July 30 covid-19 update

First, an apology to those of you who got the notification about 15 hours ago about a new blog post, only to follow the link and find nothing there. I was working on this post last night and accidentally published rather than saving it, and then immediately took it down. I'll try not to do that again!

I am changing my schedule to post Friday morning U.S. time / Friday afternoon South African time (including data through Thursday). Since moving time zones a couple weeks ago, I can't stay up late enough on Thursday to wait for the U.S. data for that day. Last week I posted Thursday night (my time), then had to update the numbers in the graphs on Friday. So from now on, I'm going to wait for Friday morning my time to post.

On to business. The U.S. has now surpassed 4.5 million confirmed cases of SARS-CoV-2 and 150,000 deaths from covid-19. In the past week alone, the U.S. has seen nearly half a million new cases, which is more than 10% of all confirmed cases to date. The death count has been over 1,000 for the past 4 days, and in the past week the U.S. has seen over 5% of total deaths to date. As you can see from the curves below, the rate of new cases is increasing. In fact, daily cases in the U.S. have tripled since mid-June, and daily deaths have doubled since early July. I'll show graphs of daily counts, taken from worldometers.info, after my graphs showing cumulative data.

In South Africa, the number of cases is still doubling every 3 weeks and 18% of cases to date have been reported in the past week, although new cases are declining somewhat day by day. Given that South Africa's testing is increasingly not keeping up with the epidemic, I'm not totally convinced that declines in daily confirmed cases reflect an actual slowdown in new infections (see discussion of test positivity later in this post for more on this). Recall from discussions in previous posts that actual infections greatly outnumber confirmed cases, likely by about 10 to 1, although we don't know the exact ratio in any country. Data on confirmed cases show that the curve is flattening over the past couple weeks. You can see this most clearly in the log scale graph of cases, which plots exponential growth as a straight line. If this reflects a decline in actual infections, we should see the curve for deaths start to flatten in a few weeks. So far, there is no decline in deaths, and South Africa recorded 1,872 covid-19 deaths in the past week, which is nearly a quarter of all deaths to date.

And here are those visualizations from worldometers.info:

I've been thinking this week about that long-ago time in mid to late March when I first started producing these graphs, and really had no idea how the pandemic would progress in these three countries. At that point, Italy had the world's worst epidemic, news reports told of Italian hospitals stretched past capacity and horrific deaths tolls, and Italy had gone into strict lockdown but the curve of new cases had not yet flattened. In those very early weeks of the U.S.'s epidemic, new cases in the U.S. were following Italy's curve so closely (about 1 week behind) that it was almost uncanny, and then South Africa epidemic took off almost exactly 3 weeks after Italy and 2 weeks after the U.S. As a reminder, Week 1 in these graphs is the week at which each country reached 100 confirmed cases, and that's why Week 1 is a different date for each country.

How different the epidemics look now. Italy went into a very strict lockdown for almost two months (March 9 to May 4), and emerged with its epidemic contained. Italy registered just 3 deaths from covid-19 on July 30. Other countries have also contained their epidemics very successfully. As I discussed two weeks ago, New Zealand has now re-opened schools and everything else, as there are no new infections. South Korea, a country of 52 million people, has seen only 300 deaths from covid-19, and has also essentially ended its epidemic. [Note to any readers from Italy, New Zealand, South Korea, or any other countries that have successfully controlled their epidemics - I'd love to hear your comments on your country's response, the extent to which life has returned to normal, and especially if I've mischaracterized anything. Please leave a comment!] The U.S. and South Africa, in contrast, haven't even stopped exponential growth of their epidemics, much less gotten transmission down to zero.

This is a virus that's spread person-to-person, and the only way to stop transmission is by stopping person-to-person contact. Masks will decrease (but not eliminate) risk of transmission, although of course we should all be wearing them when we can't maintain physical distance. Robust testing services (available, quick, free) and contact tracing are critical, although the U.S. and South Africa first have to slow down transmission enough for contact tracing of individual infections and outbreaks to be possible. (Contact tracing isn't feasible in the face of widespread community transmission; you can't contact trace an entire population.) But the key to all this is stopping the person-to-person contact that spreads the virus.

If a country could actually stop all person-to-person contact outside the household, and provide robust testing and effective treatment for those who needed medical care, the epidemic would be completely over within a few weeks. There would be no infected people left to infect other people (with the exception of people still receiving treatment in hospitals), and we could all go back to normal life. The problem is that we can't totally stop all person-to-person contact - at a minimum health care personnel still have to go to work, everyone still has to buy food, and some people have to ensure that food is produced and sold and provide other truly essential services. It took Italy about 2 months of strictly limiting person-to-person contact to what was truly essential to break the back of its epidemic, and since then, transmission has stayed very low (aided by ongoing testing and contact tracing). Other countries have had similar success, and many countries in Europe have followed a trajectory similar to Italy's.

