Tuesday, March 31, 2020

COVID: Ninety Five

A thoroughly scary, sobering piece in the Metro Section of the New York Times has the Billy Rubin blog staff wondering what will await us in just about fourteen days. If the article can be condensed into one word, that word would be: apocalyptic. Nurses and (to a more limited extent) doctors are being sent into battle without proper gear, and anyone who reads this piece can only conclude that, as the tide crests, the lack of protection will consign not only some health care workers, but also some patients, to a very unpleasant and undeserved future.

Yet it is the opening picture that's the subject here: a nurse holding a placard saying "Will Work For a New N95!" It's a clever protest--clever enough to earn its spot in the Paper of Record--and one that expresses enormous anxiety among the doctors and nurses who have moved forth to do battle. Without N95 masks, the reasoning goes, health care workers are as lambs to the COVID slaughter.

Except for one thing: that assumption is not, largely speaking, true.

What is an "N95"? The term derives from the National Institute for Occupational Safety and Health's categorization of face mask protectiveness. The "N" denotes that it isn't resistant to oils, but the 95 is the important bit, as it indicates that it filters out 95 percent of airborne particles. (There are 99 and 100-grade N masks, as well, in addition to "R" and "P" masks based on oil resistance, each with their own 95/99/100 gradations as well, for a total of nine different types of particulate protective masks. But the only one relevant for this discussion is the N95.)

The N95 mask is used in medicine for what is known as airborne precautions: when one is in the presence of a patient who may pass along an infection that can occur simply by breathing in local air. Two of the most well-known airborne pathogens are tuberculosis and measles. That is, someone with measles doesn't have to cough to infect another person, all they have to do is breathe and have someone in their vicinity. Once they breathe, the virus can remain in the air for hours. (Coughing, though, really throws infectious particles into the air.) The same is true for TB, although TB is not anywhere near as efficient at spreading. (Dr. Rubin has worked for years in TB-endemic places, with TB-positive patients, without the benefit of an N95 mask, and has yet to become infected with TB.)

N95s aren't cheap, and are generally only used in narrow circumstances like TB or measles. Most respiratory viruses do not spread by the airborne route, but instead by something infection control people call droplets. Droplet spread means that the virus or bacteria need to hitch a ride to something--usually phlegm, or the aerosols generated in a sneeze. Those loogies are governed by the laws of gravity, however, and so droplet spread means that, as long as you're not directly in the line of those aerosols or droplets (or, alternately, you're not touching the surfaces where such droplets dropped), they will quickly fall to whatever surface is closest, and as long as you're wearing a mask of some kind (like a surgical mask), you stand almost no chance of getting infected--and we'll get to the "almost" in a second.

The question then becomes: is COVID an airborne, or a droplet/contact-spread, virus?

Getting at this answer is less simple than it may seem, because there's different ways of answering the question. One way is to look at actual people who get infected--say, health care workers--and review whether they were wearing N95s or just surgical masks (which protect adequately against droplet-spread viruses and bacteria), and seeing whether there is a difference in workplace-acquired infections. But this isn't a perfect way of measuring the effect, since people can't be controlled the same way mice can be controlled in a careful experiment. Humans have a way of behaving in ways that confound the results. In the first 55,000 cases of COVID in Wuhan, the Chinese CDC in conjunction with the WHO found that many health care workers--some wearing N95s, others wearing surgical masks--became infected, which would suggest that this virus is mightily transmissible even with the most adequate protection. But then the epidemiologists in China looked at what was happening in the homes of the health care workers, and in a substantial number, they found that workers didn't get infected at work where they adhered to proper precautions, but they got the virus at home, when they reverted to less vigilant practices. (See page 11 of the report.) Thus, it wasn't at all clear whether N95 provided adequate protection or not.

Based on previous experience and study with coronaviruses, the World Health Organization felt that airborne precautions (that is, the use of N95 masks) were not necessary in caring for COVID patients unless they were generating aerosols. Thus, unless someone was doing something to a COVID patient that generated these aerosols, like intubating them, doing chest compressions, swabbing their throats, then a surgical mask would suffice. Which would mean that there should in theory be plenty of N95 masks in the world stockpile available for patient care. (Ventilators, unfortunately, are a different story.)

But a brief letter to the New England Journal of Medicine gave everyone pause when it noted that "aerosol transmission of [COVID]...is plausible, since the virus can remain viable and infectious in aerosols for hours and on surfaces up to days." Here's the pic that shows the scary data:

It's the middle, leftmost, red streak that suggests the ghastly conclusion: the decay of the virus when aerosolized--which is to say, the amount of virus in the air that can be picked up--was slow indeed, lasting for hours. You can see that its decline is much slower than seen in the other panels, which indicates the viability of the virus on various surfaces. The study immediately generated headlines suggesting that WHO had gotten it wrong (and, sort of by extension, the CDC, whose guidelines are subtly different but basically concur with WHO on the notion that there's no strong evidence for airborne spread).

