropriate to products already on the market. The FDA never recommends for or against a product, and it never requires a product It just permits it or forbids it.

Once the FDA has permitted a medical product, the control typically passes to state governments, which are free to forbid it and sometimes (in the case of a vaccine, for example) free to require it.

The CDC has no role with regard to most medical products. But if a product has a public health aspect, as COVID vaccines obviously do, then after the FDA permits it the CDC gets to decide whether to recommend it – and if so for what purposes and what groups of people. The CDC usually can’t forbid it. That’s the FDA’s call, or the states’. And the CDC usually can’t require it either. As a rule only the states can do that (which is why President Biden’s federal vaccine mandate is legally controversial). But the CDC can use any or all of the steps in the middle between required and forbidden. It can and does recommend a product for some groups and uses, recommend against it for others, and stand neutral for still others. All the other decision-makers – state and local health departments, pharmacies, doctors, insurance companies, etc. – usually follow the CDC’s recommendations, even though theoretically they don’t have to.

The President and the White House normally have nothing to do with any of this.

All of the above is the way things are supposed to work and usually do work. Things have worked entirely differently vis-à-vis COVID vaccines (including boosters) for two obvious reasons. One, the COVID pandemic is the biggest public health crisis of the past century, and there’s no way a responsible president or even an irresponsible president would adopt a hands-off attitude. And two, the U.S. government owns all the doses.

So it simply didn’t compute to FDA officials that they might issue an Emergency Use Authorization for COVID boosters without thereby “authorizing” the government to launch a booster rollout. The government owned all the doses, after all, and the White House had already championed the rollout. For reasons I don’t entirely understand, it also didn’t feel like an option for the FDA to authorize boosters for everyone over 16, as Pfizer had requested, and then let individual vaccinees and their doctors decide whether they should get one, after advice from the CDC. Instead, the FDA and its advisory committee kept asking themselves who should get one, as if that were the same question as who should be permitted to get one.

To make things even more complicated, by the time boosters were under consideration, the Pfizer vaccine was fully licensed. If the government hadn’t owned all the doses, doctors would have been free to prescribe the Pfizer vaccine “off-label” for anyone they wished. In an October 18 opinion piece in STAT (an influential online medical news site), two attorneys argued that the FDA had “overstepped its role” when it started specifying who is authorized to get a Pfizer booster and who isn’t. “The FDA’s role in approving the Pfizer Covid-19 booster is to review safety and efficacy,” they wrote. “Its job does not include balancing ‘public’ costs against benefits, nor deciding who gets priority to receive vaccines.” The two attorneys wrote that that’s supposed to be the CDC’s job, and that the FDA was treading on CDC turf. Normally with a fully licensed product, I think the CDC might make recommendations but the real power would be in the hands of states, doctors, and insurance companies. But not when the federal government owns all the doses.

Then, after the FDA had acted and the CDC was in the driver’s seat, it didn’t compute to recommend or at least okay anything narrower than what the FDA had authorized. FDA officials occasionally justified broad authorizations by pointing out that the CDC couldn’t recommend anything broader than the FDA had authorized, but was entirely free to issue narrower recommendations than the FDA’s authorizations. But in reality the CDC was extremely reluctant to do so. Instead, the discussion centered largely on making sure the CDC’s recommendation criteria matched the FDA’s authorization criteria, as if the roles of the two agencies weren’t different.

When the CDC recommended COVID boosters, by default it was recommending a government COVID booster campaign – and it was recommending it to the White House, which controlled all the doses and had already announced it was waiting impatiently to get such a recommendation.

Arguably, in other words, the FDA’s booster authorizations were narrower than they might have been if the FDA had stuck with simply deciding whether there was anybody for whom the risk of getting boosted exceeded its benefit. And then the CDC’s booster recommendations were broader than they might have been if the CDC hadn’t been trying to stay consistent with what the FDA had already said. The two agencies conflated their roles and met in the middle, in large measure because the government owned all the doses.

Insofar as possible, though, the division of labor is worth preserving. The FDA decides for whom boosters are authorized. The CDC decides for whom they’re recommended. The White House decides how to orchestrate a national rollout. States decide what to require, recommend, or forbid within their borders. Doctors decide what to advise their patients. Individuals decide whose advice to take.

