Killing viruses with light, with Jacob Swett

This week, I'm joined by Jacob Sweet, the executive director of Blueprint Biosecurity, a non-profit focused on achieving breakthroughs in humanity’s ability to suppress pathogens. We discuss far UVC, the deployment of special ultraviolet lights in particular locations, as a method for controlling the spread of infectious disease, analogous to boiling (or otherwise treating) water to kill waterborne pathogens. We discuss why it hasn't happened yet, why it probably will happen in your lifetime, and what humanity can do to accelerate.

Sponsor: Mercury

This episode is brought to you by Mercury, the fintech trusted by 200K+ companies — from first milestones to running complex systems. Mercury offers banking that truly understands startups and scales with them. Start today at Mercury.com 

Mercury is a financial technology company, not a bank. Banking services provided by Choice Financial Group, Column N.A., and Evolve Bank & Trust; Members FDIC.

Timestamps

(00:00) Intro
(00:31) The importance of indoor air quality
(01:29) Technologies for cleaner air
(02:31) The promise of Far-UVC
(03:10) Impact of COVID-19 on air quality awareness
(04:11) Understanding Far-UVC light
(06:44) Applications and benefits of Far-UVC
(16:42) Challenges and adoption of Far-UVC
(20:40) Sponsor: Mercury
(21:53) Challenges and adoption of Far-UVC (Part 2)
(23:19) Cost and benefits of Far-UVC
(26:40) The broader impact of respiratory pathogens
(29:41) Rediscovering the world for better health
(30:05) Historical perspectives on infectious diseases
(30:57) The role of sanitation and antibiotics
(33:02) Miasma theory and airborne transmission
(34:59) Impact of World War II on disease research
(38:06) The evolution of public health priorities
(42:39) Future of Far-UVC Technology
(46:03) Challenges in implementing Far-UVC
(56:47) The importance of rigorous studies
(01:00:33) Wrap

Transcript

Patrick: Hi everybody. My name is Patrick McKenzie, better known as Patio11 on the Internet, and I'm here today with Jacob Swett. Jacob is the founder and executive director of Blueprint Biosecurity, a nonprofit dedicated to achieving breakthroughs in humanity's defenses against pathogens.

Thanks very much for coming on the program, Jacob.

Jacob: Pleasure to be here.

The importance of indoor air quality

Patrick: One of the things that I think will most change our built environment over the course of the next couple years is the widespread deployment of technologies to—not quite "boil the air"—but I like conceptualizing it as boiling air in the same fashion that we've been boiling water and doing more complicated things to water quality for the last couple hundred years.

That approach has been great for humanity—much less typhoid, etc. For whatever reason, we have largely decided not to do this with indoor air quality. But that might be changing. Can you tell people who might not be super familiar with this field what interesting things are happening?

Jacob: It's an excellent point. When you look back historically at some of the biggest reductions in morbidity and mortality—ways that people have been harmed and die—it all comes from the sanitary revolution. A big part of that was making the water we drink very clean, and for reasons we could get into, air was kind of left out of that.

But it isn't because we don't have different types of capabilities now. When people think about what we could do to get the air we breathe as clean as the water we drink, a handful of technologies come to mind. One is just better ventilation, bringing fresh air into buildings. That's a good option in a lot of situations. It's not always great if you have dirty air outside, so you have to make sure you filter it.

Filtration is indeed the next one we talk about—portable air cleaners. Some people will call these HEPA filters. Sometimes they are, but there are different ways of filtering the air.

Beyond that, we get into some of the newer technologies, at least for air. Things like germicidal UV, where you can put different types of light into rooms that can inactivate pathogens. And sticking with the theme of "we've done this with water"—UV has been used for a very long time to actually ensure that the water we drink is pure and fresh. But we haven't done as much applying it to the air. It does have a history, which is pretty interesting.

And increasingly we're getting new capabilities.

The promise of far-UVC

One of the technologies that we're really excited about is a capability called far UVC. Collectively, we think there is a ton of promise with these technologies. If you were to combine them together and put them into the world around us—the built environment where we spend around 90% of our time—we could get to a world where those colds and flus and other things that drag us down, make our lives get cut short in some instances, and can lead to things like COVID-19, are potentially a thing of the past. That's what really excites us about some of these different technologies.

Impact of COVID-19 on air quality awareness

Patrick: COVID-19 was definitely a bit of an eye-opener for me. There were many things learned over that experience. One of them was that a random side effect of other non-sustainable "non-pharmaceutical interventions" we were doing was that there are four strains of the flu that circulate around, and I believe one of them is named after a place in Japan, the Yamagata strain.

We accidentally caused that strain to go extinct over the period of COVID-19 due to social distancing measures and other things. We definitely wouldn't want to be doing that level of investment in infection control in the typical year—I don't know if society would survive it—but it's interesting to think: wait, is it possible that the flu is optional?

And if the flu is optional, at what margin should we choose living in a world with the flu versus at what margin should we choose "No, wait, if it's just a matter of writing a check for a hundred billion dollars, let's live in a post-flu world and have that be an anachronism that our children study in history books and literature, much like, say, typhoid."

Understanding far-UVC light

The far UVC—how is that actually different from typical ultraviolet light just mechanically, chemically, etc.? And then we will go into why this is an option for us now.

Jacob: It's a great question. I think the place to start is when you think about light, everyone's familiar with visible light, which comes in different colors. These different colors correspond to different wavelengths—that's how physicists talk about them. And if you go beyond violet in the visible spectrum, you go to ultraviolet, which we call UV.

I'm sure all your listeners have heard of this. Everyone's been warned about UVA and UVB. There's a third type called UVC that we don't typically get down on Earth. It is stopped by the ozone layer. Actually, if you remember the scare about the ozone layer disappearing, the concern was that this UVC would get everywhere.

[Patrick notes: The 80s and 90s spent a lot of educational and cultural calories instructing children that the hole in the ozone layer would lead to all the cancers and there would be no place to bury anyone because we would run out of landfill space. We fixed the ozone problem, likely through some combination of the Montreal Protocol and exaggerating less. On landfills, uh, that was just made up.

