General update on COVID.


The Wed and Thu evening updates should be our first post-holiday normal read. tl;dr - it looks like we're still trending up a bit nationally.

Case counts appear to be up a bit from mid December, and with testing being down, the actual rate of infections is rising a bit faster than the red graph shows. Hospitalizations (this is total in hospitals, not new admissions) shows steady growth. Death counts are very high the last few days, and our current case counts suggest the weekly average will climb modestly through the rest of January.


Someone infected on Christmas Day would, on average, get symptoms around New Year's Eve, and get tested, get results and get logged by about January 4, give or take a few days. So we've probably seen most of any effect from Christmas, but anything related to New Years hasn't hit the charts yet. I'm a bit of a skeptic on a Christmas surge, as increased family contacts are offset by time away from work and school. Who knows....

UK: Remember that they're a couple months or so ahead of us on the B.1.1.7 curve. Case counts have roughly doubled in the last two weeks; effects of the new national lockdown won't move the charts until around inauguration day. But there was some good news from the UK on the treatment front, see below.

USA AND B.1.1.7:

The new variant has been found in more states, but is still at a very low but not precisely known level. IT IS NOT THE DRIVER FOR OUR RECENT INCREASES. The new variant only moves the needle when it becomes a big enough share of new cases to change the overall Rt numbers. This starts about the time it reaches 20% share, and then a month later it has nearly replaced the previous strain and Rt goes through the roof. We're not there yet and certainly weren't there a month ago...the rise we're seeing recently is all due to insufficient distancing and prevention.


The test % positive rate is getting a lot of attention in recent weeks, and that's a good thing. But many (most?) people are getting the wrong message.

Texas is running about 20% positive. That DOES NOT mean that 20% of Texans are infected. Testing isn't a random sample like a Gallup poll, it's highly focused on the most likely cases. If you double the amount of testing, the number of cases you find might only go up by a quarter, because the really obvious targets were already being tested.

A 20% rate like we have in Texas now can exist when only 2-3% of the population is infected. But test % positive can tell us how to read our case counts. If the test % positive is low, like 0.5% or 1%, that means most of your testing is done in contact tracing on asymptomatic, in on-demand screenings, etc., and you're probably catching more than half of your total cases. If it's high, like 20%, it means testing isn't widely available on demand, it's mostly testing only symptomatic people, and you true rate may be 3 or 5 times higher.

Use test % positive as a guide for how realistic the case curve is, not as an indicator of how many people are currently infected.


UK researchers have found that using tocilizumab and sarilumab at the time of ICU admission lowers the death rate by 24%. That's huge, and this treatment is now approved in the UK.

But that's not the biggest news in my opinion. This treatment also reduces the average time spent in the ICU by 7 to 10 days! That's a HUGE impact on ICU availability and strain on healthcare workers. At a time when they need every break they can get, this is big news.

And by the way, I have to salute the UK's health service for how they're sciencing the hell out of COVID treatment. They've had multiple large scale RCTs going in parallel, getting solid answers to the important questions. Their genome surveillance investment spotted the new variant weeks before it took over the country; our surveillance would tell us shortly after it happened. Their rapid risk analysis of single-dosing vaccines was thorough and correct. (Delaying second doses is the only rational choice given the UK's situation with B.1.1.7; the decision in the US is much less clear.) The UK's political leadership has failed repeatedly, and the UK citizenry is as much a mixed bag as in the US, but their research and coordination are inspiring.

Comments 0