36. Agencies versus advisory committees.

It’s worth remembering that the FDA and the CDC both have outside vaccine advisory committees (VRBPAC and the ACIP respectively). Both committees operate with excruciating transparency; these days you can review their pre-meeting slide sets, then watch the meetings themselves, in real time or afterwards as you prefer, all of it online.

That transparency serves several purposes, one of which is to put considerable pressure on the agencies not to ignore the advice that everyone who’s interested knows they received. There is also a reciprocal pressure on the advisory committees to take note of where the agencies’ thinking seems to be headed, so they won’t casually render advice that the agencies are loath to accept.

For both reasons, it’s unusual and notable when an agency decision diverges from the way its advisory committee voted. Unusual and notable – but not unprecedented and not necessarily inappropriate. If an agency weren’t free to reach a different decision than its advisory committee advised, it couldn’t afford to have an advisory committee at all, certainly not such a transparent one.

The FDA and the CDC are decision-makers (though in the case of the CDC, most of its decisions are nonbinding recommendations). Their advisory committees give them advice. Any differences in opinion between agency and advisory committee are simply part of the sausage-making process. So are differences of opinion within an advisory committee.

The problem (at least many public health professionals consider it a problem; I do not) is that the public gets to watch these differences of opinion get expressed. When an issue is hot and the attention is sizable, watching the sausage-making has the potential to undermine the public’s faith in the omniscient consensus of the experts. This was a serious concern during the September and October VRBPAC and ACIP meetings on the pros and cons of offering people COVID boosters.

Even mildly contentious booster meetings provoked expert tweets regretting that the façade of public health consensus might be crumbling. Even when advisory committee votes were unanimous, as they usually were, it was considered unfortunate that some of the discussion expressed doubts. When the doubts led to a split vote, that was considered more unfortunate still.

Fortunately (that is, fortunately in the view of many public health professionals), mainstream media seldom played up the doubts and disagreements expressed at VRBPAC and ACIP meetings. Most of the coverage focused on the outcomes: Who is to be offered a booster.

But when an agency head “overrules” the agency’s advisory committee, that’s newsworthy.

CDC Director Rochelle Walensky did in fact reach a different decision than her advisory committee advised on one aspect of the agency’s Pfizer booster recommendations. The ACIP voted 9-6 against boosters for people whose working or living situations exposed them to high risk of infection. Walensky sided with the minority. Since advisory committees don’t “rule” anything, it’s misleading to suggest that she somehow “overruled” the committee – though that’s the term virtually all the media coverage used. She took the ACIP’s advice onboard, decided that on balance the minority’s case was better, and was rightly forced to defend that decision to a public that knew in detail what advice she was turning down.

If Walensky had gone the other way, many of the same people who complained that she “overruled” the ACIP would have complained instead that she “confused” the public by failing to recommend exactly what the FDA had authorized. Their real complaint, I think, was that the public got to see the sausage getting made – that President Biden should have waited until the public health profession had settled its booster disagreements quietly, shushed the losers, and established a fake consensus he could “follow.”

I think Walensky made the right decision. More importantly, I think the process had the right balance.

But Walensky made a risk communication error when she tried to downplay her disagreement with the ACIP majority. She should have matter-of-factly acknowledged that the last time a CDC director didn’t take ACIP advice was in 2003. Then the controversy was who should be vaccinated against smallpox as a precaution in case that virus somehow got into the hands of terrorists. Usually the stakes are lower and the public is a lot less interested. Usually the agencies and their advisory committees take as much time as they need to work out their differences until they reach an outcome both can claim to support, with the public paying little if any attention.

In the face of a pandemic crisis, Walensky should have said, this time the ACIP made its recommendations quickly and the CDC director made her decisions immediately thereafter. And yes, her decisions took into consideration the ACIP’s advice, plus her own scientific judgment, plus nonscientific factors – including the fact that most vaccinated people wanted a booster, that the president wanted to give it to them, and that deviations from the FDA’s authorizations might be confusing. Seeming to pretend that these factors (especially public opinion and White House policy) had nothing to do with her “scientific” decision was also a risk communication error.

37. Need versus benefit.

Much of the controversy over boosters has focused on whether people who have already received their primary vaccine doses “need” an additional dose. Almost every article on COVID boosters frames the question in terms of whether or for whom a booster dose is “necessary” or “needed.”