I view this as a combined reminder that humanity can actually accomplish substantial positive impact against tough collective action problems and also that sometimes institutions just confabulate for a few years. Those are both useful things to remember.] 

Jacob: UV light can be conceptualized as having different colors as well because there are different portions of wavelengths. The type of UV light that we are most excited about is a special type called far UVC. Far UVC came about because, as everyone knows, UV light can damage DNA and do all these harmful things. Scientists went and kind of fought about it and said, "Well, what would it take for that to not be true for humans?"

What they figured out was that given that we're covered in skin, and skin is filled with proteins that make us up, if we pick a wavelength of light that is heavily absorbed by proteins, then when that light comes down on us, it will all be absorbed in our skin and not get down to our DNA and do any damage.

And that wavelength of light is far UVC. Basically, by picking this wavelength, all the things that you've heard about UV being bad in terms of causing harms applies to germs because they're much smaller compared to our skin layer. When it comes down to our skin, it doesn't go through and looks safe for us.

Obviously, the actual details and the science of it are a bit more complicated than that, but that is the kind of promise of this special type of light called far UVC. There are other types of UVC too. You typically call the one that's been used much more commonly up until now "conventional UVC."

Conventional UVC can be used in buildings as well, but you can't shine it on people's skin or eyes because it can be painful if you do that.

Applications and benefits of far-UVC

But this new type of far UVC is one that really looks like you could potentially put it into a room, stop germs from transmitting, and this could really be a game changer to our relationship with what we call respiratory pathogens.

Patrick: When we were discussing earlier that in many cases the air outside is cleaner than the air inside, one of the reasons is that there are a few things that could be bad about outside air. There could be environmental pollutants that are not pathogens themselves. For example, your air could evolve into a chemical mix which is not great for humans. If you get carbon dioxide poisoning, which is an episode I should probably have in the future at some point, as that is wild—that is a thing that we haven't figured out yet.

[Patrick notes: If you work from home or have a confined living space, I strongly recommend getting a CO2 meter. I use the Aranet 4, chosen by outsourcing the decision to commercial spaces I’ve spent a lot of time in that have put more cycles into considering this than I have. You can physically feel the difference in mental accuity between a reading of 800 and 2,000, and you don’t have to do anything wrong to cause a room to go to 2,000, and you only have to open a window to get it back down again.]

Patrick continues: But one of the reasons that the air outside is typically "cleaner" than the air inside simply is the air outside is exposed to the sun at least once a day or so in most places. And germs hate the sun. Sunlight is the best disinfectant—it's both a famous bit of US legal writing but also not quite strictly speaking medically true, but it's far truer than most things that a lawyer would say about pathogen control.

[Patrick notes: I always understood this bromide to be in a Supreme Court decision but it turns out it was actually in other writing by a (later) Supreme Court Justice, Justice Brandeis. The full quote, which I hadn’t run into, is amusing in the context of this discussion:

Sunlight is said to be the best of disinfectants; electric light the most efficient policeman.]

Patrick continues: So just the notion that we can create artificial suns indoors and greatly decrease the amount of pathogens is very interesting. Also, we don't need to have it in all of the rooms in all places.

One of the things that some of us unfortunately had to understand the science about in the pandemic era was there's some coefficient of reproduction of the various pathogens that are running around.

When that coefficient gets above one, humanity suffers a lot, and when it goes below one, humanity suffers much less. And interventions at the margin to get something from 1.02 to 0.98 would matter a lot, even if they were only deployed in a select number of commercial spaces.

As of the state of the science right now, if one were prioritizing spaces that humanity has access to for deploying this, where would you want to deploy it first?

Jacob: It's a good question. If you ask different groups, you'd probably get different visions for what a world of deployed far UVC would look like.

Our best guess, though, is that you would want this widely deployed in high-risk environments. Epidemiologists would tell you these are where there are large numbers of people spending a lot of time together and where people are interacting outside of their usual networks, where you can get that spread if you kind of imagine the contagion in your mind.

When you model it out, that's where you get the greatest returns on investment. This probably looks like places like classrooms, gyms, restaurants, conference venues, public transit, houses of worship, lecture halls, movie theaters, long-term care facilities—you name it.

I think places like gyms and restaurants are particularly important. We got great data on this during the pandemic. Whether you're breathing, talking, or exercising really changes your infection risk. Your listeners might appreciate this, but generally the infection risk is proportional to how much you're breathing.

So if you're in a gym, it's actually much higher likelihood that you can transmit and get an infection because you're getting your blood pumping, you're breathing a lot, you're actually emitting much more infectious aerosol. And if you're breathing a lot more in, you're doing that as well. Sometimes this can be more than 10 times higher in terms of infection risk.

Restaurants are another interesting example too, because we know that when you're breathing, speaking, or singing, you emit different amounts of aerosols as well. So if you get into a really loud restaurant and someone who's infected—in that period that everyone saw with COVID when you're infectious but don't quite feel sick yet—you can emit a lot of respiratory aerosols.

So these are great candidates for technologies like far UVC and others like Portable Air Cleaners. A pet peeve of mine is that loud restaurants are not only annoying and frustrating to be in, but they're probably bad for spreading infection as well. Putting in things like noise dampening or not playing really loud background music could actually have really huge benefits in terms of transmitting infectious disease.

Patrick: We as a society get to run a sort of portfolio strategy on these kinds of things. And one of the things that makes me quite bullish on it is that of the "non-pharmaceutical interventions" (NPIs) that society has access to, some are extremely disruptive to things that we value.

For example, in the event of a pandemic, you can tell houses of worship "please close for a multi-year period," which we did in many places. And I think that was maybe the right call at some times, probably the wrong call at others. But it's enormously costly for society versus "let's simply choose to embed these things in the built environment in a place that makes sense."

And after a few weeks of "what is the new thing on the wall," no one is ever going to care about them again.

Speaking of which, what actually is the new thing on the wall? If I walk into a room that has one of these for the first time in my life, am I likely to notice it?

