Over the last two weeks (ending May 8), the COVID metrics for the US continued to slowly decline.

Hospitalizations

The CDC hospitalization stats are becoming increasingly difficult to analyze because it continually revises the total for previous days.  These revisions sometimes reach back for weeks.  If you look at the CDC chart for any given day (like the one below), it is showing a large weekly decrease in hospitalizations.  But after the revisions come in the declines are considerably more modest.  We especially cannot rely on the data from the last week, which makes it difficult to know what is currently happening with hospitalizations.  My best guess, based on the way the data has been coming in, is that total hospitalizations and new hospitalizations are declining at about 5% per week right now, despite the CDC website consistently  showing  larger declines.

Fatalities

The decline in the seven-day average for fatalities was once again modest the last two week (668→662; -5%).  There was a small increase in the first part of the two-week period.  I cannot tell if that was an actual increase or some kind of catch-up revision.

Total reported fatalites related to COVID last week were 4,410.  That is the lowest weekly total since last July.  In years before COVID, about 54,000 Americans died on a weekly basis.  So, COVID related fatalities are now probably running at about 8% of “normal” fatalities.  The CDC excess death analysis does not show their have been more deaths than expected since the second week in March.  There may be enough late filings of death certificates to push it over but at this point we should have 98%+ of the death certificates in.

I have been using the IHME model as a reference.  However, IHME announced this week that it has revised its fatality projection based on not the reported number of fatalities, but their estimate of how many people have actually died from COVID.  Their new analysis increases the number of deaths by about 80%, which is absurd.  According to the new analysis more than 100% of the excess deaths in 2020 were due to COVID.  That would mean that without COVID there would have been fewer fatalities in 2020 than 2019.  The NCHS has not reported a year-over-year decline in annual fatalities in the decade of stats they publish online.  IHME has also increased its projections for fatalities by 137,000 by the end of August based on increased mobility data from cell phones!

The results of the IHME model had been coming in increasingly optimistic, which was at odds with doom and gloom predictions of its director, Christopher Murray.  Murray has emerged as a highly partisan, pro-lockdown advocate, who recently declared that wearing masks was the most important factor in reducing infections, presumably even greater than vaccines, a view not supported by any scientific analysis.  It appears to me that IHME has begun to manage its model parameters to conform it is director’s apocalyptic views.

Vaccines

The US passed a noteworthy vaccine milestone this week.  The number of vaccine doses delivered to the states and other agencies by the federal government now exceeds the population of the county.  While the number of doses administered per day has definitely decreased since the peak about a month ago, the seven-day average is still running over 2 million.

The percentage of Americans fully vaccinated continues to  steadily grow, now standing at 34.4%.  Some analyses of Israel’s experience, which is running ahead of ours, suggests something about 50% may be a significant tipping point.

The good news is that we have really done a good job getting seniors vaccinated.  Over 71% of those over 65 are now fully vaccinated.  That percentage has doubled in less than two months, a truly remarkable achievement.  Just under 84% of seniors have received at least one shot.  The increase in that metric, however, has really slowed, potentially indicating that we are approaching the maximum that are willing to be vaccinated.  That would be pretty consistent with roughly 15% of the country who are hardcore anti-vaxxers.  However, based on the there-are-no-atheists-in-foxholes theory, I suspect the number of anti-vaxxers among seniors may end up being somewhat lower.

I am surprised we have not seen more of an impact on fatalities from the success in vaccinating seniors.  It may just be more delayed than I have been expecting.  There is a lag time in the vaccines reaching full effectiveness and, of course, fatalities are very much a lagging indication.  Hopefully, we will see that kick in to a great degree soon.

Testing

The number of daily tests in the US dropped dramatically over the last two weeks (1,491,090→932,630; -37%), which I have been expecting.  The seven-day average for the CDC’s daily “new case” count dropped by 30%, falling below 40,000 for the first time since September.  The calculated positivity rate ticked back up to 5.7%, probably as a result of the decline in tests.

Summary

I think the media has never done a greater disservice to the public than its reporting on the concept of herd immunity.  It originally consistently described the phenomenon as some percentage that, once reached, COVID would sudden disappear.  Now it is spending its time discrediting the straw man distortion of the concept it created.

Herd immunity is not a point in time.  It is a process by which our collective immune systems drive infectious diseases from an epidemic to an endemic state.  It has happened with every epidemic that has ever occurred, and it was always inevitably going to happen with COVID – even if there had never been a vaccine.

It is miraculous that our vaccine technology has advanced to the point that we were able to develop a vaccine prior to the end of the epidemic phase of a disease, something that has never been done before.  That will significantly reduce the fatalities and serious disease that COVID would have otherwise caused.  But make no mistake, the COVID epidemic was always going to end just as every other epidemic in history has ended.  As William Farr discovered 180 years ago, there are no perpetual epidemics.

As with the Texas statistics, because the course of COIVD appears to have stabilized, I will not report again on the US numbers until the end of May unless there is some significant change in the trajectory.  I will continue to post data periodically on Twitter, Facebook and LinkedIn.