As we have mentioned before in this blog, the notion of all-cause excess deaths is a comprehensive way to track the mortality impact of COVID. Excess deaths are the difference between deaths reported in a week and the expected deaths based on prior history for a region. In many geographic areas, the patterns are suggestive that excess deaths are due to the direct effects of COVID or the indirect effects of COVID (e.g. impact of lockdowns and unemployment). It also takes into account the seasonal effects of death rates. It is a lagging indicator in that it takes a few weeks for the data to be accurate. Of course, the excess deaths may not be due to COVID either directly or indirectly as there could be increases in deaths from other causes that are not immediately discernible.
The CDC has started to make its’ data on weekly deaths available to the public. The visual graphs give a more intuitive sense of the facts.
The picture for the entire United States for the last 3.5 years looks like this:
- The red pluses indicate when there were higher deaths than expected and the yellow line shows the threshold for excess deaths.
- In 2018 there were 6 consecutive weeks where were excess weekly death rates in the 4-13% range. This was likely due to influenza.
- There were no weeks with excess deaths in 2019
- CDC estimates there has been a total of 111k to 153k excess deaths for all of 2020 to date which is a 7.1 to 9.7% increase above expected mortality.
- In 2020, excess deaths were first seen the week ending March 28th which showed an excess death range between 6.4-10.2 percent. The data is reported in ranges which represent 95 percent confidence intervals.
- There have been excess deaths in the US on a weekly basis since that time. The peak for the country was the week ending April 11th when there were excess deaths of 36.7 to 41.6 percent.
- Since June 13th, the range has been between 4 and 8 percent except for the week of July 4th which was 6 to 10 percent.
Florida data shows the following:
- There was six consecutive weeks in 2018 where there were excess deaths between 4 and 16 percent, this was likely from influenza.
- Since the week ending April 11 there has been weekly excess deaths.
- From April 11 through June 13, the range has been oscillated between 2.2 to 9.3 percent.
- In the last six weeks, the range has significantly increased with a peak on the week of July 4th where the range was 15.0 to 21.3 percent.
- The excess weighted deaths that the CDC estimates are higher than the state of Florida is reporting. For example, the week of July 4th through July 11th the reported deaths from Florida Department of Health were 498; the CDC’s weighting estimates were 690.
- Total excess deaths for the state of Florida are 2.6 to 5.9% of total deaths for the year using CDC data. This is consistent with our prior estimates of one month ago which ranged from 1.8 to 6%
New York City shows how much worse its’ experience was compared to any other area of the world.
New York City shows a peak on the week of April 11th of excess deaths of 610 to 664 percent, much higher than anywhere else in the country.
Other states in the North East and Mid-Atlantic such as Maryland show similar timing of their peak excess deaths although the rates are nowhere near as high as New York City. In Maryland the peak was the week ending on May 2nd with an excess death rate of 49.9 to 61.4 percent.
California pattern looks similar to Florida. Its peak week was July 4th with an excess death rate of 11.8 to 18.8 percent:
The data shows how different the experience of COVID has been in different geographies in the country.
We will be watching it closely. Once we see the excess death rates go below the yellow line in a state or the entire country, that will be a good indicator that the worst of the first wave of infection is over.
Will there be a second wave? That will depend on immunity, which we will discuss in an upcoming post.