Jacob: Probably not. It would probably be in the ceiling, kind of recessed in. You can't see UV light sometimes. Some of the lamps, especially the main type that they use now called krypton chloride—because of the physics of it, the krypton emits a pretty cool purple glow. So you might notice that if it's one of those sources.

But otherwise it would just be up in the ceiling, shining down invisibly, basically inactivating pathogens while you go about your day. And you really wouldn't see much or hear much or feel much. That's one of the benefits of these UV lights.

With some other types of interventions, if you have tons of ventilation, which is generally good, but if you get to really high levels like you've seen in an operating room, you probably do get like a wind tunnel effect. Or with portable air cleaners, which again, we're really excited about and recommend, but if you get a lot of them in a room, you just start to hear them, and it can be difficult in terms of talking. We're seeing this challenge as portable air cleaners are put in classrooms.

So you don't see much, which is nice. It's kind of related to the fact that you also don't see the pathogens either, which is part of the problem as well. You just see a little recessed fixture that maybe looks like a lamp that isn't on in the ceiling.

Patrick: There's a sort of spectrum of safety issues associated with the spectrum of light. As you mentioned earlier, many of us have received wisdom from when we were children: "Don't look directly at the sun. That will burn out your corneas," etc. But one of the reasons that we're choosing this particular kind of light is that this light that appears to be non-functional, if you happen to look up at it, you're pretty much fine? It doesn't actually hurt your eyes?

Jacob: That's right. For skin and eyes, based on all the currently available data, there's no evidence suggesting that there's any effects that cause significant short-term or long-term harm with this. This has been studied a fair bit.

When it comes to things like skin, we've exposed mice for over a year—basically eight hours a day, every day of the week. We've had researchers expose themselves to levels far beyond daily guidelines. We have models of the skin and the eye that are looking at this. And there's also just a lot of studies with regular people in different environments going about their day, and we don't see any harms there either.

It is probably important to note, though, that we think there probably is more research that should be done. Our goal is to see this widespread. We really want to see something that can put a dent in annual influenza and flu, and something that we're really passionate about is ensuring that we can prevent the next pandemic. So we want to see it widespread.

And when you go that widespread, you are talking about large diversities of people with different skin conditions and different backgrounds. So you really want to make sure that you're covering all of your bases. Our position at our organization is that to gain further confidence in this technology, we'll want to conduct more studies, particularly long-term safety studies.

You can try to do this mechanistically and try to say, "Well, we don't think it's going to cause anything," but there will be people out there who will say, "Well, let's wait the time and see." So our position is, let's start those studies now so we can get that data sooner rather than later. We want to really make sure that everyone everywhere is convinced as well that far UVC is safe for them.

Patrick: There's an unfortunate bit of—I don't know if it's bioethics or an aesthetic reaction around these things—but the precautionary principle with respect to making big changes in the environment. One can sort of understand given history how we ended up in this place, because the decades of spraying dioxin on the road were not great for people.

But we've essentially ended up in a situation where some people are very comfortable saying things like, "Wait, it's a huge change in the world if you are introducing UVC into classrooms. We would need decades and decades of studies before we reached enough confidence to allow that to happen." But no one requires decades and decades of studies to allow several hundred thousand people next year to die of the flu. That's just accepted as the normal, natural state.

And again, we have far more agency here than we appreciate. Which brings me to the next question: I'm assuming if one of these devices is just a lamp you put in your ceiling, then presumably they can be mass-produced in a factory and, without loss of generality, in China, put in a shipping container, sent to the United States, and then you can install them for, I don't know, a couple thousand dollars each.

Given that the cost of illness is much greater than the cost of a couple thousand dollars, why are we not yet seeing companies or institutions on the bleeding edge deploying these widely even on an experimental basis?

Challenges and adoption of far-UVC

Jacob: It's a great question. And just back to that point, which I think really is worth emphasizing: there is this real problem that if you keep the status quo the same, you are not blamed, but if you do something different, then all of a sudden it's your fault. That is a real big problem here.

With far UVC, I'll talk about why it's not being deployed more quickly. We can talk a little bit about cost too, if that's of interest. But like with many other problems, and I suspect this isn't a surprise at all for you and your listeners, there doesn't appear to be one single barrier.

It's probably also worth noting that the way we think about this is that it is being adopted. People are buying them and using them. So we think about it much more in terms of the question of "Is it happening fast enough?" We believe absolutely not. There's H5N1 right now, and the base rate for pandemics is like 3% a year. And annual deaths from influenza are a tragedy we shouldn't stand for. So it isn't really happening fast enough.

The reasons why, I think, are a handful. There are historical reasons around both UV and airborne transmission that mean we've kind of neglected this. UVC, when it was first being studied, had its heyday unfortunately right before penicillin was scaled up as a technology. Not surprisingly, people were like, "Well, we could put this light in this room and kind of statistically check whether people get infected or not and study a place, or this person gets sick and we give them this magical pill and they feel better."

People thought, "Well, I think maybe this pill is the future"—pharmaceutical interventions. And we've seen some of the limitations of that with antibiotic resistance, and we don't have vaccines for things like HIV, so there's been a bit of a shift, but that was a big historical part of it.

To some extent, it's hard to tell when it's working. It is this magically invisible light that you don't feel or anything. It's kind of like a mask: you wear it, you feel protected while it's on, but then you take it off and you get sick later and you're like, "Wait a second. I got sick, but I wore a mask." People kind of conflate these things, so that can be a bit of a challenge.

There's no regulatory body for it. You might at first think, "Oh wow, this is great. The FDA is not in my business for this." But it actually presents some pretty interesting challenges in that there's no defined arbiter of whether this is safe and efficacious or not.

Like a lot of things, some people want more evidence that it works. Some of this is reasonable. Some people have high standards. Some of it is like actuaries saying, "Well, maybe we'll pay for it, but can you tell me actually how much it will save and what it would do?"

Some people have safety questions, which I think is also quite reasonable. You grow up being told, "Put on your sunscreen, UV is dangerous." And then scientists come along—especially some of the scientists that, not these ones, but scientists writ large that changed stories during the pandemic—and then they say, "Well, suddenly this is safe." And you have questions.