In the U.S., we haven't been willing to accept the short-term pain of a strict lockdown and drastic change to our quality of life for a couple of months. As a result, we're dealing with the ongoing pain of an uncontrolled epidemic, high death tolls, agonizing decisions over whether to reopen schools (and whether it is safe to send our kids back to school even if schools do reopen), and a lack of freedom to resume our lives in safety. The Americans who keep protesting for their rights to not wear masks make me absolutely nutty; I would so much rather have the freedom to have a social life again and have my kids have the freedom to go to school and play with other kids, than the freedom to not wear a piece of fabric over my face. We had an opportunity in the U.S. to get on top of this epidemic over the summer, and reduce transmission enough to safely send our kids back to school in the fall. We've wasted that opportunity because as a society we weren't willing to put up with short-term pain for long-term gain.

Note that I'm not speaking about South Africa here because I think the issues surrounding lockdown are a lot more complex there. Stay-at-home orders are a completely different proposition when people live in shacks without running water or electricity, can't get grocery delivery, and don't eat unless someone in the household worked that day or that week. I still think the only way South Africa can slow its epidemic is by drastically limiting person-to-person contact, but I'm aware that lockdown restrictions have to be balanced with people's need for survival in a way that is much more stark than in the U.S. So I'm treading carefully when speaking about South Africa.

I've been trying to get a handle this week on what is happening in various states of the U.S., in terms of the epidemic and school re-opening. Oregon, my home state, announced this week that a county needs to see fewer than 10 new cases in a week per 100,000 population for 3 weeks in a row for all grades (kindergarten to 12) to go back to school, but if cases are below 30 (per 100,000) kindergarten to 3rd grade can return. Test positivity (percent of people tested who test positive) also has to stay below 5%, as recommended by the World Health Organization. Most states in the U.S. are above the 5% threshold, which basically means they are not testing enough to actually track new infections, and testing is not keeping up with the epidemic. If you want to know the test positivity in your state, look here. That site also has a good explanation for why a low test positivity rate is so important. In Illinois, where I live now and where my kids are supposed to start school in just over 2 weeks, schools are still scheduled to re-open in person, but there don't seem to be clear benchmarks for when to open or close schools, beyond general guidelines for state reopening.

I have scoured the internet for info on school reopening by state, and this site is the best I have found. I have not scoured all the information on that page, but with a quick skim, I am somewhat shocked by how many states haven't issued any clear guidance since May or June. (Props to Oregon for being ahead of the curve, and for such a sound public health strategy.) Furthermore, many states have only issued guidelines for safe school reopening (such as distancing and face masks), and not actual epidemiological benchmarks for when it would be safe to reopen. I noted the huge range in guidance for states that have released specific benchmarks. Minnesota has specified that counties can re-open schools if they have fewer than 90 new cases per 100,000 population in a 14-day period, compared to Oregon's cut-off of 10 new cases per 100,000 population in a 7-day period. Florida and Texas, two of the states with the worst surges of infections in the past couple months, have both announced they will re-open schools 5 days a week. According to the New York Times' tallies, Texas is at 193 cases in the last week per 100,000 residents and 12% test positivity, while Florida has a whopping 333 cases in the last week per 100,000 population and 19% test positivity. Whatever is guiding Texas and Florida's decisions, it's not epidemiology.

The U.S. National Academies of Science, Engineering, and Medicine released a report this week called Reopening K-12 Schools During the COVID-19 Pandemic. You can download it for free at that URL. I admit that I only read the Summary, but felt frustrated that guidance wasn't more specific, i.e. epidemiological benchmarks for when schools should reopen and for what grades. I feel compelled once again to say that I find the lack of national leadership in the U.S. on this issue completely incomprehensible and counterproductive. I don't understand why the U.S. isn't issuing national guidance with clear benchmarks for school re-opening, even if they can only offer recommendations and not mandate state-level decisions.

If you want to see more data for your state or county, here are some great resources. I've mentioned the New York Times' coronavirus visualizations before, but they're still awesome. For the U.S., you can drill down to county level and see cases per 100,000 population. There's also a world map shows cases per 100,000 population for every country in the world. (I have serious doubts about the quality of some of that data, simply because many countries aren't testing, so I would interpret basically all of Africa apart from South Africa with great caution.) They even offer data by region for a handful of countries. The Johns Hopkins dashboard has graphs for each U.S. state showing cases, deaths, and a timeline of opening and closing decisions. I can't speak for each state, but for the handful of states I looked at, there was a clear trend of re-opening even as daily cases were increasing somewhat, and a sharp surge in new cases after re-opening. More evidence of squandered opportunities over the past few months.

South Africa, I haven't forgotten about you. South Africa is at 18 average daily cases per 100,000 population, which compares to 20 in the U.S. (nationally). I can't find a source anywhere online that tracks test positivity, but in the last week (i.e. comparing data from July 23 and 30), there were 285,943 tests performed, with 74,117 of those positive. This yields a test positivity of 26%. This compares to a total test positivity since the beginning of South Africa's epidemic of 17% (obtained by dividing confirmed cases to date by all tests to date). [UPDATE: A reader sent me a link to Twitter account @rid1tweets, who has been producing visualizations of test positivity in South Africa. Test positivity has increased from around 3% in April to over 25% currently!] In other words, South Africa's test positivity has increased over the course of the epidemic, which means that testing is increasingly not keeping up with actual transmissions. South Africa is way above the 5% test positivity which the World Health Organization says is needed to re-open schools. Based on continued high transmission and high test positivity, I would say the South African government made a good, evidence-based decision in closing schools.