This is a weird frame. Public health professionals urge us all to get flu shots year after year, but they rarely claim that the shot is “needed.” They just say it’s recommended because it might help keep you from coming down with the flu, or make your case milder if you’re infected.

“Need” is a reasonable criterion for an otherwise unacceptable medical intervention. You wouldn’t want brain surgery, for example, unless you really needed it. But for an intervention with little or no downside, we usually talk about “benefit” instead. If a vitamin pill seems likely to benefit you, you’ll probably want to take it, even if you’re not convinced you really need it.

So I think the right question for a prospective booster recipient to ask himself or herself isn’t “Do I really need a booster?” It is “Do I think a booster is likely to do me more benefit than harm?”

But given that there is no free market in booster doses, what is the right question for the government to ask itself? Booster opponents don’t often concede publicly that a booster shot would benefit most prospective recipients more than it would harm them. But they almost never deny that this is so. Their insistence on “need” as the appropriate standard implicitly claims that the government shouldn’t bother giving people boosters merely because those doses probably confer a greater benefit than harm. Instead, they seem to be saying, the government should deliver boosters only to people who really need them.

Most people who want a booster shot aren’t worried about rare vaccine side effects. So even if they’re not confident the extra dose will benefit them enormously, they’re quite confident it won’t do them any lasting harm. For them, getting a booster shot is at worst like taking a vitamin pill they might not need. But for the government, opponents imply, offering them a booster shot they might not need is like doing unnecessary brain surgery.

A side issue that might be persuasive to booster opponents is how the need-versus-benefit distinction affects vaccine hesitancy. Booster opponents worry a lot that a booster rollout could reinforce vaccine-hesitant people’s view that COVID vaccines “don’t work.” I don’t share that worry. I think breakthrough infections are a more powerful antivax argument than boosters, so I think boosters will ultimately encourage vaccine uptake by reducing breakthrough infections.

But insofar as this is a serious concern, it should help to frame the booster controversy in terms of benefit rather than need. “Yes, primary COVID vaccine doses work. And a booster adds even more value. A booster is a benefit. If you’re content with the level of immunity already conferred by your primary doses, you don’t necessarily need the booster.”

Unvaccinated people often argue that they shouldn’t have to get vaccinated if they have already had COVID, and so have natural immunity. The usual response from public health professionals is that vaccination adds another layer of protection – which is undeniably true. Of course it is also undeniably true that boosters add another layer of protection to people who have been vaccinated. Nobody’s quite sure of the comparative immunity provided by infection versus vaccination; there are studies pointing in both directions, cited by advocates on both sides with equal overconfidence. Even more up in the air is the comparative increase in immunity provided by vaccination after infection versus booster after vaccination.

It is incomprehensible to me how anyone could claim that the scientific case for vaccinating previously infected people is so strong that their vaccination should be required, but the scientific case for boosting previously vaccinated people is so weak that boosters shouldn’t even be offered to them. But somehow many of the same public health professionals who rightly assert that all previously infected people would benefit from getting vaccinated turn around and insist that most previously vaccinated people don ’t “need” a booster.

---

The Rest of the Story

38. The actual booster decision process (part nine – the tide starts to turn).

As I said at the end of Section 22, the actual booster decision process was temporarily over on October 22.

The CDC said it would release “clinical guidance” the following week on who should get the same booster as their primary doses and who should mix-and-match. As of November 20 it hasn’t yet done so. Sticking to your primary dose is presumably the default recommendation for most people, unless there’s advice to the contrary. Maybe women under 50 should avoid J&J boosters because of a blood clot risk; maybe young men should avoid Moderna boosters because of a myocarditis risk; and maybe (this is the big one) J&J vaccinees should avoid J&J boosters because a different booster looks like it gives J&J vaccinees a lot stronger antibody response. We’ll see if the CDC ever decides to turn any of that into an explicit recommendation.

Of course both the FDA and the CDC have continued to monitor the data on waning vaccine efficacy and increasing breakthrough infections. And they were candid that if the data got more alarming, booster eligibility criteria might broaden (or as I would put it, might get more clearly and explicitly broad). The October 20 FDA news release announcing the new booster Emergency Use Authorizations, for example, included this quote from Peter Marks: “We will work to accrue additional data as quickly as possible to further assess the benefits and risks of the use of booster doses in additional populations and plan to update the healthcare community and public with our determination in the coming weeks.” By then some FDA officials were already leaking news that the agency would soon consider (explicitly) authorizing boosters for everyone over 40.