And then the last one, I think, is we're kind of just accustomed to infectious disease, at least respiratory infections. It's like, "Oh, I have a cold. It's a regular thing," and we're not really fired up by it.

There's a much broader historical reason here. You don't grow up with your sibling dying of infectious disease in lots of parts of the world. There are still some places, tragically, where this is the case, but in lots of places where you could do something like far UVC, people don't have that experience. So we're kind of in that weird area where it doesn't seem like a huge problem to people, even though I think the numbers tell a bit of a different story, and so people aren't fired up about it.

For all of these reasons, it's a complex system. It ends up being a challenge. But it isn't to say there aren't things that you can do. Indeed, what we've been focusing on at Blueprint Biosecurity is really coming up with a plan to tick these barriers off one by one and make progress on them.

Patrick: If I can just highlight one of the things that you said. I think when you compare this against an intervention like penicillin or similar, there's direct value capture. If a doctor or other part of the medical establishment administers penicillin, in that they will get paid directly for the cure.

And there's a sort of one-to-one link on "here is the intervention" and then "you felt better 48 hours later." And even if no one in society had an understanding of Statistics 101 and then somehow we got penicillin anyway, just reasoning in the form of understanding how sympathetic magic works would allow you to understand that those penicillin pills—those are amazing. We should definitely buy more of them.

Jacob: People used to just die of a bacterial infection. You get a cut on your face and that could be the end of it for you. I think it's hard to appreciate just how incredible this is. I probably would've switched my research too at the time, if I was working on UV.

Patrick: I can testify from personal history. I came closer than I would've wanted to death due to a routine bacterial infection in my twenties while I was living in Japan. A coworker found me after I had not called in for work for two days straight, which was unlike me, found me passed out on the floor of my house, took me to the doctor, got antibiotics, and I was right as rain two days later. 

We don't appreciate that we live in the age of miracles. And we also don't appreciate that we're not tapped out on the number of miracles that we can deploy.

Costs and benefits of Far-UVC

So, you mentioned you had some data to share on costs. My ambient impression is, on the one hand, you're going to tell me a number and that number will be interesting. And on the other hand, the societal estimates for amount of work lost due to the flu, etc., plus several hundred thousand deaths a year means there's basically no number that you could possibly quote that would make this irrational.

But it's good to know the number. Am I greatly off base? Is this going to be a hundred thousand dollars a room?

Jacob: No. Right now, if you were to buy a far UVC lamp, depending on how many or which one, it would cost between a hundred to a few thousand dollars.

But a single lamp cost really isn't the question—you want to cash it out in terms of how much it will cost to protect a room. If you take something like a typical classroom, and if you look at the cost of the lamp, a little bit of fixed costs for installation, and amortize the electricity costs, which is really small, over a year, you'll probably get something a little north of a thousand dollars to protect a class per year currently.

So if we have 30 children in that class, we're talking something in the order of $30 per child per year. Now, it's important to note that far UVC is incredibly nascent in terms of a technology. There's probably tens of thousands of lamps in the world. So if you think about learning curves and reductions, and if we really start to scale this, I don't know exactly where the floor would go for this, but it's very easy to believe that it would drop an order of magnitude.

When we talk to some manufacturers and talk to them about their capacity, they give projections for steep cost decreases. So we could go from something that is $30 per student per year now maybe down to dollars per student per year. And as you noted, this is a bargain.

Technologies for cleaner air

When you look at all the types of things that can happen in terms of infectious disease, whether it is the mortality, the sickness, the economic loss of labor, the loss of productivity, the learning loss if you're in a school—even to the point where, not to trivialize it relative to other factors, just the frustration of being sick and having to deal with that and go through your day—I mean, it pales in comparison to the kind of deaths that we saw from COVID-19 or what we see annually from influenza, but all these things add up.

There have been a handful of cost-benefit analyses for far UVC, and they all come out positive, with some of the benefit-to-cost ratios in the tens to hundreds in terms of returns. Now we can talk about whether you capture those benefits or not if you install these fixtures, which is one of the challenges.

But I think the cost of respiratory infectious disease in the US—this is not a pandemic, which we all know can be in the trillions—is on the order of low hundreds of billions a year. So this is costing us money currently already. [Patrick notes: A study which can do a SQL query for actual cash outlays from government and insurance providers came up with about $50 billion in 2016, not even counting indirect costs like e.g. patients or caregivers missing work, decreased output, etc.]

We think this is the case both for far UVC, but even investments in things like ventilation and portable air cleaners have very good return on investment.

The broader impact of respiratory pathogens

Patrick: I'll never say it's a good thing that any part of the pandemic happened, but it was crystallizing for a lot of people in that there was a concrete cause to point to and a concrete enemy and a concrete set of interventions taken against that issue of societal concern.

Whereas in the typical year against the flu and other respiratory diseases, there's nobody—well, there are estimates society-wide of what respiratory diseases cost, and I do believe that low hundreds of billions is probably the right order of magnitude, which is a really stunning stat to throw out in the middle of an interview: "Oh yeah, we spend a couple hundred billion on that every year."

For comparison's sake, when we're thinking of issues that cost tens of billions of dollars a year, that's like all credit card fraud worldwide. And when you get up to the societal issues that cost trillions of dollars, it's like, okay, shooting wars. So we passively tolerate things that are more lethal than shooting wars and about as costly just because they're kind of traditional.

We sort of have this background process running that "Oh, you get sick. That's kind of the human experience. The amount of sickness is fixed."

If I can just bang one drum: the amount of human sickness is not fixed. We've successfully—a much less rich, much less societally advanced subset of humanity—invented the field of public health around tracking down individual sources of infection in London and tracing them back to individual wells. And that intervention bent the curve of not just London, not just the UK, but of all humans living in cities. That's almost 200 years old at this point.

And this relatively straightforward set of things could be part of what bends the curve for us for the next 200 years.

