There was a short period, a couple of weeks ago, when it felt like every day there was a new “panic headline.” My Instagram DMs and email inbox filled with “Is this true?!” and “Talk me off the ledge, please!”
I often respond to these in real time on Instagram. A 60-second story can be enough to say “Move on with your day!” but often not enough to really explain why you should disregard these headlines or get into the process of what I look for in evaluating them. This “explaining why” is the core of ParentData, so I’m going to unpack recent panic headlines in a bit more detail here.
The goal is twofold: first, if you’re still worried about this particular panic headline, you can dial that down. And second, to give you some questions to ask yourself before you panic with future headlines.
What’s the claim? A new paper, released in the journal Obstetrics & Gynecology, argues that 1 in 10 pregnant women who get COVID during pregnancy develop long COVID. Long COVID here is defined as persistent symptoms an average of 10 months after infection.
Most important issue: Lack of baseline control group.
Details: Researchers in this paper recruited about 1,600 women who had COVID during pregnancy and surveyed them an average of 10 months after infection about a series of symptoms. These include fatigue, hair loss, thirst, gastrointestinal symptoms, and others. Long COVID isn’t one specific thing or even one specific symptom profile. Individuals are coded as having long COVID if they have a sufficient number of these symptoms. The researchers find that about 10% of the sample have a sufficient symptom level.
The problem is, a lot of people have at least some of these symptoms. If you surveyed a general slice of the population at any one time, you’d find people suffering from fatigue, nausea, hair loss, etc. This is perhaps especially true of new parents. Fatigue and hair loss are among the most common complaints I hear from new moms. In order to evaluate the impacts of COVID, we need — at a minimum — to compare the 10% of people with symptoms in the COVID group to the level you’d see in a comparable population who did not have COVID.
The authors do not do that, and, as a result, we really cannot conclude anything here about the prevalence of long COVID in this population (other than that 10% is an upper bound).
Big picture: It’s always a good idea to ask about baseline risk when evaluating claims about the impact of some treatment.
Community Guidelines
Log in