What Have Epidemiologists Learned About the Coronavirus?

By Isaac Chotiner

Source Link

Justin Lessler, a professor of epidemiology at the Johns Hopkins Bloomberg School of Public Health, models disease transmission, and has been studying the novel coronavirus. A month ago, as the first confirmed covid-19 deaths were occuring in the United States, I spoke with Lessler about some of the early findings about the disease. On Tuesday, I called Lessler again, to ask him how our understanding of covid-19 has evolved in the past month, and how epidemiologists have changed their views of the pandemic’s likely effects. In our latest conversation, which has been edited for length and clarity, we discussed what we know about whether people who have been infected are now immune, the hope that warm summer weather will halt the disease’s spread, and why testing remains the only way to prevent further rounds of mass quarantines.

What sticks out to you the most, in terms of all the things we have learned in the past month?

One is something we have learned, which is that it appears that the social-distancing measures that we have been taking over the current period seem to be working in a lot of places. It’s a little early to say for sure, but hopefully the initial signals that maybe they’re working are an indication that they really are. In terms of what we haven’t learned, I think the amount of uncertainty that still remains on the true underlying burden is a bit disappointing. I had been hoping that we would have more of a sense of exactly how many infections there have been out there. I think that still remains the biggest unknown of the whole thing.

So, by “underlying burden,” you mean the number of infections?

Yeah. We know we’re seeing a particular number of deaths, and that’s a number we can be reasonably confident about. The number of people hospitalized is also something that we can be a bit confident about. But the number of confirmed cases—it’s unclear how much that’s telling us, because testing has been ramping up a lot over the past month or so, and a lot of those increases that we’ve seen in confirmed cases are just due to more testing. So the epidemic is progressing, and we are also testing more. Those are all the direct measures. But all of those measures come out of a pool of infected people that we don’t see, and we don’t know how big that pool is.

Is the reason that we don’t know that we haven’t ramped up testing enough, or is there something about the virus that has made it more difficult to figure that out?

It’s nothing specific to the virus, and it’s a little bit more than just testing. There are two ways you can get at that unseen pool. One is if you know how deadly the virus is, or how many people become clinical cases that you can detect, and we don’t really know that yet. The uncertainty in the mortality rate per infection I still think stands probably an order of magnitude between being as high as around one in a hundred, and as low as around one in a thousand. The other way you can get at it is if you’re able to go out and do not the type of testing most people are thinking of, which is virological testing to see if you’re infected right now, but serological testing, which tests for antibodies to indicate if you’ve ever been infected. Those studies are ongoing, but we haven’t started seeing many of those results yet. So we don’t really know how many people have been infected. But it’s the serological studies that are likely to answer the question more definitively, and sooner rather than later.

There seems to be an assumption that, once someone has the disease and recovers, they are likely to have some sort of immunity in the future. But at this point, how much do we really know about it?

We don’t know much in terms of the specifics of how the virus progresses. I mean, there has to be some immunity.

Why?

Because you cleared the virus somehow. You got the virus out of your system, and your immune system did that. How your immune system did that was by having an immunological response, and that was somewhat specific to the virus. So, for that reason, you have to have at least some immunity and protection from the virus. How much immunity and protection it is remains an open question. It could be what we call sterilizing or complete immunity, where you can’t even get infected with the virus, at one extreme. At the other extreme, it could just be very weak immunity that wanes quickly and maybe dampens symptoms a bit but doesn’t stop infection. So it’s clear that there has to be some protection, but the amount of that protection, how long it lasts, how important it is for interrupting transmissions—those are open questions.

If there is some immunity, would it exist only if you had shown symptoms, or would you have immunity if you had the virus and were asymptomatic?

You’d probably have some immunity either way. I think in most cases it’s likely that your immunity would be a little bit stronger if you had symptoms, because a lot of the symptoms you get for diseases like this are driven not by the virus itself but by your immune response. For instance, fever is something that’s driven by your body fighting the virus, not by direct action of the virus. And so it’s likely that people with symptoms may have a bit stronger immunity than people without, but the level of variation is something we don’t quite know. And there’s a lot of potential complexity there. Immunological interactions are complex things. In some cases it can even make second infections worse. We think that’s quite unlikely in this particular disease.