Jacob: It's incredible. We have made really extraordinary progress against infectious disease. I think this really isn't appreciated. It's kind of back to the point that for a lot of people, at least in the developed world, it's kind of out of sight, out of mind. But when you look back at what was killing a lot of people, whole classes of infectious disease have kind of been solved in parts of the world.

Again, the fact that it hasn't been solved everywhere is a tragedy and another discussion. But if you look at waterborne, foodborne, vector-borne, and even bacterial pathogens as a class, these are kind of solved to some extent. It's really these respiratory viruses that, for quirks of history which are quite interesting, we didn't quite get around to addressing, and we remain devastatingly vulnerable to.

But I really like to emphasize to people that it doesn't have to be this way. Indeed, we can do ambitious things. What we did for sanitation was pretty remarkable. Chicago was raised like 10 feet—the entire city—to put the sanitation system in. It's a whole other story about progress, but we used to be excited about this, and I think we still can be, or at least some of us still are. We can do this.

Rediscovering the world for better health

Patrick: It definitely isn't fixed that we're like, "Oh, there's costs associated" or "Oh, there's political deadlocks" or "There's a debate in society and bad feelings about some actions taken by the public health community and credentialed scientists during the pandemic," etc. But that doesn't damn us to never making an improvement again. We can rediscover the will to have nice things.

Historical perspectives on infectious diseases

I think you're right—infectious disease is not quite solved, but if you just went back in a time machine to 120 years ago and said, "Okay, we've got some updates for you. Here's what the mortality curve looks like for five-year-olds."

[Patrick notes: One of the most important graphs in the history of the world. 

]

And against that, "We recently had a pandemic, and you're going to laugh when you hear about some of the things we did during the pandemic because we seem to be less effective than you folks were during the flu pandemic."

But they'd probably be too jaw-dropped from the first revelation to get around to the "Oh, pandemics happen. That's why you have a public health department."

But I think, off the top of my head, that most of the improvement in human lifespans over the course of the last hundred years—there is some amount that you can attribute to the improvements in nutrition broadly, and that's particularly the case with regards to the zero-to-five set.

The role of sanitation and antibiotics

But then an enormous chunk of the rest of the contribution of medicine, at least when you go by that metric, is not improved surgery, it's not even improved end-of-life care, it's just sanitation.

Jacob: It's really incredible. When you look back at the history and read some of this, there are little things that almost escape detail at first, but there are stories about how in some cultures they don't name a child until the first year of life because it's just so likely that they'll be lost before that time. So many tragedies that were part of everyday life. Makes you reflect on what we're putting up with here that we will similarly look back on with disgust potentially.

Patrick: I'll have to look for the citation, but there was a debate in some parts of the scholarly literature on whether people living hundreds of years ago just had a fundamentally different relationship with their children because children were so likely to expire in the first year.

And then other scholars said, "We have diaries. We have written evidence. We have the epitaphs that are written on gravestones about fathers writing that they're shattered forever by the loss of their child." [Patrick notes: See generally.] It wasn't a different class of humanity with a fundamentally different outlook on the value of human life. They were just laboring under the specter of pestilence constantly. And then scientific progress made that more optional than it had historically been.

I earnestly pray and hope that in a couple of decades, our grandchildren will be looking at us and saying, "Wait, really? You had a season named after a disease every year, and you didn't think that was a problem?"

Again, manufacturing technologies have improved quite a bit in the last 70 years. But I think you mentioned there was research done on this sort of stuff in the middle of the 1950s, and it's just historical contingent reasons why we didn't choose to deploy it as widely as the television was going up and to the right.

Miasma theory and airborne transmission

Jacob: It's a fascinating story, and to understand why, you actually have to go back a little further. We're all now familiar with the microbial infectious disease theory where we have microbes that cause disease. But before that, there was this theory called miasma theory, which you've probably heard of.

Remarkably, a lot of the water sanitation that we did was due to the miasma theory of infectious disease. People didn't like how the water smelled because it turns out water was sewage. It was both bad for you and it stinks.

Because of this miasma theory though, when it finally got refuted, when we understood and discovered bacteria and viruses, some of that thought about pathogens and disease going through the air was still kind of tied to that. If someone said, "I think this germ is maybe spread through the air," people would say, "Ah, no, you idiot. That's miasma. We got rid of that and we don't deal with that anymore." There were some really prominent infectious disease researchers and doctors who really pushed this point.

But over time, the weight of the evidence for airborne transmission of disease started to pile up, particularly with tuberculosis. There were some really fascinating studies done by this husband-wife scientist team, William and Mildred Wells, that really just started to definitively show that indeed you can have these pathogens that go through the air. That was right around the 1930s or so.

Getting further into the 1940s, there was actually a bunch of research. If you go back and read newspapers in this time period, it's fascinating. They're experimenting with UV, they're putting vapors into the air, they're trying to suppress dust. You can kind of see how they're getting early ideas about what we'd now use filtration for, but they would kind of oil floors.

Impact of WWII on disease research

But the problem was, as they started to make progress, as I mentioned earlier, this really got accelerated during World War II and a bunch of progress was made, and then a handful of things happened that really made it all fall apart.

One was that, along with World War II, some of that airborne disease research got classified. Some of it went into the state bio weapons programs, and it wasn't something that you talked about a lot more broadly.

More broadly, though, what happened was, as I mentioned before, penicillin was scaled up at the very end of World War II. And we had that effect of, "Well, I could put it in a room and try to study a room and see if I can show if someone comes into that room they didn't get infected there, even though they could get infected somewhere else." Or "They get sick with a bacterial infection, I can give them this magic pill and cure them."

I think compounding on that, we started to get really good with vaccines after that. Smallpox, measles, and polio got vaccines where you basically got them early on in your life and you never had to do anything again, and you never got the disease. Not surprisingly, a lot of people started focusing all of their research into this.

What was really the sanitary revolution continuing on to air, finally, where a lot of the research was done by people called sanitary engineers and really with a focus on the collective in terms of public health, started to transition over to something that was done by medical doctors and was much more pharmaceutical in nature.

When you look back at infectious disease around that time in the 1960s in particular, people kind of thought it was solved. There are some really crazy quotes from then where people say things like, "Don't go into the field. It's not gonna be that interesting going forward."