But here's the relevant description from the article of how they designed the study. The authors state that "aerosols...were generated with the use of a three-jet Collison nebulizer and fed into a Goldberg drum to create an aerosolized environment" [my emphasis]. Moreover, the study didn't attempt to recover unambiguously intact coronavirus, but instead performed something called "PCR," which looks for the nucleic acid found inside the virus. It's an indirect, and strong, piece of evidence, but it's not, as it were, airtight, as broken up bits of virus can litter surfaces for hours and days long after intact virus can be recovered--thus getting a positive PCR test but being of little biological relevance.

Which is to say, this doesn't look anything like real life. It aerosolizes the virus, sure, but quite possibly in a way that no human larynx ever could, unless it was the Cough Heard Round The World. Thus, it's an interesting and intriguing piece of experimental science, but it's simply one piece of evidence in a quilt of information about how the virus is spread--and each individual piece of information does not have the ability to stand entirely on its own.

The experimental data adds a perspective, but the epidemiology gives real-life evidence of what happened, although that data too needs to be understood for its limitations. Even the lead author cautioned against over-interpretation of her work, noting in her Twitter feed the following: “Important: we experimentally generated [COVID-19] aerosols and kept them afloat in a drum. This is not evidence of aerosol transmission.”

But this nuance--this will come as a shock--has gotten completely ignored in the fevered rush to judgement and general freakout. And it's, you know, a kinda important nuance. Today's news brings another article even more extreme in its assertions--that an "MIT researcher" (they're smart!) is quoted in the New York Post as saying that COVID "can travel 27 feet and linger for hours." The "research" done to prove this? Turns out it's little more than a commentary, and describes no genuine new study designed to truly evaluate what happens when a COVID-infected patient is in a room with someone who is wearing a surgical mask but not an N95. In fact, it doesn't even deign to consider the strong epidemiologic data suggesting that N95s are unnecessary (with exceptions noted above). What it does show is...a video of someone sneezing. And what is its biggest citation? The very NEJM study described above. Which is to say, there's no there, there.

At the Billy Rubin Blog, we don't consider that to be the kind of science to merit headlines. We like articles like this, instead, that provide a whole lotta context about the question of airborne versus droplet spread. Good reading! And a lot less anxiety-provoking than the New York Post story.

Indeed, in this environment, when fingers are being angrily pointed, this seems to just pour kerosene on the flames. There are real outrages that have occurred as part of this epidemic, and for which many people are going to needlessly die. (See, for instance: Trump, Donald.) But based on what is currently known, there is not yet good reason to believe that there's anything shortsighted about public health officials suggesting that a surgical mask is, for the most part, safe for most situations involving COVID patient care.

The outbreak is moving fast, and that could change, but as of yet, the science isn't even close to being definitive that the N95 is required in all COVID situations.


Monday, March 30, 2020

COVID: As Advertised

Scattered thoughts:

I. I am not truly in the trenches, but I'm close enough to the trenches, and I do a number of reconnaissance missions to the trenches each day, such that I can get a good enough view of this beast. Everything I have read on the professional discussion boards, from Washington State to Texas to Louisiana, and above all, to New York, all check out. No disease is pleasant to behold, but some are worse than others, and this one, for those who are most affected, is bad. In the words of a colleague, "there's nothing subtle about this virus." Of course, he was referring to those who are in our intensive care units--it is precisely its subtlety among most infected people that makes it so difficult to contain.

But his point with respect to the sickest is spot on. The Cat scans do not require years of radiology training to elicit a gasp; they don't even require attendance at medical school, as this week's New England Journal images so amply demonstrate:

I have now seen three or four such CTs on patients I'm following; I expect to see more in the days to come.

II. The notion that this was some piddling, trivial virus should have all but evaporated once Italy and Iran exploded at the end of February, although even as recently as two weeks ago a majority of Republicans still found it untroubling. Wuhan was a warning, and Italy and Iran should have led any sane individual who was paying attention to realize that Wuhan wasn't a fluke--that the virus was moving fast, and that with a one percent lethal virus, one percent of a lot of people equals a lot of people. The math behind this was evident two months ago. Quibbling over whether the mortality rate was one percent or two percent or even 0.5 percent missed the essential truth, which is that the absolute number is what's staggering in a virus that can infect with such ease.

