The State of COVID-19 in D.C.: Miserable and Unclear as to How We Improve

On the whole, this week was worse than even last week, with only Ward 1 seeing a slight decrease in prevalence–though that simply be a result of less testing. The entire city and all wards, including the ‘low prevalence bastions’ of Ward 2 and 3, are now well above the German rollback threshold of 50 new cases per 100,000 per week (0.05% in the second column below):


Ward one-week prevalence one-week % pos. two-week prevalence two-week % pos.
1 0.212% 3.9% 0.376% 2.4%
2 0.152% 2.5% 0.308% 1.8%
3 0.116% 1.6% 0.220% 1.5%
4 0.256% 5.3% 0.462% 4.2%
5 0.283% 5.6% 0.468% 3.7%
6 0.259% 3.4% 0.441% 2.2%
7 0.321% 8.4% 0.551% 6.7%
8 0.245% 6.0% 0.447% 5.3%
D.C. total 0.235% 3.8% 0.416% 2.8%

For context, in column two, the ‘good place’, which is one new case per 100,000 people per day, is 0.007%.

Using the same method D.C. health does to calculate R(t), but with unadjusted new case data*, R(t), as of Sunday, was 1.26, meaning that every five days, give or take, the number of daily new cases would increase by 26%. As I noted last week:

While next week probably won’t be much worse, partly because travel-related COVID-19 cases won’t really start hitting the system until late this week, it’s going to be bad soon. Deaths are going to keep coming through–right now, we’ve had twenty deaths in the last two weeks–and D.C. will likely join the 1/1000 residents COVID-19 deaths club by the end of the year.

Barring a miracle, joining the 1/1000 residents COVID-19 deaths club by the end of the year is definitely going to happen. Given the two week prevalence and significant undertesting in most wards (percent positive rate > 3%), I think the most optimistic scenario is that 0.5% of people are infected. If you encounter 20 people in a day, that’s about a ten percent chance one or more of them would be positive.

Instead of noting, as I have every week, how this is a failure of governance at both the state and federal level, and that we need to take the obvious steps of shutting down indoor dining (and bars), and other gathering places, and limiting gatherings to households, I want to look at the new case data over time and the R(t) over time (calculated as above). Here are the new cases per day, starting with June 1:

Screen Shot 2020-12-06 at 11.11.30 AM

Here’s R(t), starting with June 1:

Screen Shot 2020-12-06 at 11.11.43 AM

R(t) essentially is a rate of change in new cases, and so mirrors the new case data.

You’ll notice things took a turn for the worse in late October. I personally noticed that mask wearing appeared to drop then, especially among people who seemed to have arrived recently in D.C. (young intern-ish lanyard types), though this is obviously anecdotal. At the same time, we had an uncharacteristic and very long spell of good weather, and restaurant dining was doing well; my anecdotal observation was those dining often didn’t appear to be from the same household. The key point is most of the increase in cases happened well before Thanksgiving and its associated travel.

If there is any good news to be had, dining really seems to be in decline now that winter has arrived (dining outdoors isn’t really pleasant in 40ºF weather no matter how many heaters you have–and if it is, then it’s probably ‘outdoors’ dining, meaning it should be treated like indoor dining). Not having people from different households sitting three to four feet away from each other and thereby breathing on each other probably will lower transmission. The mask stupidity of October seems to have waned as well.

My guess–and it’s just a guess–is that new cases will level off and maybe even decline a bit until after New Years, when people returning to D.C. will seed a whole new round of infections, and that seeding likely will continue through early February due to the changeover in administrations.

I wish I had happier news, but, to me, this is the optimistic scenario for D.C.

As usual, even with these horrible data, we still could be only around six weeks away from returning to normal-ish, but we intentionally remain six weeks away from safely returning to normal-ish because we’re unwilling to do what it takes to make that happen.

Anger isn’t the appropriate emotion, rage is.

*It’s unclear what parameters D.C. Health uses. Also, they have data on the onset of cases, which will lead to slight differences, and they also remove duplicates. For these reasons, D.C.’s R(t) estimates are ten days behind the new case data. That said, the new case estimates from ten days before usually jibe pretty well with D.C.’s estimate.

This entry was posted in COVID-19, DC. Bookmark the permalink.