And on November 8, just as I was completing an earlier version of this column, the Washington Post reported blockbuster booster news. According to three anonymous officials, the article’s lede said, “Pfizer and its German partner BioNTech are expected to seek authorization for their coronavirus vaccine booster shot for anyone 18 and older.” That’s not the blockbuster part, which came in the second paragraph: “The request, which may be filed as soon as this week, is likely to win the backing of the Food and Drug Administration, said the individuals.”

The article went on to report that Pfizer declined to confirm (or deny) the news, and that Moderna had no comment on whether it too would seek a broader government okay for its boosters. It also noted that the FDA was unlikely to go back to VRBPAC for more advice. Apparently the reporters had no leaks from the CDC, so the article didn’t offer any prediction on how that agency or its advisory committee might respond. But the implication was clear that Moderna wasn’t about to let Pfizer boosters get a broadened government okay without asking for the same thing, and that the CDC wasn’t about to block the imminent expansion of booster eligibility if the FDA gave the required authorization.

Most interesting to me are the article’s explanations for why the FDA and probably the CDC were about to abandon eligibility restrictions (or in my judgment the façade of eligibility restrictions). There are three.

Mentioned first but only briefly is the fact that universal booster eligibility “would essentially fulfill the Biden administration’s booster-for-all-adults goal, announced last August amid concerns about waning vaccine protection.”

A bit more prominent is the point I keep stressing, that the existing eligibility restrictions aren’t really restrictive but nonetheless deter some people. “Authorizing boosters for all adults would be an explicit recognition of what is already occurring,” the article says. “People are getting access to the shots by saying they are in one of the recommended categories; pharmacies are using the honor system.” At the same time, universal eligibility “also would give the green light to people who have eschewed boosters because they don’t fit into any specific category – or think they don’t.”

But the dominant rationale throughout the article is that it’s looking more and more like boosters are “needed” after all. The lede paragraph calls universal eligibility “a move that could increase booster rates at a critical moment in the pandemic,” attributing that opinion to its anonymous official sources. The article cites several recent studies to the effect that vaccine efficacy continues to wane and that boosters are both safe and effective. And it quotes several members of VRBPAC and the ACIP about their openness to reconsidering the restrictions they voted for as recently as mid-October. For example: “Wilbur Chen, a member of the CDC immunization advisory committee who earlier opposed a broad approach to boosters, said new data means ‘it’s very fair game for us to revisit this question.’”

I wrote this paragraph on November 9: “If the Washington Post article is accurate, it seems likely that before Christmas President Biden will have achieved the universal booster eligibility he called for in August. And we all got to watch the sausage being made.”

39. The actual booster decision process (part ten – the tide has turned).

By the time the Washington Post article appeared, the tide of expert opinion was already turning. Of course a sizable minority was still skeptical about the value of boosters and the wisdom of putting much stress on their availability, much less making them available to a wider range of people. A November 9 article in the Annals of Internal Medicine, for example, argued that it was much more important to promote first doses, expand test access, and increase vaccine philanthropy.

This may still have been the view of a majority of public health professionals – but it was no longer what most top experts were thinking.

I became convinced that the tide was turning when I saw the expert response on Twitter and elsewhere to a November 11 New York Times article entitled “What We Know So Far About Waning Vaccine Effectiveness.” The article summarized a number of studies showing that vaccine efficacy was declining against infection. But its main focus is captured in its lede paragraph:

As tens of millions who are eligible in the United States consider signing up for a Covid-19 booster shot, a growing body of early global research shows that the vaccines authorized in the United States remain highly protective against the disease’s worst outcomes over time, with some exceptions among older people and those with weakened immune systems.

The article pointed out again and again that so far the vaccines are “still working” and “highly effective at preventing hospitalization.” It noted that “experts worry that a national focus on boosters will detract from what should be the country’s most important goal,” vaccinating the unvaccinated. It accurately stated that “experts have been divided over whether booster shots are necessary for those beyond the most vulnerable.” But it failed to note that the balance of expert opinion was shifting, and everyone it quoted was clearly on the anti-booster side of the debate. Moreover, the studies it chose to summarize left out some of the most recent and most alarming studies that were changing expert minds across the country.