That was enough of a lapse that most of the people who were working in these areas of UV and airborne transmission and collective protection kind of died out. A lot of the research died with some of these people, like the Wellses. That put us into a real lapse until we started to see things like antimicrobial resistance and extreme drug resistant TB that had a small handful of people kind of awaken on this.

There are people out there like Ed Nardell, who has a direct lineage from working with the Wellses, who have persisted all throughout this and get a ton of credit. But mostly what happened is that it was pretty small up until COVID, and we reawakened the understanding of airborne transmission and have seen a lot more research since then.

Patrick: I think the history of medical research has been just fascinating because we generally have this schema where scientific progress goes up and to the right, and that hasn't always happened. We lost the knowledge that scurvy was caused by vitamin C deficiency, and then that got rediscovered by, I think, one ship's doctor through patient experimentation.

We shouldn't make too much fun of our ancestors for saying, "Oh, you idiots, that couldn't possibly be airborne transmitted," given that we had all the king's horses and all the king's men come to that conclusion in 2020 against mountains of evidence which should have been available to them.

[Patrick notes: To be maximally charitable, the authorities invoked insufficient evidence to come to the conclusion that the coronavirus was airborne transmitted. This was poor epistemics and a poor communications strategy, a toxic combo the public health authorities deployed repeatedly.] 

The broader impact of respiratory pathogens

I think right around the World Wars, for a variety of reasons, there have been sort of agglomeration effects and sorting effects with regards to particular industries, which I think tended to draw a lot of the talent away from fields like public health and redistributed it elsewhere in the economy, including into the more pharmaceutically-minded parts of the healthcare ecosystem.

As a result, up until the pandemic, if you were to ask the people in charge of public health, "What are your top priorities this year and what are you doing about them?" frankly, they did not have what looked like, in retrospect, great answers to that question.  Some webpages are disappearing from the Internet because the contents of them are too embarrassing to look back on. [Patrick notes: Some are not.]

But again, focusing on the positive and focusing on the future, we have options about what we do over the course of the next 20 years. It matters incredibly whether this is broadly deployed in 2030 versus 2032 versus 2035.

I mentioned on Twitter, partly because I had fallen down this rabbit hole, that I think this is going to be broadly deployed in the United States in the same fashion as an indoor smoking ban by 2030. As the number of days until 2030 diminishes, I think I'm a little bit over-optimistic on that one.

But I don't expect to still be waiting for this at retirement. It matters enormously when it happens. That was a drum we banged incessantly at VaccinateCA, where it's not enough that the vaccine has arrived and everyone is going to get it eventually. The exact day on which you administer it to someone matters incredibly.

So you should act with a level of urgency and focus associated with that. Our mantra internally was "Every day matters and every dose matters," and I think that would be instructive with regards to non-pharmaceutical interventions like this and other big society-level projects. It isn't enough to simply get all the research done and all the political economy issues solved and get the grand compromise and a stamp by the 15 regulators and have it happen eventually in our working careers sometime.

It matters on a week-to-week basis. There's an invisible graveyard in the delta between those two timelines. And we should care about that much more than we sort of ambiently decide to care about things.

Jacob: I totally agree. And there's also a clock ticking on pandemics too. A lot of people think they're like once in a century, but when we've looked at rigorous analyses of what the base rate is, it's like 2-3% per year. That's a 30% chance each decade. And that's not accounting for the fact that this might be increasing due to abilities to engineer biology.

Whether it's state bio weapons programs or lab leaks or bioterrorists, these risks are nontrivial. As you know, every day matters with the deaths that we're seeing from infectious disease. But it'll really matter whether on that day zero, when that next pathogen comes that could be a pandemic, whether we have these capabilities. That's one of the things that similarly animates us to try to make progress as quickly as we can.

Patrick: I think the increasing interconnectedness of the world and increasing population of the world are also sort of risk factors for pandemics. One of the mental models I tried to give people during the COVID years was: just like we use compute capacity on various server farms in various places to train AI, viruses get, essentially, compute capacity. And instead of running on an Nvidia chip, it runs inside a person.

The virus, not being sophisticated enough to understand what a border is, is pretty much ambivalent as to who gives it the compute capacity. Given the increasing interconnectedness of people and the great difficulty that, empirically, we would suffer in trying to do something like close the borders in the event of a pandemic, the risk goes up over time, even if one discounts the risk of human-engineered pathogens, which unfortunately is probably getting worse over time as well.

So if we know what we think the broad shape of the glorious future looks like, and we know what the current state is, what are the things that are actually going to happen over the course of the next few weeks and months to drive this forward? How could people be involved?

Future of far-UVC technology

Jacob: The vision that we have, which is our best guess for how we get to the world where far UVC is widespread, looks roughly like this: We're already getting the early adopters who are excited and are installing it. We will continue to climb up that value chain where people who get big benefits from it will be compelled to use it.

This could be things like professional athletes or Olympians where the difference between a respiratory infection or not is very easily the difference between medaling and being remembered by history or not, and kind of seeing that happen.

All the while, there is more data being generated on the safety and efficacy side that makes it more compelling. And as you get more adoption and more of that data, the cost-benefit calculus continues to shift, and we continue to see more adoption. Maybe we start to, at some point, get recommendations that it goes into place.

Our hope is that when you get to the point of having the gold standard for evidence of a cluster randomized control trial, where you can definitively show the value of this, you can start to get to the point where maybe insurance starts to incentivize it, and you get ways that it starts to get more widespread.

When we look at other types of interventions—fire safety is a really interesting example here that we could go into—what you see is that insurance starts to cover it. And then at some point when the evidence starts to mount enough, usually as it spreads wider as these incentives go, you maybe start to get legal liability. People saying, "Well, we have this thing that we know works, why weren't you using it?" And that can compel adoption.

And then ultimately you get to the place where you get to regulations where people say, "Wow, it's pretty irresponsible to not have a fire extinguisher in this place given the evidence that we have." And that ends up pulling along the laggards.

