Pushing back against Long COVID misinformation

The CDC and Census’s Household Pulse Survey is one source showing the broad impacts of Long COVID.

It’s an unfortunate reality in the Long COVID media landscape that a lot of journalists and commentators write about this condition without really doing their research. I frequently see articles that misunderstand key aspects of Long COVID or dismiss patients’ experiences.

Two recent stories (one in the Washington Post and one in Slate) make these mistakes, in fairly high-profile outlets. I’m not going to link to the full pieces here to avoid giving them more attention, but I wanted to share a bit about what these stories get wrong, to help readers recognize similar issues in other pieces.

The Washington Post piece discusses results from a data analysis project that the news outlet did in collaboration with Epic, an electronic health records company that has access to anonymized data from millions of people. Researchers evaluated whether patients had sought medical care for common Long COVID symptoms following a positive COVID-19 test. According to this analysis, Long COVID symptoms have become less common during the Omicron era compared to earlier COVID-19 waves—which may sound like a promising conclusion, until you recognize the analysis’ flaws.

Epic and the Washington Post didn’t share the full data behind this analysis, which makes it difficult to fully evaluate. But the public methodology provides enough information to be critical. First of all, relying on electronic health records for Long COVID leaves out a lot of people. Many of the people most vulnerable to this condition have limited access to PCR tests and to the doctors who might help them diagnose new symptoms, and PCR tests in particular have only gotten harder to access since early 2022.

Second, this analysis focused on people who had new symptoms after COVID-19—and excluded people who had existing symptoms that might overlap with Long COVID before they got hit by COVID-19. This focus leaves out people with preexisting conditions (i.e. other chronic diseases), who tend to be at higher risk for Long COVID. Epidemiologist Deepti Gurdasani has a helpful Twitter thread explaining this issue in detail:

Meanwhile, the Slate piece dismisses Long COVID, arguing that the condition is “neither as common nor as severe” as experts have warned it might become. Like the Washington Post piece, this article bases its conclusion on flawed research that focuses on biased health records, and other types of biased data. For example, the author argues that people aren’t getting social security disability benefits in large numbers; but Long COVID patients face many barriers to this program, as I’ve covered for KHN.

The Slate article fails to cite Long COVID patients or experts who have cared for those patients. It also fails to include extensive research showing that Long COVID can mean lasting damage to many organ systems. In fact, Slate issued a correction to the piece shortly after it was published, explaining that, yes, research has “identified biological changes” associated with Long COVID. (The author initially wrote that there were “no biomarkers” for Long COVID.)

I’ve signed onto a letter demanding that Slate retract this article, along with hundreds of other journalists, researchers, and clinicians with Long COVID expertise. You can read it and consider signing here.

These two articles fall into a similar trap: they cherry-pick evidence to suggest that Long COVID might not be as common as some experts and patient-advocates say it is. But they ignore substantial evidence showing how widespread and how damaging the condition actually is. And furthermore, even if Long COVID is getting “milder” or “rarer” with Omicron, so many people have been infected by it (and by prior variants) that there are still millions out there who need help. Arguing over prevalence distracts from the true story: people are suffering, and they need support.

It’s important to note that the majority of journalists covering Long COVID are doing so in good faith, simply trying to understand a complex and confusing condition. But it can be easy to make mistakes (like citing the wrong evidence, or failing to talk to a person with Long COVID). Plus, some bad actors have shown up in the media again and again; the author of the Slate article, for example, previously wrote a highly discredited story for New York Magazine.

For other journalists covering this topic, I’m always happy to help answer questions or point folks to resources. The tipsheet I compiled for NICAR might be a good place to start. And for editors assigning these stories: please make sure you’re giving assignments to a credible writer.

More Long COVID data

COVID source shout-out: Moving closer to Long COVID biomarkers
Scientists are moving closer to biomarkers, or clear biological indicators, of Long COVID. A new study—posted this week in Nature ahead of full publication—identifies clear differences between blood samples of people who have the condition and those who don’t.
Sources and updates, September 10
Sources and updates for the week of September 10 include monoclonal antibody costs, viral persistence in Long COVID, and Medicaid unwinding.
Sources and updates, September 3
Sources and updates for the week of September 3 include a new CDC updates page, Long COVID research, and people who are more vulnerable to severe COVID-19.
Sources and updates, August 27
Sources and updates for the week of August 27 include funding from Project Next Gen, wastewater testing for more viruses, health misinformation, and more.
The NIH says it “inappropriately” censored Long COVID patients on social media
The National Institute of Health (NIH) is under fire for censoring comments from patients on social media — the latest in a trend of heavy criticism from people living with Long COVID for failing to listen to patients and implement …
Sources and updates, August 13
Sources and updates for the week of August 13 include Long COVID rates, vaccination benefits, and a wastewater surveillance webinar.

Leave a Reply