“Excess Deaths” From COVID-19

Several weeks ago, I began tracking total U.S. fatalities last year compared to this year and noted in this blog that there had not been a large increase in the total number of fatalities for the year.  Since then the CDC has begun a similar analysis that it calls “excess deaths.  The CDC started this website which shows how the fatalities in 2020 from February 1-May 2 compare to the average for 2017-19 for the same period.

I believe this is the only objective way to determine the true severity of the COVID outbreak for two reasons.  First, over 80% of the reported fatalities have comorbidities.  As result, it is difficult to determine whether COVID was the actual cause of death.  Second, there have unquestionably been many fatalities where the person may have died from COVID but undiagnosed, especially in the early going.  Looking at how total fatalities compare to previous periods negates both of these factors.

Let me start with a caveat about this data.  There is a significant delay in reporting fatalities.¹  The CDC has estimated that there have been about 30,000 more COVID fatalities since they put this table together.  Some of those will be for the days after May 2, but some will be “backdated” to prior to May 2.  As a result, the CDC’s calculation of the number of the recent weeks’ excess deaths are very likely understated.

The CDC breaks out excess deaths in a table for each state and New York City.  That table is cumulative for the entire period and, therefore, includes weeks in late April and early May for which the data is not complete, as I outlined above.  While it is interesting for the purpose of comparing the severity of the disease between jurisdictions, the ultimate number of excess deaths for each state will rise.

With those caveats in mind, here are the key takeaways from this data so far:

• For the middle part of April, it appears that people were dying in the U.S. at a rate about 30% higher than during that time period in previous years.  The previous high was during the 2018 flu season when fatalities ran about 10% above normal.

• New York City has accounted for almost all of the excess deaths.  For the period, it has recorded 230% of the expected fatalities.  That is both a shocking and frightening level.  It would be disastrous if other cities in the U.S. were to reach that kind of level.  Fortunately, so far, no other area is even close.
• There are only twelve states that are over 100% so far.  Texas is at 94%.
• The most recent data we have that appears to be mostly complete is the week ending April 18 (Week 16).  Total U.S. fatalities through that week, according to the most recent data, are now running about 2.7% higher than last year and about equal with 2018.

I think it is very important that we continue to monitor this data, especially as we begin to try to return to some normalcy.  It is, at least in my mind, the only true objective measure of the severity of the impact of COVID on our country.

¹ The NCHS reports weekly fatalities as part of its influence surveillance website and it shows the “percentage complete.”  However, since I have been tracking that data there have been significant revisions even after NCHS has reported the week was 100% complete.