[arcticbull]

The CDC puts the average lag at 13 days [1] and the uptick began in early October. That was not reflected in death counts 14 days later. I suggest waiting and seeing, there are other reasons why the number of cases may be skyrocketing, including broader access to testing. More testing will increase the denominator without increasing the numerator.

Keep in mind that giant spike in cases in July led to a reduction in COVID deaths in August and September according to the charts.

Case counts and death counts have been materially diverging for months now, because the early results were marred by adverse selection bias. With few tests available, testing was only done at the point of admission for critically ill patients, who naturally were the most likely to die of COVID. With broader availability of tests, the full picture starts to come into focus.

It's worth paying attention of course, and not hand-waving the spike, but let's not pre-judge yet. We'll have to see what this spike means in the fullness of time.

[1] https://wwwnc.cdc.gov/eid/article/26/6/20-0320_article

[roca]

Test positivity in Sweden has been rising steadily and is now over 10%. So no, it's not "broader access to testing" --- in fact they aren't testing nearly enough and the rise in case numbers is probably even steeper than reported. https://ourworldindata.org/grapher/positive-rate-daily-smoot...

Lag effects are complicated. In many places young people are the main spreaders of the virus but don't get very sick, and it takes multiple cycles of infection for the virus to reach more vulnerable people (e.g. via care workers who are pretty careful but will inevitably get infected when they live in a COVID-saturated environment). So yes, we shall see.

[arcticbull]

Test positivity isn't really what I was referring to re: adverse selection and limited testing. I'm suggesting that yes, more people may be testing positive but they're people with lower grade symptoms who wouldn't have sought out a test before or weren't in a position to get one before.

[m4rtink]

The percentage rising could be a big indicator of things starting to go bad. Here in Czech Republic the percentage was around 4 percent for a long time and things were looking good - tens maybe a hundred cases a day, couple hundred dead. Then in late summer/early autumn it started to rise, crossed 10, then 20 and went up to more tan 30. At that time there were 15000 next cases detected daily, hospitals under strain and daily deaths above hundred.

Preatty drastic measures have been implemented (non essential shops closed, nigtly curfew, restaurants takeout only, all schools closed, mandatory masks in all public spaces) and the numbers are going down to about 2000 per day most recently.

The measures seem to have been effective, but a a cost of about 6000 more dead - and more can be expected due to the inherent delay.

So for that reason, percentage of positive tests really should be watched carefully.