I am seeing some good news for South Africa in this week's data, as South Africa has somewhat flattened the curve of new cases over the past 3 weeks. Last week I was skeptical that this slowing of new cases was real, and thought it might be a result of testing not keeping up with actual infections. Test positivity in South Africa is high (as of last week's data, 26%) and nowhere close to the 5% recommended by World Health Organization as a measure that a country is testing enough... but test positivity has been fairly stable in South Africa, so I don't see evidence that South Africa is falling further behind in testing. (You can follow @rid1tweets on Twitter to see data on test positivity rates.) We haven't yet seen deaths per day fall in South Africa, but I am hoping to see deaths slow in the next week or two.
The U.S. curve may be flattening a little, but of course this varies so much by state. Here's the curve for Illinois (showing new cases per day), where I currently live and where obviously we are not on a good trajectory. (Visualization from Johns Hopkins Coronavirus Resource Center.)
One topic that's been on my mind this week (and that I haven't mentioned often enough in these posts): racial disparities and the horrific toll this disease is taking on people of color. I heard this week of a person of color on staff at my institution (Wheaton College) who has lost THIRTEEN family members to covid-19. Nationally, Black Americans have 2.5 times the risk of dying from covid-19 as white Americans, and other minorities are also at much greater risk. There's lots more data at COVID Racial Data Tracker and APM Research Lab's The Color of Coronavirus project (including data by state). This visualization is from the Data Tracker site.
I don't have data by race for South Africa but am guessing the situation is much the same as the issues are much the same. Compared to whites, other racial groups are much more likely to be working in occupations that expose them to risk and to live in conditions in which they can't practice good hygiene and physical distancing. Compared to whites, other racial groups have less access to healthcare, and are more likely to have other health conditions and co-morbidities (themselves the result of poverty and lack of healthcare access). This is a very familiar story in public health: disadvantaged groups are almost always more vulnerable to disease and have worse outcomes when infected.
Ed Yong at the Atlantic provided another excellent piece of reporting this week, on what went wrong in the U.S. response ("How the Pandemic Defeated America"). Despite a bit of hyperbole (I'm not sure "the U.S. squandered every possible opportunity to control the coronavirus"... just most of them), Yong nails so many important points. I am resisting the urge to quote vast chunks of the article, but will settle for this:
The COVID‑19 debacle has also touched—and implicated—nearly every other facet of American society: its shortsighted leadership, its disregard for expertise, its racial inequities, its social-media culture, and its fealty to a dangerous strain of individualism. SARS‑CoV‑2 is something of an anti-Goldilocks virus: just bad enough in every way. Its symptoms can be severe enough to kill millions but are often mild enough to allow infections to move undetected through a population. It spreads quickly enough to overload hospitals, but slowly enough that statistics don’t spike until too late... How will the U.S. fare when “we can’t even deal with a starter pandemic?,” Zeynep Tufekci, a sociologist at the University of North Carolina and an Atlantic contributing writer, asked me. Despite its epochal effects, COVID‑19 is merely a harbinger of worse plagues to come. The U.S. cannot prepare for these inevitable crises if it returns to normal, as many of its people ache to do. Normal led to this. Normal was a world ever more prone to a pandemic but ever less ready for one.
I believe this kind of stock-taking and analysis is critical, not to bash the U.S. or any other country, but because this is not the last major pandemic we will see in this century, and maybe not even in this decade. Among ways the U.S. wasn't ready: a weak and under-funded public health infrastructure (local health departments have cut 25% of their jobs just in the last 10 years!), FDA red tape that meant at least one U.S. lab was sending perfectly functional diagnostic tests to Africa before any tests were approved for use in the U.S., a national stockpile of medical supplies that had never been replenished after the 2009 flu pandemic and so was critically low on supplies, and a president with a strong track record of spreading misinformation who had intentionally purged government of career civil servants and experts and then undermined trust in the experts that were left. A friend of mine (and also an estimable scholar at Johns Hopkins SPH), Sarah Dalglish, is quoted at the end of the article as saying, “When you have people elected based on undermining trust in the government, what happens when trust is what you need the most?”
I also checked in on Sweden this week, as I hadn't in a while. The bad news: to date Sweden has about 10 times as many covid-19 deaths per million population (571) as other Scandinavian countries (Norway at 47, Finland at 60), and even more than the U.S. (491), although I'm quite sure the U.S. will catch up. The good news: Sweden's number of cases have dropped sharply, and deaths have dwindled to 1 or 2 per day. There is still a lot of debate over whether Sweden's strategy (of encouraging social distancing but not imposing a strict lock-down) was successful and to what degree it helped the economy. I myself wish I had an answer to the question of to what degree Sweden has achieved herd immunity, and whether this is part of why Sweden has flattened its curve so dramatically. Put another way, did new infections and deaths slow to a trickle because a significant number of Swedes became immune even though they were being exposed to infected people, or did social distancing (even though not mandated) prevent people from being exposed, even in the absence of herd immunity? Either way, the reality is that as active cases fall (Sweden now has just a few hundred new cases per day), it becomes increasingly unlikely for someone to be exposed to an active case. Just as epidemics can spiral up (increase exponentially), they can also spiral down.
