We're all doing our own research

We're all doing our own research

It would be nice if the CDC's advice were more accurate and useful

Matthew Yglesias

As the vaccines for kids have rolled out, several people I know have gotten their 4-year-olds vaccinated, even though the FDA has only approved the vaccine for ages 5-11. It turns out they are not asking kids to show photo ID when they get their shots, so if you want to break the rules and get a 4-year-old vaccinated, you can. I assume there are limits to this — a baby is obviously not 5 and nobody’s going to let you get away with pretending that she is — but the point is, you can get away with stuff. And so my friends did.

Were they right? Well, I think that they were. And they came to this conclusion by doing their own research.

In the circles I travel in, that phrase is almost always a joke poking fun at anti-vaxxers. But it’s what vax enthusiasts skirting the rules to get doses for 4-year-olds are doing, too. And more broadly than that, we are basically all doing our own research during this pandemic because the prescriptive public health guidance is not, and has never been, informative or accurate enough.

(Ethan Miller/Getty Images)

Doing your own mask research

I don’t really favor mask requirements, but I follow the rules where they are in place, and if I am going to wear a mask in order to reduce the spread of Covid-19, then I feel I ought to wear a mask that works well.

That usually means a KN-95 mask. They’re easy to carry, comfortable enough, and have a good fit on my face.

But how do you find out which masks are the good ones? You really have to do your own research on this one. The CDC’s mask advice features some allusions to mask quality, but it’s incredibly vague, and based on the pictures, they seem to be envisioning people primarily wearing cloth masks.

Here’s the thing about cloth masks, though. Remember that big study out of Bangladesh showing that encouraging population-scale masking is effective at halting Covid? You can read about it in The New York Times and The Washington Post and elsewhere. One of its key findings is that surgical masks are much more effective than cloth masks, even if the surgical masks are handled fairly roughly and reused.

We also tested the filtration of surgical masks that had been worn, crumpled up in pockets and purses, and washed with soap and water up to 10 times. These masks still prevented more virus particles from passing through than typical cloth masks. Masks with even better filtration or fit than surgical masks, such as KF94 or KN95 masks, may provide even stronger protection than surgical masks if worn properly.

Unfortunately, we do not have large-scale study evidence on the efficacy of KF94 and KN95 masks. But we do have engineers like Aaron Collins who have experimented on the basic physics of mask filtration and shown that these kinds of masks really do filter much more effectively than cloth or surgical. And that conclusion is echoed by reputable journalists in places like Scientific American.

But precisely because this is me doing my own research, my confidence is not as high as it ought to be. Ideally, the NIH would have funded a bunch of mask-related studies over the past year to give us really clear and convincing evidence about mask-efficacy in a range of scenarios.

I think the lack of emphasis on mask quality represents the continuing fallout of the public health agencies’ original opposition to promoting masks at all. When they flipped on this, we were still dealing with PPE shortages that they didn’t want to exacerbate by encouraging mass hoarding of surgical masks. So at the time, they went all-in on the virtues of cloth masks as better than nothing, really pushing “wear a mask, any mask.” That mask flip-flop was costly to their reputation, so re-revisiting it to say “wear a high-quality mask after all” is something they don’t want to do (and that might be perceived as confusing), so we just stick with advice designed for the circumstances of spring 2020 even though the situation has totally changed.

Doing your own vaccine research

Some of my friends who got Pfizer for their initial vaccination went out of their way to schedule Moderna as their booster. This made sense to me since, if you do your own research, the Pfizer —> Moderna sequence is second only to the Moderna —> Moderna sequence in efficacy. Moderna —> Pfizer and Pfizer —> Pfizer are still really good, but definitely inferior to Moderna.

I’m in the Pfizer Pfam, and I’m opting for Pfizer for my booster. This is first and foremost because an appointment was available at a convenient time and place. But I’ll also say that while the side effects I experienced from my Pfizer shot were short-lived and ultimately harmless, it did hit me like a ton of bricks on that first day. And based on having done my own research, it seems that serious fatigue and muscle pain are more common as side effects with Moderna.

My suspicion is that the Moderna combo of higher efficacy + bigger side effects is driven by the fact that Moderna gives you 100 micrograms of vaccine versus only 30 for Pfizer. Indeed, the Moderna shot is so big that for booster purposes, they only give you a half dose. But you don’t need six years of medical school to unearth the fact that 50 is still bigger than 30.

These dose sizes were arrived at through a fairly arbitrary process. Moderna chose to design a Phase 3 trial around a 100 microgram dose. It seems very likely that had Moderna chosen to design a trial of 25 microgram doses, that would have passed the safe and effective test too. But in that universe, studies might show Moderna as slightly less effective than Pfizer.

There are plenty of scientists who’ve done studies on fractional dosing, primarily because they are interested in stretching supply to meet demand in poor countries. But America’s public health institutions are too inflexible1 to fund these kinds of things and adjust domestic dosing strategies. They also don’t want to discourage vaccination by talking about side effects, so they don’t tout the reduced side effects of Pfizer. And they don’t want to encourage wastage of Pfizer doses by having everyone clamor for Moderna, so they don’t tout the improved efficacy of Moderna. Instead, the CDC’s website says you just should do whatever — so you need to do your own research.

