Probable Regional Peak Reached on April 10th, 11th

A notable trend emerged in this week's data for confirmed COVID-19 cases in DC area, that the day-to-day increase was less than 10% for five straight days. This stands in contrast with mid and late March, when that figure was consistently higher than 15%. Hence the obvious question - has the DC area hit a peak.

Is There A Peak?

Probably.

It may sound simple, but a single day dip does not constitute a trend. Thus, the first thing we would need to see was consecutive days of general downward movement to make the call. It would appear we have that, with an apparent peak date of 10-11 April. 

It was key for me to see several days of the <10% figure, both to account for day-to-day noise, as well as day-of-week effects. By the latter, I mean the trend where weekends have consistently shown apparent dips, which later prove to be somewhat artificial. I'm not going to expand on the issue at this point, but I think the 12 April datum really shows what I mean.

One admitted blind in my analysis is testing capacity. As this is an amateur pursuit, I haven't had the time or desire to add that element to my numbers, so I can't rule out the impact of testing capacity.

Does The Apparent Peak Mean Anything?

Probably.

I'll use data from Italy, which has been hit quite hard by COVID-19, to make this case. Italy instituted a nationwide lockdown on 9 March. It had an absolute maximum peak (thus far) on 21 March, 12 days following lockdown.

Here in the US, Washington state was hit early with COVID-19. Their governor instituted a stay-at-home (i.e. lockdown) order on 23 March. A quick review of case data indicates that the new infection peak was on 3 April, 11 days following lockdown.

The governors of Virginia and Maryland, and the mayor of DC, instituted lockdown orders on 30 March. If we use the Italian and Washington state examples hitting peak 11-12 days following lockdown, we would anticipate a peak on or around 10-11 April. This obviously lines up very well with the DC region's apparent peak dates of 10-11 April.

What's Next?

The first thing is to continue to look for <10% growth, and gradual decreases beyond that as well. For instance, the 16 April datum shoots up a bit, but that alone isn't sufficient to buck the recent trend.

Beyond that, it's still to be seen. Italy is still under lockdown through all of April and into May, so their experience may provide insight as to what to expect here in the DC area. I peeked at the Chinese data*, and it seems that the tail for new cases came close to zeroing out about four weeks after peak. This would suggest 8-9 May for the DC area for when new cases may be very low, but I wouldn't put nearly as much faith in this estimate quite yet.

By Phoenix7777 - Own workData source: 疫情通报 Outbreak notification, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=87342125

Interestingly enough, China instituted a lockdown for Hubei Province (the epicenter of the outbreak) on 23 January. With a peak date of 4 February, the spread in China was 12 days following lockdown, right on par with the apparent spreads for Italy and the DC region.

*I know there's a considerable discussion regarding the quality of Chinese data, and I will not be engaging in it. I simply don't have the time or desire to do so.