Back to the blog as the U.S. heads over a cliff called Thanksgiving

I didn’t mean to disappear from this blog for 3 months! Life as a new professor has been sprint after sprint after sprint, and in the crush of daily lectures and dual modality teaching and all the things I've never done before, I've hardly had time to think about the pandemic, much less write about it. For the U.S. and the rest of the northern hemisphere, the pandemic has been in a terrible but relatively stable state through late summer and fall. My life has been surprisingly normal in that time, at least compared to so many in the U.S. and world. My kids have been attending in-person school half days (with all the precautions, including open windows in the classroom), and their elementary school reported its first infected student only last week. I’ve been teaching students mostly in-person (with all the precautions), with a few online students, and other students (and me) cycling in and out of isolation/Zoom attendance based on symptoms or exposure. Miraculously, there have been fewer than 20 SARS-CoV-2 cases among students on my campus (the New York Times map here shows only 9 cases for my college, Wheaton College in Wheaton, IL). I hope to write more on schools sometime soon, but I'm more and not less convinced than I was 3 months ago that schools are not hotbeds of transmission, as long as students are wearing masks and keeping physical distance.

I taught my last in-person class for the semester on Friday; our remaining two weeks of instruction before Christmas will be online, and the college has already announced that we will be starting spring semester online. I feel incredibly proud of Wheaton students for all the mask-wearing, rule-following, and sacrifice that got us to Thanksgiving still in-person (at the beginning of the semester, I really thought we’d be online within a month) – even as I know that these measures have been incredibly tough on many student’s mental health. I feel incredibly blessed that my life has been so relatively normal in this time period – even as it’s been incredibly tough to not be able to do things like playdates and having friends over for dinner and attending church and being within hugging distance of anyone not in my immediate family. I feel incredibly grateful that of my multiple extended family members who have now had covid-19 (including one who spent time in the ICU), everyone is still alive.

Enough about me. Let’s talk about the pandemic. My attention has snapped back to the numbers in the last few weeks as the U.S. epidemic has entered a new, critical, terrifying stage. (I’ll distinguish here between terrible and terrifying – we’re now at code “utterly terrifying”.) Here’s one of the New York Times’ excellent visualizations (and if you go here, you can explore their vast array of graphs and resources interactively).

The bottom line is that we've gone from very high-prevalence epidemics in the spring in New York and a handful of other coastal cities, to a surge over the summer (in which 4th of July gatherings are clearly implicated), to uncontrolled spread in much of the country and especially the rural middle of the country. (Following graphic also from NYT.)

On Friday Nov. 20, when I started working on this blog, we had nearly 200,000 new cases in the U.S. and nearly 2,000 deaths. (As always, counts over the weekend are a bit lower.) The epidemic is doubling in less than 3 weeks; according to NYT the 7-day moving average (which removes some of the “noise” seen in daily counts) was 170,539 on Nov. 21, nearly double the 7-day average of 85,592 on Nov. 2. I fully expect the epidemic to double again in the next 7-10 days as a result of the Thanksgiving holidays. A national survey a couple weeks ago found that 40% of Americans are planning to attend Thanksgiving dinners of over 10 people. If you want to know what your risk of encountering an infected person at a Thanksgiving dinner of 10 people is, this handy map will show you for your county. (Of course, this is an estimate that is based on a number of assumptions, including how many undetected infections there are for every confirmed case.) For my current county of DuPage County, IL, the risk is 36% (based on a ratio of 5 undetected infections for every 1 confirmed case). In many counties in the U.S., the risk is far higher. The CDC is begging Americans not to travel over Thanksgiving weekend, while the White House is criticizing my home state of Oregon for rules banning more than 6 people at a gathering, but I’m just proud and relieved that Oregon is not bright red on the map above.

I want to scream as I read reports of busy airports and full flights. But the real risk, in my opinion, isn't airports or planes. It's when a group of people not living under the same roof gather indoors and eat and talk for hours. I can’t think of a situation more conducive to the spread of the virus. I think we're increasingly learning that you're pretty unlikely to get this virus from passing someone in the produce section, and highly unlikely to get it by touching an apple they touched. I want to see us distinguish better between levels of risk - and at the pinnacle of that risk pyramid is prolonged close contact indoors. Although masks reduce risk, having everyone at your Thanksgiving dinner mask up after the pie does not in any way make the gathering safe. We're entering a phase of the U.S. epidemic in which transmission is mostly happening in small, indoor gatherings - and until we hit herd immunity (which we're still nowhere near), there is no real upper bound to that soaring red line of new infections.

As always, most transmission of this virus is ridiculously easy to prevent, as long as we do the simple and painful things. I think I could actually boil those "things" down to ONE thing: Don't spend time indoors with people you're not living with, period. Yet a vast number of Americans are planning to do just this in the next week. It’s not hard to do some simple math. Let’s assume that just half of Americans who wake up on Thanksgiving Day infected with SARS-CoV-2 sit down at a table with uninfected people, and that each infected person infects 4 more over a day of laughing and feasting. Boom – we just doubled the epidemic, and there’s still lots of time to double it again before Christmas.