When we look at these other types of public goods, we often see then governments and philanthropies end up having compelling evidence to fund it for those who can't afford it. So we anticipate that far UVC will probably follow a pretty similar kind of trajectory, and that it can be supported throughout this process by having that evidence generated to go along with it.

We like to give people the analogy that you need something similar to FDA approval because if you were to go into—name your big tech company or big organization where you want to have it—the challenge is that all this data, even the compelling data that's out there now, is in scientific papers. It's kind of hard to parse. We've done our part, we've put out a blueprint to make it a bit more accessible. But there's no arbiter for this, like the FDA, that just says "this is okay."

You go into this initially and you think, "Oh, well great, we don't have to deal with the FDA! This is awesome." Then you quickly realize, "Well, when a lot of these big organizations—bureaucracies in some sense—where someone can speak up and say 'We don't want it,' who are you to point to as the source of authority?" We think you have to kind of recreate that in a way that ends up getting you to the same type of situation.

Patrick: I also think that people underappreciate that a large organization is not simply one brain making platonically ideal choices for itself. It has an org chart, and unfortunately this falls in white space in the org chart. There is no director of air quality.

There actually is someone at Google whose job is making sure that the requisite number of fire extinguishers is installed. There are many someones there—relatively low on the org chart—but I express a very high confidence that all Google offices have fire extinguishers. But it's not that person's job to make sure that UVC is installed until it becomes their job due to some combination of "Well, the insurance company put it in our annual assessment" or "It's now a part of city code."

So accelerating the proof points to get to there helps, and then accelerating some people to have some sort of commercial incentive to figure out, "Okay, even if Fortune 500 companies don't have a process that knows that it wants to buy this right now, we'll figure out a sales motion that gets it to them anyway."

Which brings me to something that I've had an idle curiosity about. It seems likely to me that this is going to get adopted either for commercial spaces that have the most expensive people in capitalism in them—the Googles and McKinseys in the world where they pay people six figures a year and don't like them being absent from work—or it is going to get installed in someplace like, without loss of generality, a Chipotle, where they pay their workers much less than six figures, but they have much more rigor around tracking absences and understanding the direct cost of absences.

In the white-collar workforce, people get sick an average amount every year. What are you gonna do about it? That's why you have 60 employees in the division rather than 57.

So what's your best guess: are we going to see it first in Chipotle, or are we going to see it in a Google office?

Jacob: It's a great question. My sense is a little bit of both. But what we're getting at here is that it will go in double based on where there are commercial incentives—where the money can be followed and someone can make a case that this is saving money.

I think when you look back at other types of threats that humanity has had to address, fire is one I'm very fascinated with. Fire was something that was a scourge of humanity for millennia. Every major city has burned down. And it wasn't just like one day we said, "Oh, we're gonna get our act together and solve fire."

What happened was insurance was invented, and insurance companies got very interested in this, and they started to say, "Well, hey, I'm not gonna insure your house if you make your chimney out of wood. I'm going to look into this a bit. I'm going to charge you much more if you're a higher risk for me." And they started to look around and say, "Well, we're paying out a lot of money. Maybe we should have fire departments that can help us with this."

Patrick: That's a forgotten bit of history in a lot of places—that the fire departments were originally insurance schemes where one, you would only get your house treated by the fire department if you were paying into the insurance scheme. But two, the economic incentive for the insurer to ensure that your house wasn't going to burn down: "Oh shoot, we have to replace the house if it burns down. Better put out the fire if it happens."

Jacob: Yeah. Markets and everything. And so I think what will happen with far UVC will be that where we have these clear financial incentives—and this is where I think some of the data generation needs to go as well. When you're looking at what the types of benefits are, you need to ensure that you're tying back these benefits to what people care about.

For schools, it needs to be tied back to what people care about. Lots of schools get paid by the state based on meeting a number of children there that day. If they fall below it, they don't get paid. So they care about absences. You need to track absences as the endpoint of your study to see that. So wherever we can tie these things together is where I think we'll get these incentives.

And I think that will be an important component of expanding that market early on until we kind of get to these levels where we start seeing insurance or reliability or regulations. But crucially, the thing that we're focused on is that given how diverse these could be and the different number of stakeholders, our sense is that given that there's no FDA here, the level of evidence will have to be quite high.

What Google or Chipotle will want might be different. So we want to make sure that all of them kind of feel assured because, as we've already discussed, the cost-benefit analysis for a perfectly rational actor, which we know isn't the case, already looks pretty good. But that won't actually be who we'll be dealing with in terms of these decision makers.

We outline in our plan for far UVC a pretty rigorous safety and efficacy agenda, but it's really with this in mind of generating all that data so that this calculus can pan out and make progress. And of course we hope as well, when it comes to thinking about things like pandemics, there's potentially philanthropic or government interest in protecting, whether it's critical infrastructure or certain spaces, well in advance so that we can accrue those benefits.

We talked about the annual cost of respiratory infections, but the bigger one is that if you take that 2-3% per year risk, which is the risk around a pandemic at the scale of COVID, and you amortize the cost of COVID over this—trillions of dollars—it actually far dominates everything else.

So making this more and more compelling and getting through some of these blockers around concerns are what we think will be important to get this to where it needs to be.

Patrick: It wasn't obvious to me that the cost of infectious disease would be dominated by amortized cost of pandemics versus just the slow ongoing tragedy that is the flu.

I will drop a link in the show notes to your blueprint for getting UVC deployed more widely for people who want to read the case for that. I do think it's a matter of providing overwhelming evidence and then glide paths to getting it into places.

It seems like it's a layup for congregate care facilities, which as we learned during the COVID pandemic were some of the places where the most just bordering on unspeakable human tragedies happened. But are hospitals also not pretty high up on the list here for places that should be installing it? Just speaking of where disease is likely to be present at a higher base rate and where vulnerable people are.

Jacob: Absolutely. Hospital and clinical settings are some of the markets that are really being targeted by a lot of the companies that are developing this technology because it goes back to being able to close the loop. They're treating patients; if they're sick, it can be your problem.

