June 4 covid-19 update

As the U.S. has been rocked by violence and injustice in the last ten days, I've been less focused on the pandemic than I have in months. It's been a painful reminder of all the ways that people are in danger quite apart from the coronavirus - although of course Black Americans are still dying at more than twice the rate of other Americans from covid-19. Public health is social justice, and all of the injustice needs our attention. I know there have been many reports of police brutality during South Africa's lockdown, but this story of a 15-year-old boy in Eswatini being killed by police while playing soccer during lockdown hit me particularly hard, maybe because I used to live in Eswatini. (Eswatini, by the way, reports only 3 covid-19 deaths to date.)

As crowds of people gather to protest around the U.S. (some wearing masks, some not), lots of people are concerned that the virus may be finding a perfect opportunity to spread. What I have said in the last couple post is still true: the U.S. is not seeing a large spike in new cases as states re-open, and deaths have declined (to around 1,000 per day, down from around 2,000 per day a month ago), but we're not really winning against the virus, either. We're increasingly learning that the virus is passed much more easily indoors than outdoors, and summer temperatures may further slow down its spread. There's so much we still don't know.

I was fascinated to read about Japan this week, which has followed a fairly unique path in controlling its epidemic, resulting in very low numbers of deaths despite having one of the oldest populations in the world. The Japanese government apparently couldn't legally impose lockdown, although it did urge people to voluntarily stay home and limit their social interactions (much like Sweden). Surprisingly (to me at least), Japan has also done relatively little testing and contact tracing. What it has done is emphasize the 3 Cs: avoiding closed spaces, crowded places, and close-contact settings, and especially avoiding "clusters" in which these three overlap. I'm sharing this info and the flier (available online here) because I think it's good guidance for all of us as life returns to something more normal and we constantly have to weigh the risk level of various activities. Plus, this strategy has clearly worked in Japan - along with widespread mask wearing.

Sweden, like Japan, has not enforced a strict lockdown although it has banned large gatherings and urged its citizens to stay home and social distance (and they have largely complied). We've discussed that despite this, Sweden has seen far higher fatality rates than neighboring countries. Now the architect of Sweden's response, epidemiologist Anders Tegnell, has stated publicly that Sweden may have gotten it somewhat wrong. Here's a quote from Tegnell: “If we were to encounter the same illness with the same knowledge that we have today, I think our response would land somewhere in between what Sweden did and what the rest of the world has done... Clearly, there is potential for improvement in what we have done in Sweden." The same articles notes that avoiding lockdown has not meant that Sweden has avoided severe damage to the economy. Sweden is facing its worst economic crisis since WWII and GDP is down 7% in 2020 (on par with the rest of Europe).

Brazil is another very sobering case study about the costs of not having a lockdown. Brazil now has the second most cases of any country in the world (after the U.S.) and the third most deaths of any country in the world (after the U.S. and U.K.). The University of Virginia's dashboard shows 1,336 deaths today (June 4), but I've seen even higher numbers in news reports. Whatever the exact number, it's staggering and on par with the ~2,000 deaths per day that the U.S. was seeing a month ago (Brazil has about 2/3 of the U.S. population), but Brazil's epidemic doesn't yet show any signs of slowing down. President Bolsonaro has been widely criticized for not only failing to provide effective leadership in the face of the pandemic but actively discouraging social distancing and other preventive measures in order to bolster the economy.

There's a lot of debate right now about what effect different age distributions will have on fatality rates from the pandemic, given that older people are so much more likely to die. South African friends, you may have seen the criticism from the private research group Pandemic Data Analysis (PANDA), which I think is a group of actuaries, claiming that South Africa's death toll will be much lower than predicted by other modelers due to its relatively young population. I think it's valid to ask the question, although we should also acknowledge the other big unknowns, such as whether risk of death will be higher in people infected with HIV and TB (which if true would increase fatality rates in South Africa compared to many other countries). However, Brazil also has a relatively young population, although not quite as young as South Africa. To list some countries from youngest to oldest population: median age is 27 in South Africa, 33 in Brazil, 38 in the U.S., 41 in the U.K. and Sweden, and 48 in Italy (numbers rounded to aid comparison). And yet Brazil has reached the same daily death toll as the U.S., despite its younger population, with factors such as the lack of lockdown and weaker healthcare system doubtless contributing.

While South Africa has the advantage of an even younger population, it has the disadvantages of a weak healthcare system and other serious co-morbidities, which makes me think it's not outside the realm of possibility that it could reach the same level of deaths per million seen by the U.S. and Brazil (which would be about 400 deaths a day). So far, South Africa's case fatality rate (CFR) is 2/3 that of the U.S. and less than 1/3 of Italy's (see figures in graph below). I'm sure that these differences in CFR (and deaths per million population) are largely due to differences in ages of who is infected, but age can't be the only factor. The U.K., for instance, has lost a higher number of people to covid-19 per million than Italy, but has a younger population. The U.K. is tragically following the same pattern we see in the U.S. of disproportionate impact and deaths among people of color.

Another very sobering fact that emerges from comparing Brazil to South Africa is that South Africa is actually seeing a higher rate of growth (of new cases) right now than Brazil. South Africa has doubled its cumulative cases in the last 13 days and doubled its cumulative deaths in the last 12 days, whereas it has taken Brazil 16 days and 18 days, respectively, to do the same. This is a point I've been making for several weeks: South Africa's current trajectory is not good, but cases and deaths still remain relatively low because the early lockdown prevented exponential growth early in the epidemic. Brazil and South Africa reached 1,000 cases in the same week: Brazil on March 21 and South Africa on March 27. On April 21 (1 month after its 1000th case), Brazil had 43,079 cases and 2,741 deaths. On March 27, South Africa entered a strict lockdown and a month later (on April 27) had 4,793 cases and 90 deaths. Granted, Brazil has 210 million people to South Africa's 59 million people, so we would expect Brazil to have more cases and deaths. Yet it's clear to me that South Africa's lockdown had a major impact, and is responsible for a lot of lives saved.

You can see South Africa's current trajectory in the graphs below, and particularly in the middle (log scale) graphs. A reminder that CFRs are calculated with a 2-week lag (dividing deaths to date by cases 2 weeks ago to account for the lag between diagnosis and death). These graphs aren't showing the fact that Western Cape has nearly two-thirds and more than three-quarters of South Africa's deaths to date, but CFR in Western Cape is actually somewhat lower than nationally (3.8%). More info at WC's coronavirus dashboard.

A final note about a topic that many of us are keenly focused on: the re-opening of schools. I know schools in South Africa are re-opening, and U.S. schools are trying to figure out what to do when the school year begins in August/September. A former colleague of mine has written an op-ed and longer journal article arguing for re-opening schools, and that this will not greatly increase risk for either children themselves or the community at large. Here's another article making similar points. We've known since early in the pandemic that children have a very low risk of dying from covid-19 or experiencing other serious complications, notwithstanding that a very small number seem to have very serious complications such as symptoms resembling Kawasaki disease. It's been harder to answer how many children are actually becoming infected and how effectively those children pass the virus to higher-risk individuals (adults and especially older adults), even if the children experience mild or no symptoms. Put simply, the concern is not just how re-opening schools will impact children themselves, but transmission of the virus in the whole community. But a few months into the pandemic, there is persuasive evidence that children do actually have a low risk of infection and are less contagious than adults if infected (the science behind this is fascinating - just read the articles). In fact, transmission studies of thousands of people have failed to find a single case of child-to-adult infection. This is all good news, and argues for re-opening schools as well as sending kids back to school being a safe choice, for those of you on the fence about that.