Probably the most egregious version of this was around mRNA boosters for people who got an initial Johnson & Johnson dose. As Arin Dube notes, statistically literate people saw there was clear evidence for this long before it was officially recommended by the CDC or authorized by the FDA, and so those in the know (including several of my friends) just went and did it.

The point is that in broad culture war terms, the people using good masks and diligently getting their boosters (or vaxing their 4-year-olds) are the people we associate with slogans like trusting the experts. But in fact, the Covid hawks just as much as the Covid doves are circulating pre-prints, looking at weird foreign studies, and in practice ignoring the major American public health institutions because they don’t produce information that is timely, relevant, and clear.

Doing your own research on Thanksgiving

This basic dynamic is unfortunately not limited to Covid-19.

To take a somewhat trivial example, the CDC says you should cook turkey breast meat to 165 degrees, which will make it dry and stringy. Kenji López-Alt, by contrast, recommends that “for turkey that is moist and juicy, aim for breasts that register 150°F in their deepest section and legs that register at least 165°F.” Is that unsafe? If you read López-Alt’s cookbook, “The Food Lab,” he explains that there is no magic temperature at which food goes from unsafe to safe. What happens instead is that bacteria are killed by a combination of temperature and time. The CDC, for simplicity’s sake, is picking a temperature level at which you generate something like an instant kill. But if you have a lower temperature for a longer time, it’s equally lethal.

In other words, the CDC isn’t just being conservative here (the way they are when they say you should never eat runny eggs); they are obscuring the nuances of the underlying science in order to generate simpler and more unequivocal recommendations.

I think CDC food recommendations provide important context for understanding their coronavirus recommendations because it’s such a slow-moving area of science. The CDC has had many decades to consider and refine what it wants to say to the public about cooking Thanksgiving turkey. It’s not a question of the science being unclear, it’s a question of the agency’s philosophy about what to say. This gets particularly complicated if you’re talking about trying to cook the bird with stuffing inside it. The CDC notes that the stuffing itself needs to come all the way up to safe poultry temperatures. But that leads to overcooked meat. So what’s the fix? Here’s López-Alt:

I tried cooking a stuffed turkey using my standard Easy Herb-Rubbed Turkey method and baked it until the stuffing reached the requisite 145°F before pulling the whole thing out and allowing it to rest. By this stage, the breast meat of the turkey was at around 155°F near its center, and all the way up at 180°F on its exterior layers. Needless to say, it was dry as bones.

So what's the solution?

It's actually quite simple, and even Alton [Brown] himself has gone back and recommended a similar method since his earlier disdain for stuffing: Just heat the stuffing before you put it in the turkey.

Now I can see the CDC saying this is all too complicated. People need simple directions. Wear a mask, any mask. Take a vaccine, any vaccine. Cook your poultry to 165. But people who actually care about cooking or Covid or whatever else are going to inevitably find themselves doing their own research.

Public health for an educated public

If you read coverage of the polio vaccination drive from the postwar era, compliance with officialdom was much higher back then. Some of that was surely the prestige the state garnered as a result of winning World War II.

But I also think it’s relevant that back then the population had a much lower level of education (the average American hadn’t finished high school) and very limited access to information or communication technology. The public health experts of the 1950s could successfully execute a strategy based on priestly mystique and information control. All you really needed to do to absolutely dominate the messaging space was to persuade three television executives to play along.

Simplifying things for public consumption was a strategy that could actually work. Yes, you’d face a trade-off between the quantity and nuance of information provided and the simplicity of the message — but the choice really was up to public health officials.

Today the population is much better educated, and we have a much larger population of people who are comfortable working with and discussing quantitative information. We also have many, many more channels for the dissemination of that information. And because of web search, a Scientific American article about mask quality isn’t something that runs once and is swiftly forgotten. It’s something that anyone who gets curious about mask quality and decides to Google it will swiftly uncover. So in my little corner of the universe, the Moderna > Pfizer result is something that everyone knows and many people act on, even though it’s not in the official guidance.

It seems to me that this approach is not actually working well to achieve its intended aims. One would like to tell the anti-vaxers that they should just listen to the public health agencies. But it’s not just that the public health agencies have proven themselves to be fallible (everyone is fallible), but that the people who are most amped-up about Covid are also ignoring the public health agencies in favor of doing our own research. Their headline recommendations on schools, for example, allude briefly to “ventilation” but don’t say anything specific about HEPA air filters, which you can easily buy on the internet and plop down in classrooms, creating a mitigation layer that causes no disruption whatsoever to the basic function of the school.

So on ventilation, on masks, on vaccines, and on everything else, despite the putative benefits of simplifying the message, nobody can actually use the appealingly simple message “go to the CDC website and do what they say.” And I don’t think that getting increasingly mad at people who do their own research poorly is a real answer to the problem.

1

Note that the lack of flexibility is not always the fault of agency personnel. This story about syphilis killing babies makes the case that the way Congress allocates money to the CDC makes them incapable of responding to new outbreaks with sufficient flexibility.