Published on June 3rd, 2020 | by Zachary Shahan0
“Pneumonia Deaths” & “Flu Deaths” Jump Enormously In USA
June 3rd, 2020 by Zachary Shahan
I’m going to keep this piece short, but it seems to be an important data-focused followup to a previous story I wrote regarding an important public health matter. The earlier story was, “Overall Death Count Jumps Mysteriously In Countries Around The World — No Clear Explanation.” Yes, there’s a heavy dose of sarcasm in that piece.
We know why the overall death count is much higher than normal. There are various reasons why there’s a gap between that figure (O) and the normal average (n) plus the covid-19 (c) count (O > n + c). There could be an undercount for covid-19 because there haven’t been enough tests conducted and many people who died from covid-19 were given a different cause of death. It could be that the data collection is just delayed. Or it could be that some people/agencies are assigning the wrong cause of death in order to keep the covid-19 death count lower. In reality, it’s probably a combination of those factors, but the degree to which it is one or another is not clear at all at this point.
In at least a few states, though, it appears that the specific data confusion is that covid-19 deaths are being counted as pneumonia. This first place I ran across this was in a tweet in response to a tweet about a reddit post without a clear source. It appears, though, that these numbers come from new weekly state reports provided by the CDC (Centers for Disease Control and Prevention).
This is really important. Florida is cooking the books on COVID-19 deaths. Look at the excess death rate. In Florida the number of deaths from Covid so far are likely 3200 more than they are saying. I suspect other states are also doing this, particularly those with GOP control. https://t.co/zU8c3a6Td0
— Howard Dean (@GovHowardDean) May 28, 2020
As you can see above, based on those numbers, thousands more people are dying in some states (Florida, Texas) from pneumonia and the flu than you would expect based on historical data on this topic. That said, there’s a bit of an issue with the numbers.
This reporter and this reporter, using CDC numbers, found a much lower number for Florida pneumonia and influenza deaths — still significantly higher than previous years, but not by nearly as much as the tweet above indicates. I was confused, as were others. The problem is, the CDC is publishing different numbers in different places. The sources of the more eye-popping numbers are provided here — which were CDC numbers for 2020 and CDC numbers for earlier years. Clearly, it’s getting a little confusing now, but the point is that pneumonia and influenza deaths are up either significantly or hugely in several states. The easiest way to consider the scale of the problem is perhaps to just look at the national story.
Here’s a graph for the US as a whole showing the percentage of all deaths that supposedly came from pneumonia or influenza:
During flu season last year, it peaked at around 8% share of deaths. This year, the peak seems to be 16%.
Clearly, something is amiss. I’m not trying to delve into intent at all, though. These could have been misclassified because the person was not tested for covid-19 (recall that we’ve had a great shortage of these tests) and the symptoms pre-mortality were most similar to pneumonia or the flu. Or it could have been more nefarious. Who knows?
The good news is that there are people who are trying to get to the bottom of the data problems in order to quantify things as close to accurately as possible.
“Currently, health officials and statisticians are researching how many of the states’ ‘excess deaths’ over the last several months should be attributed to the coronavirus,” Erin Banco of the Daily Beast reports. “One study by the New York City Department of Health and Mental Hygiene published earlier this month said that there were thousands of ‘excess deaths’ in the city from March 11 to May 2. About 18,879 of those deaths were explicitly tied to the coronavirus. But the study said there were also an additional 5,200 deaths that were not identified as either laboratory-confirmed or probable COVID-19-associated cases, but could have been tied to the virus in some other way.”
Also, from the CDC webpage where I got the graph above, they noted: “Weekly mortality surveillance data include a combination of machine coded and manually coded causes of death collected from death certificates. Percentages of deaths due to pneumonia and influenza (P&I) are higher among manually coded records than more rapidly available machine coded records. Due to the additional time needed for manual coding, the initially reported P&I percentages may increase as more data are received and processed.”
The general take-home point: Official covid-19 death counts are really high (107,000 and counting in USA), but the actual numbers are probably much higher than the official counts.