For a sampling of the expert vituperation that followed, check out this Twitter thread. Writing mostly to each other, top experts called the article “misleading,” “incomplete,” “skewed,” “bad messaging.”

A month earlier I think this article would have seemed unexceptionable. But expert opinion was changing, and the Times was behind the times.

The Washington Post, on the other hand, was keeping pace with the change. Consider the lede of this Post article, published on the same date, November 11: “Anxious about a surge of coronavirus infections enveloping Europe as cases tick up in the United States, senior health officials in the Biden administration are pressing urgently to offer vaccine booster shots to all adults. But support for the renewed push is not unanimous.” Unlike the Times article, this Post article is carefully balanced, citing a wide range of expert and official opinions.

(The Times caught up quickly. On November 16, it took note of a “significant softening of opposition among public health experts.” On November 19, it published what might be its first unmistakably pro-booster article. Entitled “Boosters for all?,” it pretty much answered its own question in the affirmative.)

The November 11 Post article also included this important information: “Some state officials aren’t waiting for Washington to act. California state officials, worried about a potential rise in cases, on Wednesday urged all adults who are six months past vaccination to get a booster. In Colorado, where a surge of covid-related hospitalizations is stressing the health-care system, state officials unilaterally declared this week that everyone 18 and older was eligible for a booster if they waited the specified time after initial vaccination, and they ordered pharmacists not to deny anyone.”

California and Colorado were the first two states to decree universal booster eligibility. By November 18, just one week later, there were a dozen, including Kansas, New Mexico, Louisiana, and Rhode Island. New York City reached the same decision.

These state and local jurisdictions didn’t defy the CDC eligibility criteria. They interpreted them liberally – just as millions of booster-seeking individuals had interpreted them liberally; in my judgment, just as they were meant to be interpreted liberally. A Colorado executive order, for example, declared the entire state “high risk for exposure or transmission of COVID-19 and therefore eligible for the safe and highly effective COVID-19 booster shot.” And the head of the New York City public hospital system said in a November 15 New York Times article: “I view all New Yorkers, because of the density of our city, of being at higher risk.”

And insofar as these state and local governments were departing from the CDC’s recommendations, that was totally compatible with their respective roles. State and local governments normally hew to CDC recommendations, but in principle the power remains in their hands, not the CDC’s. The fact that the U.S. government owns all the COVID vaccine doses might have complicated the situation, but not with the White House squarely on the side of broadening booster access.

Even so, the impression of state and local rebellion was harmful in the view of most public health professionals. Peter Hotez put it this way in a November 20 Washington Post article: “The states made the right decisions, but the optics are awful in appearing to go rogue and undermining the federal agencies.”

The cart was getting ahead of the horse again.

40. The actual booster decision process (part eleven – universal booster access).

There were more than enough reasons for the FDA and the CDC to broaden booster access – or as I see it, to clarify that booster access was broad already. The three biggest reasons:

• The evidence was stronger in mid-November than it had been a month earlier that vaccine efficacy was waning; that breakthrough infections were increasing; that severe breakthroughs might be increasing too; that boosters had a good safety profile; and that all these things were true for everybody, not just the elderly and infirm.
• The pandemic was looking worse again. A number of European countries were on fire, as were a few U.S. states. National case counts had stopped declining, plateaued, then started to rise. As the weather got colder and the holidays got closer, the short-term prognosis was looking grimmer than it had a month earlier.
• The October eligibility criteria weren’t deterring anybody who really, really wanted a booster. But they seemed to be deterring a lot of people who really, really ought to get one. Booster uptake was slower than expected. Survey data showed that large numbers of people whom the experts considered prime targets for boosting weren’t sure they were even eligible.

The third bullet point deserves some additional explication. As I noted earlier, the October porous-but-seemingly-restrictive eligibility criteria worked pretty well from an individual liberty perspective. They allowed determined people to insist on getting boosted and they simultaneously allowed more lackadaisical people to decide not to bother. Both could cite the eligibility criteria in support of their decisions. Neither had reason to complain and neither was vulnerable to criticism.

But the criteria worked less well from a medical perspective. Determination to get boosted is bound to be more highly correlated with variables like education, class, and race than with actual vulnerability. So healthy, highly educated, upper-middle-class white people were getting the boosters they badly wanted, while less healthy, less educated, working-class people of color didn’t necessarily get the boosters they badly needed.

