On Sunday, March 19 2023, the Washington Post published an in-depth, front page article with the headline: “Long-covid symptoms are less common now than earlier in the pandemic.”

The many significant problems of the story fall within two categories: (1) The use of limited research to draw misleading claims, and (2) the choice to emphasize analyses and perspectives that minimize the scale and impact of long COVID while burying caveats and criticisms. The end result is that the article created a dangerous misrepresentation of the prevalence and severity of Long COVID, downplaying its existing risk to society – even while including compelling stories of the profound impact it has on individuals:

1. The featured research does not represent the full extent of Long COVID:

The research in the article is based on an analysis of electronic medical records (EMR), developed in collaboration with the Washington Post, Epic Systems and Kaiser Family Foundation, a nonprofit health-policy organization.

EMR-based research can be useful but it is well known to have a dangerous weakness: it only includes data on who is in medical care. Research has consistently shown that many people with long COVID are not in care, and struggle to find competent and affordable medical providers. In addition, providers report being underprepared to diagnose and treat it. As a result, patients either drop out of medical care or never receive care for Long COVID and associated symptoms or diseases.

As reported in Public Herald by journalists, researchers, physicians, and others who cover Long Covid and other infection associated illnesses, “People with Long Covid have been denied medical care based on the assumption that their illnesses are psychological in origin — an issue which appears to disproportionately impact women; Black, Latinx, Indigenous, and people of color; LGBTQ+ patients; and other already marginalized groups.”

The well-documented barriers to health care access experienced by those marginalized groups limits the capacity of EMR to estimate rates of diseases across populations. Yet the story features a chart and heading asserting that Long COVID symptoms are “consistent across race and ethnicity.”

This graphic is much more visible to readers than the powerful quote in the story from Alba Azola, co-director of the Johns Hopkins Post-Acute COVID-19 Team in Baltimore. Azola explains that “Health equity plays a huge role regarding the type of patients able to seek care for long COVID… the people who make it to our … clinic are rich, white and able to access medical care.” (emphasis added)

As an under-served and poorly-understood condition, Long COVID is being undercounted. For example, multiple sources within the Long COVID community have reported that their visits to emergency rooms were not charted as “Long COVID” and did not document the actual symptoms presented, but instead were recorded as “anxiety.”

As we will address in our longer analysis, the research decision to exclude those who had similar symptoms prior to Long COVID further weakens the study’s power to fully calculate prevalence and severity.

These and other critical misrepresentations within EMR can end up excluding Long COVID symptoms and allow for major flaws within the type of research conducted by Washington Post.

2. Editorial decisions minimize the scale and impact of Long COVID:

The structure and emphasis of the article minimizes the scale and impact of Long COVID, from the headline and throughout the content. For example, the longform article buries this shocking figure:
In a country where at least 200 million people have been infected with SARS-CoV-2, according to federal estimates, the rate detected by The Post translates into about 14 million U.S. residents who survived the virus and are struggling with long-lasting effects that often alter their lives. (emphasis added)

The reader travels 1000 words into the article before this shocking number appears. It is likely an undercount, given the research methodology problems – but even this number reveals a much larger crisis than the existing headline suggests.

While the article includes caveats and quotes that question some of the minimizing effects, they are not nearly as prominent as the headline, the headings of charts with misleading research findings, and what’s emphasized in the text itself.

We, the undersigned, demand that the Washington Post:

-Publish a series of news and opinion stories with a more comprehensive and accurate representation of Long COVID

-Release the data and records used for the analysis in this report for public review and peer review purposes

-Demonstrate a commitment to engage with patient-led research and community-accountable, participatory research that centers those with lived experience

-Conduct a transparent analysis of the partnership priorities and editorial processes that led to this article, including:
-Lack of recognition of the limitations of EMR-based research
-Insufficient analysis of widespread barriers and disincentives to the accurate diagnosis, care and treatment of Long COVID and associated diseases affects EMR research
-Exclusion of disabled and chronically-ill editors, reporters, and researchers with lived experience.

Long COVID Justice, in partnership with:
Long COVID Action Project (LCAP)
MEAction
Patient-Led Research Collaborative (PLRC)
Strategies for High Impact (S4HI)
On behalf of additional individual and organizational signatories