As an epidemiologist, watching the COVID-19 pandemic unfold has been a bit surreal. Terms like basic reproduction number, or R0, have gone from fairly finicky things only spoken about in the musty hallways of universities to common parlance across the globe. A paper on antibody testing, instead of being largely ignored by people with better things to do, hits the news worldwide and makes its entire author team famous. And amid all this noise, there is a constant, exhausting flow of nonsense and outright disinformation.
It feels, to a great extent, as if the world has gone mad.
The latest front in this seemingly endless war on claptrap is something that has been bouncing around for a long time: that COVID-19 deaths are not actually COVID-19 deaths. The basic idea is that a lot of people who supposedly die from COVID-19 actually had other conditions like heart disease or diabetes, and so you can’t blame the coronavirus when it’s really the underlying issues that caused them to die.
Now, to anyone who has studied how deaths are recorded and what we mean when we say a “COVID-19 death”, this is pretty ludicrous. Countries go to great lengths to properly record deaths, because death information is vital not only to funding decisions but also things like insurance, court cases, and the like.
Knowing this, it came as quite a surprise that apparently the CDC has come out with a huge announcement — only 6% of COVID-19 deaths in the United States were caused by the virus itself!
This would be big news if true. If only 6% of people who died of COVID-19 actually died from the infection, with the others dying from other things while they had coronavirus, it would reduce the death toll of the dread disease substantially. Sadly, this is complete and utter nonsense.
Let’s look at why that is.
This claim first surfaced recently in a tweet by a conspiracy theory account that was retweeted by Donald Trump before being deleted by twitter. If that sounds crazy to you, read the entire Forbes article, because it’s a pretty wild ride start to finish.
But the basic idea has now spread across social media, and even the mainstream news, so let’s look at what actually happened. The CDC has released a breakdown of COVID-19 deaths in the U.S. One of the tables that you can access shows COVID-19 deaths broken down by what the website describes as “comorbidities”. Above the table, there’s this short blurb:
Now, on first reading you might be forgiven for being confused. It says right there that only 6% of the deaths recorded by the CDC as involving coronavirus had COVID-19 as the only cause. But if you look at the table below, it becomes very clear that this doesn’t mean that these deaths aren’t coronavirus deaths. For example, let’s look at the first section — Respiratory diseases:
The first line of these so-called “comorbidities” is influenza and pneumonia. Now, it might not be surprising that people who died of COVID-19 had pneumonia, but that’s what you’ve got in the table. Similarly, respiratory failure, something that the coronavirus directly causes, is listed here as a “comorbidity” that 55,000 people had. The single biggest “comorbidity” in that table is the section labelled “other”, which includes everything from neonatal jaundice (P59.9) to chest pain, unspecified (R07.4), making it a bit useless as a category.
So what’s happening here? Well, it’s pretty simple — in the U.S., deaths are recorded using standardized death certificates. On these certificates, completed by medical certifiers, there are several spaces to fill in — one for the immediate cause of death, and then several lines for the underlying causes of that. As an example, say someone has lung cancer, and dies in hospital of an infection after having a lung removed. The immediate cause of death is the infection, which occurred due to complications of the lung removal, which was ultimately caused by the underlying issue of lung cancer. In the same way, someone who gets COVID-19, which causes respiratory failure, and then dies of kidney failure due to being on a ventilator would have at least three things on their form — the immediate cause, kidney failure, the secondary cause, respiratory failure, and the underlying cause, COVID-19.
When you see that “only 6%” of people had COVID-19 as the sole reason listed on their death forms, what it means is that there were only a small fraction of people who died of the disease who didn’t have any other underlying or immediate causes noted by the medical certifiers. This is completely unsurprising, as it’s pretty rare that someone wouldn’t have at least one issue caused by coronavirus prior to their death, and all it means is that in 94% of cases people who had COVID-19 also developed other issues, or had other problems at the same time.
One way of looking at the precise number is to ask how many COVID-19 deaths had coronavirus as the UNDERLYING cause. That is, the cause that precipitated any other issues, or the thing that actually killed a person. The CDC has actually estimated this, and puts it at >95% of all COVID-19 deaths, meaning that the vast majority of deaths recorded as caused by coronavirus in the U.S. were caused by COVID-19.
In other words, most COVID-19 deaths were almost certainly caused by COVID-19.
Hardly surprising, but still a depressing fact.
COVID-19 Deaths Worldwide
One important thing to note is that a lot of this is specific to the U.S. There are, oddly enough, quite a few other countries around the world, and they all have their own way of recording deaths. In some places, for example India, there has been a lot of criticism that the death reporting is shockingly bad and thus the country may be missing some or even the majority of their COVID-19 deaths. In other places, like Belgium, the death reporting is so good that it may explain the high COVID-19 death rates — they are simply picking up coronavirus deaths that other places have missed.
However, one thing remains true: most countries go to great lengths to ensure that deaths are correctly classified. Death reporting is incredibly important, and in most places it’s a detailed process that has to be checked carefully. In most cases, we can say with some certainty that deaths attributed to COVID-19 are, at best, a solid count, and at worst probably an underestimate. If anything, it’s likely that we are missing quite a few deaths that have been caused by coronavirus, but for whatever reason not picked up in our reporting systems, and thus the death count is actually higher than the reported figure.
And I say all of this as someone who has spent quite a bit of time studying COVID-19 death rates. I’ve now co-authored two studies looking at the fatality rates of COVID-19, and can say with some certainty that they are pretty good estimates, if probably a little low.
Unfortunately, while it may sound like a great thing, it’s total nonsense to suggest that most deaths attributed to COVID-19 were caused by something else, or that the people would’ve died anyway. From studies of excess fatalities we can say with a lot of certainty that the people who were killed by the coronavirus would be with us today if not for the infection.
As of today, the global toll from COVID-19 is 846,000.
Those deaths were, depressingly, caused by the disease itself, no matter what viral tweets may say.