In September and October, most experts and public health professionals thought this was okay, mostly because they thought boosters weren’t very important anyway, since they judged that fully vaccinated people were adequately protected from severe outcomes. By November they had come to think that maybe it’s important to get vulnerable people boosted – which made it important to be clear that everybody is entitled to get boosted.

In other words, I think most experts and public health professionals are still uncertain and divided about whether it ought to be a priority – even a secondary priority – to provide boosters to young, healthy, fully vaccinated people. But the majority are now convinced that it is indeed a priority to get not-so-young people boosted, especially if they have comorbidities. And the majority are now convinced that the best way to do that is universal booster access.

But at least sometimes, universal access to encourage the vulnerable to get boosted is morphing into universal encouragement to get boosted. Here’s how the FDA’s Peter Marks put it in an interview: “I think this is pretty simple now: If you are over 18, and you have been vaccinated … it is time to go get a booster. Doesn’t matter which one you get, go get a booster.”

Once this change of heart had been accomplished, the denouement was astonishingly swift:

• On November 18, the FDA authorized boosters for everyone 18 and older, period. It framed the amended authorization as mostly about avoiding confusion. Peter Marks focused on “streamlining the eligibility criteria” to “ensure booster doses are available to all who may need one.” The FDA acted without consulting with VRBPAC.
• On November 19, the CDC convened a three-hour meeting of the ACIP. The advisory group voted unanimously to recommend that boosters be permitted for everyone 18 and older. That was the action item on its agenda. But it went further, voting – again unanimously – that boosters should be recommended, not just permitted, for everyone 50 and older.
• The main thrust of the ACIP meeting wasn’t really on new data or the worsening conditions on the ground. It was on simplifying the eligibility criteria in order to convince as many vulnerable people as possible to get boosted. The president of the Association of State and Territorial Health Officers told the ACIP members that every state in his organization favored the change. He added: “The current guidelines, though well-intentioned and thoughtful, generate an obstacle to uptake of boosters. In pursuit of precision, they create confusion.”
• That evening CDC head Rochelle Walensky endorsed the two ACIP recommendations.

Universal booster eligibility is now U.S. policy, just as the White House envisioned in August. How many vaccinated people choose to get boosted in the coming weeks remains to be seen. So does how getting boosted will affect their health and the course of the pandemic in the United States.

Something else also remains to be seen: How watching the COVID booster sausage getting made will affect the way Americans view the public health policymaking process. Harvard risk communication expert Jay Winsten thinks the impact will be harmful. “After months of confused and contradictory messaging, it’s baked into the coverage now and into the public psyche,” he told the Washington Post. “The way they have handled this has done real damage to the agency’s credibility with a lot of people.”

Unlike Winsten, I am hopeful that watching the sausage getting made will have a salutary effect. Not that I disagree with Winsten on what the effect is likely to be. People saw the public and the politicians taking the lead instead of following the experts’ lead. They saw how much uncertainty and disagreement underlay expert and agency pronouncements. Where I think we disagree is whether these are good things or bad things for the public to see.

---

What’s Wrong/Right with the Process

41. The two ways the booster decision process went “wrong.”

The booster decision chain of influence ran from public to politicians to agencies to experts – exactly the opposite of the ideal process in the judgment of most public health professionals. That’s one thing they hated about the process.

The other thing they hated about the process was its transparency. As a rule, public health professionals try to reach some kind of working consensus behind the scenes. Not secretly, exactly; their doubts, dilemmas, and disagreements are there to be analyzed if you look for them – especially now that experts spend a lot of time tweeting at each other about, well, their doubts, dilemmas, and disagreements.

But this time it was all vividly public. You didn’t have to go looking. Everyone got to see the public health policy sausage getting made. To most public health professionals, that transparency was almost as infuriating as the nearly intolerable fact that the public and the politicians were largely dictating the sausage-making outcome.

These two characteristics are exactly what I loved about the process. I liked the outcome too; I see precious little rationale for opposing boosters. But even if I disapproved of the outcome, I would still be delighted with the process.

The next two sections explain why.

42. Don’t follow the science.

I have long opposed the “follow the science” incantation. In 2001 I wrote a website column sarcastically entitled “Trust Us, We’re Experts.” I followed that up with another column entitled “Sound Science.”

