What does "I believe in science" actually mean?

Masks and the limits of scientific inquiry

Nov 03, 2020

You hear a similar point being made in lots of places: we should be “trusting the science” when it comes to COVID.

But what does that actually mean?

I’ve always found the statement “I believe in science” to be a bit odd. After all, isn’t belief something that requires faith? Why should science require faith?

It’s not like there’s a natural law of “science” that everything must obey. Science isn’t a belief system. Science is a process, it’s not an endgame. It’s about observing the natural world, testing hypotheses, and finding out what doesn’t work. Science is an asymptotic quest for the truth; it’s not a search for eureka moments.

And translating scientific observations about the natural world into public policy is never straightforward. There are imperfections in our experiments, there are tradeoffs in real life. There’s never a perfect answer.

I wrote a newsletter a few weeks ago about mask deniers because it’s a perfect example of the disconnect between the “I believe in science” crowd with the folks who want to Fire Fauci.

Today, I wanted to revisit what we know about masks, with the idea of getting into the science a bit to think about the limitations of knowledge and, more importantly, where “following the science” falls short with regards to policy implementation.

First, a quick digression on why masks are even a thing

The SARS-CoV-2 virus that causes COVID-19 isn’t floating around the air on its own waiting to infect us. It’s also not sitting around on surfaces waiting for us to pick it up (which is why a lot of the surface cleaning that you see is just pandemic theater).

The virus spreads through our respiratory secretions. Take a look at this comment from a recent article in the Journal of the American Medical Association written by, among others, Anthony Fauci:

SARS-CoV-2 is primarily transmitted by respiratory droplets exhaled by infected individuals; these droplets span a spectrum of sizes. Larger droplets fall out of the air relatively quickly while close to the source, usually within a 6-foot distance. Smaller droplets, often referred to as aerosols, are also present at close range but may remain in the air over time and greater distances, decreasing in concentration as they move outward from their source.

So because respiratory droplets carry the virus and masks block some respiratory droplets, masks reduce COVID spread.

That makes sense, so what’s the argument against masks if it’s so obvious?

There are a lot of objections to masks, some in good faith, and some in bad faith. We’ll stick with the good faith ones for this newsletter.

From what I’ve read, the two biggest concerns that people have are 1) about the safety of masks; and 2) about how effective they are at actually reducing the spread of the pandemic (and no, I didn’t already answer that question with the link above).

There are also the concerns about social interaction, especially with children, and the associated psychological ramifications, but the people making this particular argument don’t seem to be as concerned with whether masks work as they are with at what cost do they work.

I’m sure that almost everyone reading this has worn a mask by now.

I think we can all agree that they’re uncomfortable. If you’ve worked out in a mask, you’ve probably noticed that masks make exercise feel harder as well.

But is there any chance that they actually compromise our respiratory status?

And are they actually going to prevent us from getting COVID? After all, aren’t 85% of people with masks getting infected?

Let’s take a look at some science to figure out what we believe

There’s been a lot of fear-mongering about masks lowering oxygen levels.

A recent study in JAMA attempted to test the the hypothesis that masks lower oxygen levels.

We’ll start with the authors’ conclusion: masks don’t lower oxygen levels.

But, this isn’t politics where we can just see a headline and click retweet. In science, you can’t be satisfied with the conclusion that the authors tell you to believe. You need to dig into the details of the study to actually understand it.

In this study, the authors took 25 Canadian adults, gave them a non-medical face mask and a pulse oximeter, and told them to self monitor their oxygen saturation before, during, and after wearing masks.

They found that participants wearing masks actually had their oxygen saturation levels increase, from 96.1% before wearing a mask to 96.5% while wearing a mask, and then 96.3% after wearing a mask.

So what does this tell us?

Honestly, not a whole lot. I am pro-mask but I am also pro-not-exaggerating-the-conclusions-from-your-research, so even though I want to crow that this “proves” masks don’t compromise anyone’s respiratory status, this study doesn’t actually provide us with that assurance.

Why do I say that? There are a handful of reasons, of which I’ll just list a handful:

We can’t extrapolate these findings to people with lung or heart disease, since these patiets were excluded
We also can’t extrapolate these findings to medical masks, since that’s not what the participants used
The people in this study weren’t doing any type of exercise, so we can’t draw conclusions about the safety of masks while working out
Respiratory status isn’t just about your oxygen levels, it’s also about your ability to ventilate, i.e. exhale carbon dioxide. We don’t know what happened to the carbon dioxide and whether increased work of breathing through a mask impacted this. Plus you’d need to draw blood to test this, which the researchers didn’t do
We don’t know whether respiratory rate increased or whether work of breathing increased. Maybe the participants needed to breathe much faster or started using their work of breathing increased to the point that they could potentially tire out if wearing a mask for longer

And this is just a small sample of the questions that are raised with even a brief critical appraisal. To say nothing of the leap of faith that you take when extrapolating conclusions from a sample of 25 Canadians to tens of millions of others throughout the world.

So what does it mean to “follow the science” here? I honestly don’t know because the science is very limited in this example.

Exaggerating the certainty of the research lends itself to losing trust with the population

The headline from the article above is easy: “Masks don’t lower oxygen levels.” But as we’re getting into, the nuance of conclusions that can be drawn isn’t as straightforward.

Let’s look at another example.

This article with it’s nifty visualization of COVID transmission off last week as a way of describing the importance of masks to reduce viral spread. I encourage you to click and take a look for yourself. Here’s a screenshot, it’s quite easy to understand and the pictures are compelling!

So what’s the problem you ask?

You can quibble with some of the statistics, but the biggest issue here is that the authors from this article overstate the importance of aerosol transmission (viral particles indirectly transmitted from hanging around in the air) and don’t directly report the much larger importance of proximal transmission (viral particles directly transmitted from a person in close proximity).

If you do the math on their numbers, like ID physician Andrew Morris did in this Twitter thread, it’s clear that the authors exaggerate the likelihood of spread in different indoor scenarios from aerosol transmission and implicitly underplay the importance of close mask-less contact.

Is this really a big deal?

Well, not to me. I think ventilation is good, that we should wear masks when we’re in public places indoors, and that keeping a few feet of physical distance is a good idea when socializing.

But if you’re a mask skeptic, an article like this can feel like propaganda rather than science.

What’s the real argument for masks then?

The real argument for masks isn’t any one article about the relative impact of masks in any single situation.

Masks are a multiplicative intervention. If you reduce the likelihood of transmission by even 10% for any one person and then you multiply that benefit across hundreds of millions, you start to bend a pandemic curve (not flatten, because I think the time has come to retire that metaphor).

Masks are compound interest for our society’s fight against the virus.

They’re among the most powerful tools that we have. But I worry that when we make weak arguments for them or exaggerate findings from small, clinically useless trials that we undermine trust and increase the risk for all of us.

Gregory Katz's Substack