The fact that the U.S. now has twice as many confirmed cases per million population as Sweden makes me think that it's not that Sweden reached a natural stop point for transmission (i.e. herd immunity), but rather that social distancing is continuing to keep the virus in check. At least, our continued high transmission in the U.S. indicates we are nowhere near herd immunity here.
I'm also interested in Sweden as an example of what happens when you don't close schools, even when there is high community transmission (which there was several months ago in Sweden, and primary schools never closed although high schools and universities did). This study (summarized here) is from a few weeks ago but still relevant. The public health agencies of Finland and Sweden compared data and found that keeping schools open, even with high community transmission (Sweden), did not result in higher transmission in the country than closing schools (Finland). There were NO deaths of primary students due to covid-19 in either country. There were some deaths of teachers, but Sweden found that mortality rates among teachers were no higher than in other professions. I'm not saying this example is universally applicable - there are a lot of ways in which the rest of the world isn't Sweden. Yet I don't see any evidence that Sweden's decision to keep schools open made an already widespread epidemic worse.
I'll say it again: I think we need to re-open schools (at least at the primary school level) as soon as possible, even though it's not risk-free for students or teachers. In epidemiology the counterfactual is really important. Yes, some teachers will die of covid-19 if teachers return to the classroom. Some teachers will die of covid-19 even if teachers don't return to the classroom and teach online from their houses, but are exposed to risk in other ways. The important public health question is not whether any teachers die if we re-open schools, but whether there will be significantly MORE deaths of teachers if we re-open. (I'm focusing on teacher deaths here because given how low risk of death is for children, I think it's far more likely that re-opening schools will be implicated in deaths of teachers than deaths of chidren, even if the U.S. doesn't manage to totally avoid student deaths as did Sweden and Finland.)
The answer to this, of course, is highly dependent on level of transmission in the community, and the chance that students (and teachers) are entering schools already infected. But we can't wait until community transmission (and risk of in-school tranmission) is zero - and even the most conservative school re-opening plans recognize that. I don't mean to sound at all hard-hearted, but rather point out that there are no risk-free scenarios here, unless one is in New Zealand, where there literally is no one with SARS-CoV-2 walking around who could infect you. As individuals and a society, we have to choose what level of risk we are comfortable with. And as I discussed in my July 2 post on this, there are a great many risks to children of staying out of school.
The Washington Post ran this important but depressing piece on vaccines this week, and I was reminded that even if we have a successful vaccine by the end of the year, we are likely a year or two away from a return to anything like normal life. Incredibly dedicated researchers have been working at a breakneck, superhuman pace to develop a vaccine, and there is now one vaccine being in a Phase 3 clinical trial (the last phase, in which it's tested on thousands of people). But even once we have a successful vaccine, it may take months or years to fully distribute (not even considering that a significant portion of Americans are anti-vax and will refuse it). The vaccine may only reduce risk by 50% (the FDA will require a vaccine to be 50% effective to be approved for release), may take weeks to confer even that level of immunity, and may require a second shot or boosters. Here's a good quote from the WaPo article:
Mulligan said he believes people should view vaccines in much the same way they have regarded reopening — as something that must occur in gradual phases to be safe and could even double back on itself as we learn more. Governments and companies are investing billions of dollars to ramp up the vaccine supply now, but even so, it won’t be possible to vaccinate everyone in the first week or even the first month after the first vaccine becomes available. The world will become safer, bit by bit, not all at once.
Which means that there is no silver bullet coming any time soon that is going to allow us to safely reopen schools. For the foreseeable future, our tools are going to remain what they have been: hygiene, physical distance, and stopping all non-essential risky activities so that the essential risky activities can go forward (kids returning to school being very high on my list of "essential").
As schools have re-opened in some U.S. states in the last week or two, there are already reports of outbreaks in schools (and subsequent closures), although I haven't seen any reports of serious illness or deaths among students or teachers/staff. But once again, the counterfactual is important. Would just as many children (and adult staff and teachers) be infected if they were out of school? Dare I say that some children might be socializing in more risky ways if they weren't in school, such as closer than 6 feet away and without masks?
Chicago schools announced this week that they will be online only, and I am waiting (with dismay) for schools in my current district on the outskirts of Chicago to follow suit. There remains only one major metropolitan area in the U.S. that is planning even partial in-person school this year: New York City. Where the pandemic hit hard and early, but which has now maintained a fairly low rate of transmission for almost 2 months, and which currently averages less than 20 deaths a day. Here's New York state's curve (of new cases per day), from Johns Hopkins Coronavirus Resource Center.
Once again, controlling this virus can be done - and if urban, crowded NYC can do it, any place can. I hope it doesn't take the kind of catastrophic death rates seen by NYC to make people in other places get serious about the simple, hard behavior change that needs to happen to slow transmission (masks, hygiene, but most of all physical distance). Maybe the reality of our kids being out of school for months more will finally galvanize Americans to do what it takes?