My blurb for “Sound Science,” written twenty years ago, captures perfectly how I felt and still feel about claims that the public and policymakers should always defer to the experts: “The term ‘sound science’ is almost always a way of pulling rank, of harnessing the high stature of science on behalf of one side in a policy debate.” I went on:

This column exposes some of the pretenses that typically underlie the term: the pretense that your scientific support is stronger than it is; the pretense that your actions are grounded in science when they are grounded mostly in other considerations; and the pretense that your disputes with critics are about science when they are mostly about trans-scientific issues. The column is also about my clients’ tendency to believe their own pretenses – to forget that they are using or even misusing science to achieve their goals and imagine instead that they are science’s virtuous handmaidens.

Public health decisions should of course be grounded in scientific advice, but those decisions nearly always raise nonscientific considerations as well: among others, considerations of politics, ethics, communications, and above all values. To some extent, scientific experts are influenced by these nonscientific considerations; to some extent, they know they are; to some extent, they admit they are. But as I pointed out twenty years ago, they often pretend and more often genuinely imagine that their positions on nonscientific issues are also scientific – and therefore entitled to special deference from nonscientists.

Expertise is always a set of blinders. Deep knowledge isn’t broad. So even if scientists are willing and able to distinguish their scientific from their nonscientific judgments – which is rare – relying on the nonscientific judgments of scientists is a serious mistake. It’s not just that their expertise doesn’t extend as far as they claim. Their expertise actually gets in the way. Despite some stunning exceptions, scientific experts are notably poor at making nonscientific decisions.

Frankly, I’d just as soon scientific experts didn’t try. In my ideal process, experts in fields like epidemiology and virology render expert advice to political leaders (about policy) and to the public (about individual decisions). Other sorts of experts render advice in their fields of expertise as well. Political leaders and the public do their best to hear and understand everyone’s advice, and to integrate what makes sense to them into a coherent decision. In this schema agencies like the FDA and the CDC are hybrids. Sometimes they function as scientific experts, advising political leaders and the public; sometimes they function as political decision-makers, recipients of scientific and nonscientific advice, charged with integrating what they have learned into a coherent whole.

Insofar as I have a quarrel with the White House effort to push a COVID booster policy, it isn’t that the White House failed to “follow the science” or to follow the scientists. It’s that the White House ever pretended that that’s what it would or should do. One of the ways Biden distinguished himself from his predecessor, both as candidate and as president, was by repeatedly and explicitly saying that he wouldn’t disdain the experts’ advice the way Trump did. Promising to listen to the experts is fine. Promising (or seeming to promise) to always do what the experts tell you is awful. Actually keeping that latter promise would be more awful still.

When scientific experts complain that a politician isn’t “following the science,” what they often mean is that the politician isn’t following them, the actual scientists. Following the scientists is often impossible because scientists often disagree. Worse, they often hide their disagreements, so you could easily end up following the winning side that forced the losers to shut up. Even worse than that, scientists like everyone else have opinions that go beyond their expertise. Their expert judgments are affected by their inexpert opinions – and they often pretend or imagine that their inexpert opinions are expert judgments too. Getting ethics advice, communication advice, or policy advice from scientists isn’t “following the science.” Often it is getting bad, blinkered, inexpert advice.

The FDA and CDC vaccine advisory committees (VRBPAC and the ACIP respectively) are made up almost exclusively of two kinds of people: vaccine experts and clinical practitioners. Lots of other areas of expertise are missing from the table. A September 25 Washington Post article on the Pfizer booster decision quoted risk communication expert Jay Winsten’s point that the ACIP “includes infectious-disease specialists, obstetricians and pediatricians who grappled Thursday with questions they have no expertise in, such as whether offering boosters might undermine public confidence in the vaccines’ efficacy.” It makes sense to broaden these advisory committees by adding social scientists, including, yes, risk communication experts. Or it makes sense to confine these advisory committees to questions within their wheelhouse – scientific and medical questions. It makes no sense at all to seek the advice of science experts on non-science questions – or on policies with important non-science aspects – and to think that following their advice is following the science.

The same is true of the agencies themselves. If the FDA and the CDC confine themselves to questions of science, medicine, and public health, then they can’t be the ultimate decision-makers – simply because the ultimate decisions must embrace other concerns as well: liberty, law, politics, economics, etc. If they want to be the ultimate decision-makers, they must acquire and demonstrate much broader expertise than they currently possess.