Let’s talk about deaths. The U.S. just passed the 1500 mark for the 7-day average of covid-19 deaths per day. (The Worldometers visualization below shows a 7-day average of 1510 deaths as of Nov 21.) Reported deaths are going to lag behind cases by several weeks, as it typically takes several weeks for people to die (and additional time for the death to be reported). The horrifying implication is that by the time we see a big surge in cases, it’s already too late to stop the big surge in deaths that is coming.

How many people will die, given the recent huge upsurge in cases? This excellent Atlantic article presents evidence from genomic epidemiologist Trevor Bedford that there’s an average 22-day lag between cases and deaths, and that given this lag, the “lagged case fatality rate” is 1.8%. (This does not mean that 1.8% of all infected people will die of covid-19, since the case fatality rate is only calculated among confirmed cases, and for every confirmed case there is an unknown number of mild or asymptomatic infections that don’t get diagnosed.) In the visualization below (from the article), you can see how closely deaths mirror cases given this lag, and also that the ratio of deaths to cases hasn’t changed much in the last few months.

The Atlantic article predicts we'll hit 2,000 deaths a day by the first week in Dec. I can't imagine how death counts won't double in the next few weeks, given the fact that the case count just doubled in 19 days, so I fully expect we'll be seeing an average of 3,000 deaths a day by mid-December. (Multiplying our current 170,000 average daily cases by a case fatality rate of 1.8% yields 3,060, a nearly identical estimate.) If the epidemic doubles again over Thanksgiving weekend, we could see death rates double again by Christmas, to 6,000 deaths a day – especially as hospitals become increasingly overwhelmed. In many parts of the country, they are already overwhelmed. More than 1 in 5 hospitals is already facing staffing shortages, and Iowa (which is bright red on the map above) has already run out of staffed beds (data in article linked above).

If these predictions seem apocalyptic or exaggerated, they’re really not – they’re just simple math based on what happens in exponential growth. What’s critical here is the reproductive number (R0 or R-naught). If each infected person infects, on average, less than one other person, the epidemic winds down and eventually disappears. If each infected person infects, on average, more than one other person, the epidemic increases exponentially – and if each infected person infects on average just 2 others, the epidemic doubles. We've seen this happen in the last 19 days, and the only question is whether we'll see the epidemic double again, and double again.

We’ve had perhaps the best news of the pandemic in the past week, with both Pfizer and Moderna announcing that their vaccine trials have shown efficacy of 95% and that they were applying for FDA emergency authorization. This level of efficacy is way higher than anyone was expecting; FDA was prepared to authorize a vaccine with efficacy as low as 50%, and efficacy of 70-75% would have been good. So efficacy of 95% is almost unbelievably good news.

But we have to temper that good news with a cold dose of reality: Vaccines don’t save lives, vaccinations do. Meaning that the most effective vaccine in the world doesn’t help us unless we can actually get people vaccinated. The challenges in this case are daunting, historic, mind-bogglingly massive. Both vaccines have to be administered as two injections several weeks apart, and have to be stored at cold temperatures. (My understanding is that the Moderna vaccine only needs a normal cold chain, but the Pfizer vaccine needs super-cold temperatures that can only be provided by dry ice.) There are multiple million-dollar questions to be answered over the coming months (given the stakes to human well-being and economies – maybe billion or trillion dollar questions?) – How fast can hundreds of millions, or even billions, of doses of vaccine be produced? How fast and how can they be distributed? Will people be willing to be vaccinated? Of the willing, how do we prioritize who gets the vaccine first? The scientists have done their job to deliver successful vaccines at an unprecedented, absolutely stunning pace. Now it’s up to the supply chain managers, public health professionals, and health care workers to translate successful vaccines into a successful global vaccination campaign – also at lightning speed.

I can’t find words to express my rage that the Trump Administration, after delivering on Operation Warp Speed (the public-private partnership that guaranteed massive U.S. government purchasing of successful vaccines), seems to be devoting no discernible energy or leadership towards a vaccination strategy. Instead, the administration’s full focus seems to be on contesting an election the whole world knows is over, and obstructing the ability of the Biden administration to hit the ground running in responding to this national and global crisis. Biden and his team can’t so much as talk to Fauci and the other leaders of the U.S. pandemic response, much less access the vast government resources guaranteed to an incoming administration and that in any normal election they would have had access to within days of the election. I wish with all my heart that public health was never political – that we could all just look at the same evidence and draw the same conclusions about what would save the most lives, and together pursue a sound, evidence-based strategy across party lines. In reality, this debacle is just the latest reminder that public health can never escape politics, particularly small-minded, self-serving, disgustingly cold-hearted and myopic political calculations. Make no mistake: the Trump’s catastrophic lack of leadership at this critical juncture is costing lives.

I’m not going to make any promises about how often I’ll be able to blog in the coming weeks and months, but I am hoping to blog more regularly as my semester winds down and the U.S. epidemic heads for an incredibly dire winter. I'm hoping also to very soon turn my attention to what is going on in the rest of the world, including South Africa. I think I’m going to give up producing my own visualizations, as there are so many much prettier and more sophisticated visualizations out there. As always, I’d love to know what you’re reading and finding interesting, to add to my ever-longer and never-finished reading list.