I know that everyone has been (justifiably) distracted by other things, but the #COVID19 epidemic in the US is looking pretty dire with 125,552 confirmed cases reported Friday by @COVID19Tracking. 1/10
Confirmed cases have continued to tick up across the US, though with the Midwest and Mountain West contributing to most of the recent increase. Data from @COVID19Tracking. 3/10
If we look across states and plot confirmed cases on a log scale we can see a steady linear-on-a-log-scale trend in the past 4-8 weeks across most states. This pattern is indicative of exponential growth. 4/10
Exponential doubling of COVID-19 is proportional to the instantaneous reproductive number Rt at a particular point in time. If Rt is greater than 1 then the epidemic is growing, if less than 1 it's shrinking. In the following I plot Rt estimates from rt.live. 5/10
Broadly, Rt varied widely across states in May-Aug corresponding to differing societal behavior and state-level "reopening" policies, but during Sep and Oct states have largely equilibrated in terms of Rt. 6/10
Notably, 49/50 states are currently estimated to have Rt > 1 and in 40/50 cases Rt is estimated to be larger on Oct 28 than on Oct 14. 7/10
This indicates that the current surge has no evidence yet of cresting. It's exponential "velocity" is the same or greater than two weeks ago; we've not yet seen signs of deceleration. 8/10
I've written previously about the dynamic relationship between COVID-19 cases, deaths and societal response (
I expect both to play a role in countering the current surge, but I won't hazard a guess at this moment as to when this occurs and what we eventually suffer in terms of daily case load. 10/10
• • •
Missing some Tweet in this thread? You can try to
force a refresh
The US is reporting over 2000 deaths per day from Dec 1 and I believe will do so consistently throughout December based on daily case loads above 120k starting early November. 1/4
A drop in reporting over Thanksgiving weekend has made for some difficulty in directly comparing 7-day averaged deaths, but the trend is clear. Red bars are daily reported deaths from @COVID19Tracking and black line is 7-day sliding average. 2/4
The simple projection of 1.7% of reported cases into deaths 22 days later has remained largely accurate, although drop of reporting during Thanksgiving weekend is quite clear. We'll know soon whether 7-day average returns back to projection. 3/4
After posting about sharply rising #COVID19 cases Friday, there were multiple replies to the effect of "but deaths aren't going up". As should be obvious to most at this point, (reported) deaths lag (reported) cases. This thread investigates. 1/8
There is a lag between when a case is diagnosed and when the individual may succumb to their disease and there is a further lag between date of death and when the death is reported. 2/8
Here, I compare state-level data from @COVID19Tracking for cases and deaths and find that a 22-day lag maximizes state-level correlations. 3/8
I wanted to respond to news of #COVID19 death in Santa Clara County on Feb 6. This is an interesting, if slightly puzzling, data point. 1/9
We know that there is very little genetic diversity in global samples of SARS-CoV-2, which points to emergence in ~Nov 2019 in Wuhan, China. We know that once community spread is established it ramps quickly in the absence of social distancing. 2/9
Rate of increase in confirmed cases as well as genomic data suggests a 3-4 day doubling rate. If we look at the abundantly sequenced WA cases we see evidence for an introduction in late January that leads to the majority of cases in March. Figure from bedford.io/papers/bedford…. 3/9