So if you’re a political leader in the midst of a pandemic, you should definitely listen to scientists – the outliers as well as the majority – and do your best to distinguish their expert judgments from their inexpert opinions. Listen to other sorts of experts on other sorts of issues. And listen to the public, both because you work for them and because policies don’t work well if the public won’t accept them. Then filter all of that through your own political wisdom, your own sense of what the science says, and your own sense of whom to trust and whom to discount. Come up with a policy – listening to all but following none. Finally, take responsibility for what you decided. You’re not following, you’re leading.

Leana Wen got this exactly right in a September 28 Washington Post op-ed entitled “The booster shot debate shows that public health is not only about science. It’s about values.” I’m tempted to quote the whole thing, but here’s a bit of it:

The process was made far more muddled than it should have been, not because President Biden got ahead of scientists, as some critics claim, but because he didn’t set the expectation of who should be making this call.

As I’ve noted previously, the Biden administration has confused “follow the science” with ceding policymaking to scientific agencies…. [W]hen a decision rests on not just the science but also the interpretation of it, it’s entirely appropriate – and indeed necessary – for the public, through our elected leaders, to determine the outcome….

That’s not playing politics or disregarding science. Rather, it’s listening to constituents and respecting their values, which is what people elect leaders to do.

43. Don’t speak with one voice.

I am also a longtime opponent of letting experts hash out their disagreements behind closed doors – at least not when the issue at stake matters to the rest of us. Without transparency, the process by which experts “reach consensus” is all too often a process by which they enforce an artificial consensus, cowing the minority into silence.

Here I diverge from the views of most of my risk communication peers, whose mantra on the subject is “speak with one voice.” (Somehow they manage to pair this mantra with equally vehement support for transparency, as if the two ideals were compatible when experts disagree.) I think hiding or suppressing dissent is harmful when it backfires – when the minority viewpoint leaks out and the public starts talking about “coverup.” It’s even more harmful when it succeeds, misleading political leaders and the public about the extent of expert doubts, disagreements, and dilemmas; and depriving them of the chance to take the dissenters’ opinions into consideration.

For a detailed explication of my rationale for wanting expert dissensus to be visible when it exists, see this 2006 column, “‘Speak with One Voice’ – Why I Disagree.” Here’s an excerpt from the introduction:

The obvious problem with the “speak with one voice” advice is the frequent presence within the organization (or coalition) of more than one opinion. In such cases, speaking with one voice necessarily means suppressing discrepant voices…. The consensus view (no pun intended) is that this is not too high a price to pay for message consistency.

That’s the key question: whether it is wise or unwise for risk communicators to try to hide the diversity of opinion within their organizations. This can be partitioned into three questions – whether suppressing dissenting voices usually works; how much harm it does when it fails; and how much good it does when it succeeds.

If I were rewriting that column today, I would focus more on the harm that “speak with one voice” does when it succeeds. Hiding scientists’ disagreements hides scientific uncertainty. It causes the public to believe, wrongly, that the answer is obvious. When new evidence leads to a different answer, the public feels misled and learns to distrust the scientists, and maybe even science itself.

44. OMG, more to read!

In my view, both “follow the science” and “speak with one voice” have led to recurrent, severe defects in U.S. management of the COVID-19 pandemic. I have made these points in 50+ website pieces and media interviews since January 2020. Interested readers can consult any of the following, in which I point to some ways public health officials and experts have misrepresented the science, gone beyond the science, or stifled disagreement about the science.

The worst examples (and they are plentiful) are when public health officials and experts tell a “noble lie,” intentionally misleading the public about the science for what they consider a good reason; then find ways to gag or undermine those who try to dispute their dishonest claim, insisting that the dishonest claim is “the science” and nonscientists should feel obligated to follow it.

Below in chronological order are some of my articles, website columns, and interviews addressing how “follow the science” and “speak with one voice” have damaged COVID risk communication. For examples from earlier infectious disease outbreaks, consult my website’s Infectious Diseases Index.

Most public health professionals are unhappy with the booster decision process because the politicians didn’t follow the scientists and the experts didn’t speak with one voice. I am delighted with the booster decision process for precisely those same reasons. First we all got to launch the sausage-making process, and then we got to watch it.

Copyright © 2021 by Peter M. Sandman

---

Contact information page:    Peter M. Sandman