There is a virus well known to humans that should have taught us this lesson. Its mortality rate is an order of magnitude less than COVID, but its threat can be found in its extremely high transmissibility, for it is probably the most transmissible virus known to humankind. It is measles, and prior to the measles vaccine, it leveled children: only one in a thousand, which seems laughable, except that a child with measles can infect between 12 and 18 unvaccinated people. Millions of kids died every year, and even now, after a massive campaign that has reduced the mortality rate of measles by almost 75 percent, there are still nearly 150,000 children who die from the disease each year.

COVID's not that good at transmission--an infected person probably infects somewhere between two and three people--but it's good enough in an age of transcontinental travel to light up the globe in only two months.

III. There are at present three countries left in the world that do not have documented cases of COVID, and last week I was on one of the last planes out from one of them. Sometimes I do question my own sanity.

IV. I wonder if we had been lulled into a kind of virus stupor with COVID because we had been so acclimated to the fearsome case fatality rate of Ebola. Get Ebola, and until recently, it was basically flip a coin as to whether you lived or died. Actually it was flipping a trick coin, since the mortality rate was a bit more than 50 percent until finally we established that two specific treatments definitively worked. A one percent case fatality sounds so pedestrian compared to Ebola's case fatality that I wonder if it even caught the epidemiologists by surprise. At any rate, some historian in the years to come may want to juxtapose the initial response to COVID in terms of the deadening of a stimulus-response brought on by Ebola preoccupation.

V. Irony: I worked with Ebola patients during the West African Outbreak five years ago. I was millimeters from that virus. I examined patients while they sweat on my PPE, held them while they vomited and while they cried. I gave one of them back rubs regularly. I held their hands and walked them from the suspect ward to the confirmed ward. On occasion I stuck needles into their arms in order to draw blood tests. And during my time working in the unit, I slept like a baby. I was never scared of what I was doing. It wasn't that I thought I was going to be okay--far from it--it's just that I was able to approach my work with a calm that never abandoned me.

This thing? It's not so much that I'm scared, but I am definitely tense, and in a way that I never was in the Ebola Treatment Unit. The whole damn hospital is Hot--indeed, the entire state is Hot. And there is something grim and menacing about that.

VI. More on books to read another day, but if anyone wants to step back from the daily COVID cacophony yet still learn something about this moment in history in which we find ourselves, there may not be a better book than John Barry's The Great Influenza. It is impossible to read this book now without feeling a sense of deja vu. If you really do have the time on your hands, this is a good place to start--but more on some other pandemic/plague books worth reading another time.

More soon.


Wednesday, March 4, 2020

A COVID-19 Analogy: Snowstorm

Here at the Billy Rubin blog we're fond of medical analogies to blizzards, and as I see article after article breathlessly talking about coronavirus-this and coronavirus-that, and what still seems to be mass panic across the world, it may be helpful to think about the coming COVID-19 epidemic in terms of how those of us who live in snowy climates think of bad snowstorms. Which is to say, they can be bad and disruptive--and even deadly--but people have learned to cope with them. And with the proper approach, they can be weathered, but that doesn't mean they don't put a lot of strain on the system.

A massive snowstorm makes travel difficult at best; that drains the workforce of workers who make or do things. (Yes, there's telecommuting, but no, that won't work for, say, plumbers and electricians, who are, you know, important to making things run.) Schools shut down so even people who might be able to make it in are otherwise occupied. Overall, the economic impacts are real, but temporary, and mostly remain confined to cold-weather areas and don't ripple all the way through the economy.

People die in snowstorms, generally by power and heat loss, or through heart attacks of people with heart disease who head out to shovel snow (I saw this during my residency more than once). Do a lot of people die? No. But does a bad snowstorm strain the system if a bunch of frostbitten or collapsing snow shovelers require emergency medical care? Yep.

That's this coronavirus in a nutshell, though with some important differences. The first involves economic impact: the epidemic isn't just limited to a small part of one country, but will shut down the workforce everywhere. Supply chains of products are international, so the effects of China's aggressive containment policies will very likely have a negative impact on the global economy for months. And a snowstorm lasts days; this is already two months old, and it's hard to guess when the workforce brownouts will stop--though I don't think six months is an unreasonable estimate.

And people will die, just as they do in bad blizzards. Any given individual who gets infected is unlikely to die, just like your chance of dying in a snowstorm is fairly low. But that doesn't mean you can't attribute a number of deaths directly to the storm. And again, the storm will hit the entire country (indeed, nearly every country). In aggregate, it will add up to a lot of people.

Just as with snowstorms, there are protective measures that reduce risk, and the most important of these is handwashing.as well as "social distancing"--keeping yourself at arm's length (though the CDC says 6 feet) from people, particularly those with symptoms. It won't reduce the risk to zero, just as no intervention can prevent the hazards associated with a blizzard. But it isn't hopeless, and shouldn't cause mass panic.

Analogies are tricky, but hopefully this one helps give a little context to the numbers.