It’s been more than 3 years since Covid changed the world forever, and both researchers and the press still have no idea how to cover it.
In my day to day clinical work, I see a fair number of people who are suffering from PACS, or post acute Covid-19 syndrome, and I even wrote about it a few months ago:
In that newsletter, I wrote about how my experience taking care of these patients has convinced me of a few things:
Many long Covid patients are really suffering
Doctors don’t understand the biology very well
Most patients with long Covid that I’ve seen fall into 1 of 2 groups:
But unfortunately, a huge amount of what I see reported in the popular press and published in scientific journals doesn’t actually describe these patients very well.
Instead, we get stories fearmongering about Covid causing a risk of hypertension or even worse cardiovascular disease.
None of the trials linking Covid to development of medical conditions are reliable
This is really hard research to do well, but really easy research to do poorly.
If you really want to study this in a manner that we can draw reliable conclusions from, you would need to take a large group of people, randomly assign them to either get Covid or not get Covid, and then follow them for years, ensuring that the group you didn’t assign to get Covid never gets infected.
But that’s not a possible experiment to run because we don’t live in a simulation.
Since the spectrum of Covid infection ranges from death to asymptomatic infection, and people can get Covid more than once, it quickly becomes basically impossible to sort out the impact of Covid infection without hopelessly confounding your research.
Do you think a study looking at insurance claims or electronic health records is reliable?
I would argue that it’s not useful for drawing any conclusion - someone who becomes captured by a medical record or an insurance database when they have Covid is likely to be very different from someone who isn’t.
People who seek out medical care can confound your interpretation in multiple ways here:
Maybe the person who comes to medical attention is way sicker than the person who doesn’t
Maybe the person who comes to medical attention is way more neurotic than the person who doesn’t
Maybe the person who comes to medical attention has more underlying health concerns than the patient who doesn’t
And that’s only the quick list I could come up with in 30 seconds - there are a million other ways that this research is confounded.
Research this confounded isn’t just tough to interpret - it’s useless
I’m really not exaggerating here.
When you see headlines like this one, there isn’t really anything to say about it. This is bad science that shouldn’t get your attention.
I’m going to take a little bit closer look at this study, just to show you how easy it is to realize this is all nonsense.
This study was published in the journal Hypertension and publicized by organizations like the American Heart Association.
In this study, they compared patients hospitalized with Covid to patients hospitalized with the flu who did not have a previous diagnosis of high blood pressure and then followed their medical records to see who developed a record of high blood pressure.
Here’s a look at some demographic information on the two groups of patients:
That table may feel a little bit busy, so just look at the very top where the ages of patients are described… do you notice anything?
Patients with Covid were older than patients with the flu.
Risk of hypertension tracks with age.
So older patients are more likely to develop hypertension.
Why would you look at this and conclude that Covid causes high blood pressure? Wouldn’t the more normal conclusion to draw be that older people are more likely to develop high blood pressure than younger people?
Are we too gullible to discuss this science?
Most of the research that you see on long Covid doesn’t pass the smell test to think that there’s really something useful to draw there.
This article is no different than thousands of other ones that have made news.
The research is confounded in a way that makes it useless for drawing conclusions, but then it’s presented to us as though it’s “food for thought” or a “reason to exercise caution.”
Long Covid is treated by researchers and the media the same way that exercise and nutrition are treated: not worthy of studying rigorously because low quality work gets attention and money.
Scientists publish low quality research that they are incentivized to exaggerate.
Journals with editors that know the work is hopelessly confounded publish this work because it helps with their impact factor when other articles cite it with their own low quality studies.
The media discusses it because we click on it and share it.
These ideas then seep into the public discourse, leading to confusion, fear, and anxiety.
You have to keep your own information diet here - the media is not going to help you and journal editors aren’t going to filter bad ideas or protect you from researchers hopelessly exaggerating the relevance or importance of their low quality findings.
And so when you see something that is a “link” or “associated with,” it’s probably an observational study that doesn’t establish causation and isn’t worth your time to read the headline, let alone the article.