There Is A More Accurate Way To Compare Coronavirus Deaths And Flu

Christopher Ingraham, Washington Post

Published

Saturday, 02 May 2020 12:14 PM PDT

During the months of the coronavirus pandemic, some politicians and experts continue to promote hamhand comparisons between covid-19 and seasonal flu to gain political points.

There are many ways to uncover this fundamentally flawed comparison, but one of the most obvious was this week published by Jeremy Samuel Faust, a physician at the Brigham and Harvard Medical School Emergency Hospitals.

As Faust explains, the problem is summarized as follows: The annual influenza mortality rate published by the Centers for Disease Control and Prevention is an estimate generated by incorporating laboratory-confirmed deaths into a mathematical model that attempts to correct for underestimation. Covid-19 mortality represents the number of characters in a person diagnosed based on the virus being positive or meeting certain clinical and epidemiological criteria.

Such a comparison is a comparison of apple and orange varieties, because the former is the “bloated statistical estimate” and the latter is the “real number.”

To make a more accurate comparison, you should start with the number of directly confirmed influenza deaths that the CDC tracks each year. The last seven influenza seasons, dating back to 2013, varied from 3,448 to 15,620 deaths.

Please note that these numbers differ significantly from the CDC’s final official influenza mortality estimates. For example, between 2018 and 2019, 7,172 confirmed flu deaths are finally estimated at 26,339 to 52,664 deaths. Again, the CDC plugs the confirmed mortality into a model that seeks to reconcile what many epidemiologists believe is a serious underestimation.

Now let’s add a bar for this season’s covid-19 deaths. At the time of writing it was 63,259, and even higher when I read it. Note the significant change in the y-axis to accommodate the covid-19 mortality scale.

This year’s data is not always complete, as the flu season remains 22 weeks. No more deaths from the flu since the worst of the season has passed. However, the death rate for covid-19 has peaked at around 2,000 people per day. No one knows where to go next.

Using the apple-to-apple comparison, it can be said that the coronavirus has already killed eight times as many people as the flu. By the time we get the data for the whole season, the difference could be at least 10 times, or even full magnitude.

Faust writes that the coronavirus is “not like the flu, but much worse.”

One of the most difficult things about this pandemic is understanding the serious uncertainties surrounding the seemingly solid amount. On the surface, the comparison of influenza and coronavirus deaths seems like a simple proposition. Examine the numbers in both formulas to see which is larger.

However, the challenge is complicated as soon as influenza mortality is reported as an estimated range rather than an aggregate, and is significantly different from the individual counts based on the tests and diagnoses used in covid-19. . And those covid-19 deaths are probably underestimated as we can’t test and diagnose anyone. Soon things that used to look like simple mathematical exercises turn into confusion of algorithms, estimates, and uncertainties.

Many of us are in the middle of this pandemic, so those who first encounter that uncertainty can react in one of two ways. Some people choose a single number that corresponds to prejudice, creating artificial certainty so as not to gain political points or upset their preconceptions. That is, the heads talking to politicians use incomplete flu data to downplay the outbreak. The other person raises his hand to proclaim the truth, and ironically believes that he can “make statistics say whatever you want.”

But instead of trying to understand this uncertainty yourself, you have a third option. That is to look to the experts who have spent their entire career asking these questions. People like epidemiologists and doctors, Faust and his colleagues are trained to draw the best possible conclusions from uncertain data and need to update those conclusions as new information comes in. I understand what may be.

And while not all experts may agree in some respects, if you listen closely enough, you’ll find something like an important consensus. You can then use that consensus to inform policies that help save lives and protect the economy.