But the point is that there’s a lot of complexity in the way that immunological interactions work. Where I think we can be confident enough is that, over time, even if it’s not strong on the first infection, people will accumulate immunities such that they’ll be protected at least from severe outcomes from this virus on subsequent infections.

So if we were hoping for something, it would be that the virus spreads faster than we think, is less deadly than we think, and, once you’ve had it in any form, you have great immunity?

That would be the best scenario right now, that essentially there are many more people out there who have been infected now than we think, so that we’ve accumulated these high levels of community immunity, or herd immunity, that will protect us from subsequent waves.

Have we learned anything in the past month that makes that scenario more or less likely than when this thing first hit?

I think there are a few pieces of evidence that might point either way. I think it seems unlikely that it would be really, really widespread. I think there’s a moderate version of that scenario that’s still possible. But just take the number of deaths in a place like New York. The population is over eight million people, and just over eight thousand people have died. [New York City recently revised its death count to more than ten thousand people.] If the death rate is one out of a hundred, that implies that eight hundred thousand people have been infected, which is reasonable. But if the mortality rate is one out of a thousand, that assumes that eight million people have been infected in New York City, which is everyone. So if you scale those up by too big of a factor, you quickly get to a point where it implies that an unrealistic percentage of the city has been infected. So that sort of puts an upper limit on that scenario.

But some people think it spreads a little bit faster than we thought—those people have estimated reproductive numbers that were higher than we had been assuming from earlier estimates. So those maybe indicate that there could be more infections out there. But that’s all to say that there’s evidence, and it’s not clearly supporting either a lot of asymptomatic cases and a very low infection-fatality ratio, or not a lot of asymptomatic cases out there and a relatively high infection-fatality ratio. You could probably go out there and find somebody who would argue vehemently for either side. But I think it’s still one of the primary sources of uncertainty in the epidemic.

We’ve been aware of age discrepancies in terms of who suffers serious bouts of illness. But a month ago, there were still a number of unanswered questions about exactly how the virus was working with kids and how it functioned with people of different ages. What have we learned in that time?

I think it still looks like kids get infected. They don’t get clinical illnesses at the same rates as adults, but, as more and more people are getting infected, it’s becoming clear that they’re not universally protected. We are starting to see some severe cases in young children. We still don’t quite know how well they transmit compared to adults, or how well asymptomatic people in general transmit. But I think it’s becoming increasingly clear that asymptomatic people are playing a significant role in overall transmission.

What makes us think that?

There’s several strains of evidence pointing in that direction—a few case studies here and there and some more insight into viral shedding, and some analysis of the over-all dynamics of infection. “Asymptomatic” is a squirrely concept, and what we really mean here is “not detected and likely not detectable.” If you look at the shape of the epidemic curves, it is hard to explain them if children are not contributing to transmission at all, and if asymptomatic younger people are not contributing to transmission at all.

I still don’t think it’s clear that we could say, “O.K., you’re symptomatic, and you’re not symptomatic, and you, the symptomatic person, are, say, three times more transmissible than the asymptomatic person.” I don’t think we’re there yet. But I think we do understand that asymptomatic people, which includes people with mild symptoms who we just wouldn’t recognize as being infected, or super sick, can transmit and are likely helping to drive the epidemic.

There is hope that warmer weather in the summer will contribute to keeping cases down, which is in turn based on what may be a correlation between warm places and lower incidents of the disease. Do we have any real evidence for this beyond the correlation?

Not a huge amount of evidence. There’s some indication that there may be a seasonality in other coronaviruses, so that would support that idea. The seasonality may lower the amount of infections in the summer, if it does exist, and buy us some extra time, and amplify the effects of the shutdowns and social-distancing measures that we’re taking right now to perhaps push the second wave of the epidemic further into the fall. But I think the idea that it’s somehow going to save us and eliminate the disease is a pipe dream.

What makes it a pipe dream?

One is that we know that disease can be transmitted in warmer places, even though transmissibility may have been less in those places. There has been significant transmission in places like Singapore and Thailand, where warmer and more humid conditions exist year-round. So that’s the big reason. And the second is, from other diseases, I don’t think we really have an example of a disease where a combination of social distancing and getting lucky in the seasonality has eradicated it. There are really only two diseases we’ve eradicated—smallpox and rinderpest, and rinderpest isn’t even a human disease. We’re close on polio. But in all of those cases, we had a vaccine to work with. So I don’t think we should have any thought that this will just go away.