Sometimes there's a little bit of misaligned incentives with whether the insurance pays or whether you pay. But oftentimes these are closed enough, and this ends up being a place where you can do this. Again, making those markets is what's so important.

When you go back to the big companies, another important question will probably be in terms of whether they use it: Do they self-insure? Are they the ones kind of incurring some of the cost? Can it be closed on the books and recaptured in a way where people really see this as the appropriate return on investment?

Patrick: I think given that while we don't have a single-payer system in the United States, a lot of the total cost of medical care is under Medicare/Medicaid. And there are things like payouts conditional on your clinical endpoints.

If there is not a subtle signal, but a very strong signal in the data that the number of infections incident to surgery goes down remarkably after having this installed, then we might be able to influence a large portion of the medical establishment to install it just because the hospital administrator will say, "Well, that's $40 extra per patient starting N weeks after installation, this is a no-brainer for us."

Jacob: And if you go back to the 1930s with the conventional UV that I mentioned, hospitals were some of the first places they applied this. Some of these studies even back then showed benefits in terms of surgical site infections.

It was actually a motivation for the scientist David Brenner to reduce surgical site infections where he really came across and got much more interested in this far UVC as well. So it's a recurring theme, even where history has repeated itself. But it goes back to exactly the point that you're making—this is where these markets can be made and closed in pretty tight concert. And we expect this to be key to the widespread deployment of far UVC.

Patrick: We mentioned earlier there is no person that is doing the intervention on any given Tuesday while this intervention is operating, which unfortunately makes it a little more invisible—a little less connection between the intervention and the improvement in outcomes. But the flip side of that is that there's no person who has to be convinced to do it on any given Tuesday after installation. Light is working above you.

One of the most frustrating bits of infectious control research was that getting surgeons to wash their hands is an enormously effective technology at reducing infection and death by infectious diseases. The problem is that you have to convince surgeons to wash their hands.

And there was a multi-decade battle over that despite the germ theory of disease not being very controversial by that point. But just convincing high-status people: "No, really, we understand you washed already, but you have to go back to the sink." There's a great phrase in Japanese, "shitsukoku" (しつこく)—you have to really go in there and you're going to do it past the point where it feels like it's giving any benefit because, here's the curve.

So, we won't, knock on wood, have to have a debate with every surgeon for decades until they retire to say, "Now remember to turn the UV light on," because the UV light is always going to be on when the lights in the room are on.

Jacob: It's a great point about the surgeons. And one of the things that actually makes some of these interventions like far UVC a bit harder to study is that, as you noted, it's really hard even when you have the study where you say, "Wash your hands or don't, and then have the surgery and you see the outcome there." There's a very direct correlation there and ultimately causation that you can study.

But the difference when you're studying something in the built environment is that instead of studying people, you are studying a room. This is something that my colleague Richard Williamson, who a lot of the blueprint is actually credited to, really likes to point out. This is quite different when you look at pharmaceuticals versus some of these other interventions where you're treating a person versus a place.

This can make it much harder. I think this is one of the reasons we've been slower to adopt some of these interventions. Again, the easiest one is: give someone who's sick with a bacterial infection an antibiotic, they get better. We don't need a large n to kind of understand that works.

Vaccines are kind of the next level where you're actually trying to prevent an infection now. So this has gotten a level harder, but we can track you—we can give you the vaccine and then look and kind of say, "Did you get infected or not?"

And this is where it gets even harder for things like masks and respirators and these other ones, where it starts to get a bit harder. We say, "Well, did you get the infection while you were wearing it, or did you get an infection at some other point in time when you weren't wearing it? Or did you prevent the infection while you were in the classroom and then ended up going out somewhere into town later and get infected there?"

This has been something that's haunted this type of disease control for a long time. Some of the history with far UVC, or at least for upper room UV, is that there were a handful of really good studies that showed this level of reduction in disease burden. And as they started to replicate it, they took varying degrees of caution and care in the studies that they designed.

Sometimes they did not control sufficiently for outside infection. So they study it in one school where kids are actually pretty well off: they go to school, that's their main mixing, they don't get infected there, they go home to their nice houses, and we can see a reduction in disease.

But when they took it to other places where kids were less well off and crammed on buses and went to crowded housing, what they found was that you could probably see that they were not getting infected at school, but they were actually still getting infected.

So this is a really interesting point that makes some of this a little bit harder to study, but still possible nonetheless. And it's one of the reasons we're really gearing up for the kind of sufficiently powered randomized control trial that you would need to be able to demonstrate this definitively.

Patrick: And there are also some felicities here in that if the person has an ongoing relationship with the room, then you can have more visibility into whether they are affected by things that happen in the room or not.

But when you mentioned places that are high on our list to deploy—places where people are transient, such as shopping malls, houses of worship for many people, transportation systems, and similar—these do not make it easy for you to retrospectively say, "Okay, can we get a list of everyone who went through subway station in the three-week period here where we turned off the UVC?"

So I expect that what we'll have to do is get just overwhelming evidence that it works when we know people's names who are in the room, and then just assume the UVC is not operating on a mechanism where it does a quick question and answer with each virus and says, "Is this person here a lot?"

So there is always going to be a gap in the science and a gap in the literature. And at some point we just have to decide we're going to leap the gap and make reasonable decisions based on the evidence available to us. And I hope that happens sooner rather than later.

Jacob: Me as well.

Patrick: I feel like we could discuss this for hours upon hours, but unfortunately all good things must come to an end. Jacob, where can people read more about UVC if they want to get UVC deployed?

Jacob: Our website Blueprint Biosecurity has an almost 300-page document that covers everything that you could imagine and want to know about far UVC [PDF link]. So I would encourage people to read that.

If you're interested in that and you're excited to contribute, don't hesitate to reach out. We can be contacted at info@blueprintbiosecurity.org, and I'm always happy to chat with people as well. I'm on Twitter at @JacobSwett

Patrick: Thanks very much for taking the time today. I learned things, and I hope everyone else did too. And for the rest of you, thanks very much for your time and attention. As always, we'll see you next week on Complex Systems.