As Faust describes it, the issue boils down to this: The annual flu mortality figures published by the Centers for Disease Control and Prevention are estimates produced by plugging laboratory-confirmed deaths into a mathematical model that attempts to correct for undercounting. Covid-19 death figures represent a literal count of people who have either tested positive for the virus or whose diagnosis was based on meeting certain clinical and epidemiological criteria.
Such a comparison is of the apples to oranges variety, Faust writes, as the former are “inflated statistical estimates” and the latter are “actual numbers.”
To get a more accurate comparison, one must start with the number of directly confirmed flu deaths, which the CDC tracks on an annual basis. In the past seven flu seasons, going back to 2013, that tally fluctuated between 3,448 and 15,620 deaths.
Now, let’s add a bar for this season’s covid-19 deaths, which as of this writing stands at 63,259, and which will be even higher by the time you read this. Note the drastic change in the y-axis to accommodate the scale of covid-19 mortality.
This year’s data are necessarily incomplete, as 22 weeks remain in the flu season. There are not likely to be many more flu deaths, as we are well past the worst of the season. But covid-19 mortality has plateaued at around 2,000 deaths per day. Where it will head next is anyone’s guess.
Using an apples-to-apples comparison, we can say that the coronavirus and the disease it causes, covid-19, have already killed eight times as many people as the flu. By the time we get data for the entire season, the difference appears likely to be at least tenfold, or a full order of magnitude.
The coronavirus, Faust writes, “is not anything like the flu: It is much, much worse.”
One of the most challenging things about this pandemic is making sense of the profound uncertainty surrounding the many quantities that might appear, at first glance, to be rock solid. On the surface, comparing flu and coronavirus deaths seems like a simple proposition: dig up the official numbers of both and see which is greater.
But that effort gets complicated as soon as you realize that flu mortality is not reported as a tally but as an estimated range, which is far different from the individual counts, based on testing and diagnoses, used for covid-19. And because we can’t test and diagnose everyone, those covid-19 deaths are probably undercounted as well. Soon, what once appeared to be a simple mathematical exercise turns into a mess of algorithms, estimates and uncertainty.
People encountering that uncertainty for the first time, as many of us are during this pandemic, are likely to react in one of two ways. Some cherry-pick a single number that comports with their biases, creating an artificial certainty to score political points or avoid upsetting their preconceptions. That’s what the politicians and talking heads using faulty flu data to downplay the outbreak are doing. Others throw their hands up and declare the truth to be unknowable, indulging in the cynicism that believes you can “make statistics say whatever you want.”
But rather than try to make sense of this uncertainty ourselves, there’s a third option: turning to the experts who’ve devoted their entire careers to these questions. We can listen to the epidemiologists and physicians, people like Faust and his colleagues, who are trained to draw the best possible conclusions out of uncertain data, understanding that those conclusions may have to be updated as new information comes in.
And while the experts might not all agree on some points, something like a critical consensus emerges if we listen to enough of them. Then, that consensus can be used to inform policy that helps save lives and protect the economy.