When we talked a month ago, the hope was that we’d put social-distancing measures in place, we’d flatten the curve, and then, hopefully, things would be more manageable. It seems like if things continue, we may get to that place. But it also seems like opening up the economy or ending social-distancing measures is going to bring the disease back. Do you have any kind of optimism that we’ve learned something in the past month that will help prevent that fate?

I think a month ago we were still thinking that maybe we’d gotten on top of this early enough that more moderate social-distancing measures could really flatten out the curve. I think what became clear around the time we talked is that things had gotten a little bit out of control and it was necessary to have more extreme social-distancing measures and not so much flatten the curve as slam it down. And so that’s where we started getting all of these stay-at-home orders and things like that. But the impact of those orders is that you short-circuit the transmission of the disease before there’s a chance for significant immunity to grow in the population. So, like you said, if people start going back to business as usual, the population is still at least partially at risk for infection, and we could see a big second wave.

So I think the big focus now needs to be that there has to be a plan for the post-stay-at-home-order era, where we find something that allows the economy to open back up, allows people to by and large go about their daily lives, but still prevents widespread circulation of this virus and the overwhelming of our health-care systems. And right now I think the only reasonable strategy that we have, with the tools we have now, is a “test, trace, and isolate” type of strategy, where we do extensive testing for the virus, and, when we find positive people, we do contact-tracing and find their potentially infected contacts. And then we only have quarantine and isolation among those people.

Right now we’re essentially doing quarantine isolation in a very unfocussed manner across the entire population, sort of a blunt hammer approach. And if we could get that to be targeted exactly at those people who are potentially infected and could achieve similar levels of reductions in viral transmission by doing that, I think that would be ultimately the key to being able to go back to something that looks more like business as usual. They seem to have achieved something like this in South Korea, and hopefully we can do something like that here.

So it comes back to testing, which is what it was a month ago.

Right. But we’ve made a lot of strides in testing since then. The challenges still exist, but we’ve ramped up testing in the U.S. to a huge extent, and I think that, provided that certain supplies like reagents for running the tests are available, or we can come up with alternatives, it should be possible to have one of these extensive testing programs. You would have very easy, voluntary testing for people who had even the smallest inkling of having symptoms, and perhaps even mandatory testing for people who had certain critical jobs. And then whenever you found somebody who tested positive, you would aggressively try to find their close contacts and ask those people and their close contacts to isolate. Our health system is state-specific, so it would certainly be at the state or sub-state level.

Would it be the serological testing you talked about earlier, where we’re testing people for whether they may have had the disease in their system in the past? Because then the people with immunity could take on certain tasks in society, and would be at less risk.

Yeah, I think that could play a really important role as well. I think the challenge there is that we need to not just know that somebody has been infected but know what we call the “correlates of protection” are. What does the immune response, measured from blood, from serology, look like for someone who’s actually protected from infection, compared to someone who maybe had a more mild immune response and isn’t protected? I think the serological tests play a massive role in getting back to business as usual and understanding the virus, simply because they allow us to know how many people are infected and potentially immune and get a better sense of what happened. But if we’re going to start saying things like, “O.K, you’re immune, so we would prioritize you for working on covid-19 wards,” we need to have a really good sense of what actual correlates to protection are.

Have you seen any signs that the countries that tried any variety of the herd-immunity strategy have seen any positive results?

Not that I’ve seen. I think the question is, how much tolerance do you have for large waves of infections hitting your health-care system? As far as I’ve seen, there haven’t really been any cases where a large epidemic has not been overwhelming, where people of a country with a large epidemic have managed to not have their health-care system be overwhelmed.

What have we learned from other countries in the past month?

I think three things that are critical, and they’re broad things. One is that we’ve learned that, if you don’t take any efforts to combat this disease, you will get overwhelmed. Those are the lessons from places like Italy and Spain and the early days of Wuhan. I think the second is that social distancing, at the level of stay-at-home orders and lockdowns, does work. Certainly that has been the lesson from China. Though they may have achieved levels of lockdown and social distancing that we’re not willing to implement here in the United States, it does look like it’s having an impact in other places as well. And then I think the third thing is that there may be a path focussed around testing. And that is what I think the lesson is from South Korea. We’re also learning bits and bobs here and there about how deadly the disease is and things like that, and that information is critical. But those are the three bigger overarching themes.