For the last two weeks (ending October 10), the US COVID metrics have mostly moved sideways.  There is, however, a hint that we may be seeing some increase in the spread of the virus, probably related to schools reopening and the relaxation of some suppression measures.

Hospitalizations

The number of COVID patients hospitalized in the US increased during the last two weeks by 12%, wiping the gains from the previous two weeks.  Patients in ICU units and on ventilators increased by smaller amounts, 3.6% and 2.0%, respectively.  The levels are running about the same as six weeks ago and about 60% of the peak in July.  With the possible exception of Wisconsin, there are no areas were hospital capacity is overtaxed.  As I have said before, I think hospitalizations are the most reliable metric so this will bear watching carefully over the next few days.

Fatalities

Average daily fatalities continued to drift lower, averaging 683 versus 770 for the previous two weeks.  The 4-week lagging case fatality rate is still under 2% but did not show any additional improvement.

In August, I wrote this post about the declining lethality of COVID.  The evidence for this trend has become so ubiquitous that it is now being picked up in mainstream media, like this story in Washington Post.  In addition, it appears that one or more of the new therapeutics will soon get emergency use authorization from the FDA and be more widely available in the coming weeks.  This should further drive down the fatality rate.

The  IHME model lowered its fatality estimate for year-end from 360,000 to 347,000.  Still, that would mean that the daily fatalities would have to more than double for every day between now and the end of the year.  That is become increasingly less believable and I suspect we will continue to see them lowering their estimate.

Testing/New Cases/Positivity

Testing increased for the two-week period by 14%, rising from 6.28 million to 6.8 million.  New cases also increased, but at a slower pace than the increase in testing – 391,000 to 435,000 (7%).  As a result, the positivity rate fell again and was under 5% for the last two weeks for the first time since the pandemic began.

As I have said many times here, new cases and positivity rates are the worst metrics we have.  The reporting from the states is all over the map and more researchers are raising how issues like false positives in PCR tests and the change the clinical presentation of those being tested may be distorting  epidemiological analyses based on test results.  The COVID Tracking Project recently published this excellent blog post,  describing the problems and limitations of new case/test results analysis.

There were many breathless accounts in the media and social media of “spikes” and “surges” in new cases like this notice I received on my smart phone.

But almost none mentioned that the increase on testing resulting a lower positivity rate.  Every time I think that media cannot sink any lower in its sensationalization of this pandemic I am proved wrong.  Please take any headline you see about how bad things are with a huge pinch of salt.

Summary

Ten days into October, we are still not seeing any major indications of the much-dreaded fall/winter surge.  However, the fact that hospitalizations have stopped declining and begun to inch back up makes me a bit nervous.  I continue to believe that we are in, and will likely to continue to be in, the tug-of-war between the relaxation of suppression measures and growing naturally acquired immunity.  As a result we will likely see the metrics ebb and flow until we get a vaccine to deliver the knockout blow to COVID.