Request for NIH to Retract RECOVER Study Regarding 12 Symptom PASC Score For Long Covid

© Long Covid Action Project

Dear National Institutes of Health & RECOVER Program Researchers,

I am writing this first* letter in solidarity with the Long Covid Action Project (LCAP) to call for a retraction of the recent study “Development of a Definition of Postacute Sequelae of SARS-CoV-2 Infection” published in The Journal of the American Medical Association. The study, along with the National Institutes of Health (NIH) press release titled "Large Study Provides Scientists Deeper Insight into Long Covid Symptoms", raise significant concerns for people living with Long Covid.

Specifically, various aspects of the study are misleading and could complicate Long Covid access to care.

Despite the best efforts of NIH and its significant contributions to Long Covid (a.k.a. PASC) literature, the current effort falls short and must be addressed.

LCAP has identified several reasons why the most recent NIH RECOVER paper should be corrected and considered for retraction. They include:

  1. The 12 point symptom-based scorecard in the study has high potential for misinterpretation by the medical community. It can be assumed that it has been validated through peer-review for clinical diagnosis of PASC. In fact, the RECOVER paper claims to have developed “a PASC scoring algorithm that provides a framework for diagnosing PASC.” To date, the Long Covid scoring system in the real-world population is untested. Without tested evidence of the scoring system's efficacy to accurately define and characterize Long Covid, it should not have been released to the public and clinicians as a scorecard for measuring positive or indeterminate cases of Long Covid. LCAP already has examples of clinicians misinterpreting the paper to use the 12 point cutoff score to diagnose and sub classify patients, as seen in the following example:

    1. Additional concerns from people with Long Covid have been raised that this could be misused by insurance companies to approve or deny coverage.

  2. The media has misinterpreted the study’s findings, largely as a result of how NIH wrote the press release. NIH needs to issue a revised press release with statements to the press that include considerations made in this letter by people with Long Covid.

    1. I have concerns about how the study was sold to the media. The media has reported the study widely as containing the 12 Key Symptoms of Long Covid, but no one from the NIH, that I’m aware of, has made an effort to correct this misinformation. The media has also published the points system contained in the report without properly explaining them within the context of all studies on Long Covid. To date, there are thousands of studies about this disease but only a specific few that were given attention to draft the RECOVER paper scoring system. This misrepresentation in the media has alarmed the Long Covid community. The media reporting as a result of the NIH press release has lent itself to the current trend of misrepresenting Long Covid.

    2. I would ask that NIH RECOVER release a specific statement clearly explaining that their assessment was intentionally designed to prioritize Long Covid symptoms that are the least overlapping with other illnesses rather than to accurately characterize Long Covid. And that the scoring system is not a validated way to diagnose Long Covid and should not be used for that purpose.

  3. Using the word “definition” for Long Covid excludes children and is false and misleading:

    1. The paper claims to have developed a definition of PASC by using self-reported symptoms from an adult cohort (admittedly excluding children). Specifically, the authors state,“A definition of PASC was developed based on symptoms in a prospective cohort study.” The term definition in this context typically means an act of determining or setting limits, a sharp demarcation of outlines or limits, or clarity and distinctness

    2. However, Thaweethai, the paper’s lead author, admits the method used “cannot find everyone” with PASC. He also admits to using a label of “indeterminate” to exclude those who may have PASC from those that are past the artificially derived “optimal threshold” for identifying PASC positivity. Quite literally, Thaweethai admits that people with PASC can be in the indeterminate group. Simply said, this does not fit the definition of the term “definition”.  

    3. The wording is therefore misleading regardless of the intent of the authors. The paper does not present a definition or anything resembling a definition to the scientific community, clinicians, or patients. The document should remove all reference to the term definition, or specifically admit that it failed to accomplish what it claims.

  4. The 6 month threshold could prevent both adult and childhood access to life-saving and life-altering care and early diagnosis. Six months is identified in the paper as the cut-off for being PASC-positive or PASC-indeterminate using the 12 point cutoff score. This would allow ongoing NIH research to define Long Covid without including data from those who appear to have recovered from the disease prior to 6 months, or have been evaluated as such. This goes against common medical knowledge that “anything that helps prevent & detect the disease at an early stage is crucial.” The authors should issue a correction stating that one to six months is sufficient to identify Long Covid cases, and that 6 months is appropriate for determining Chronic Long Covid in patients until biomarkers are established from proven etiology. Furthermore, it is not a given that symptoms collected and scored at a certain time point from infection will stay the same in time both in a single patient, and across the patient population.Data is important in all research regardless if it reduces the percentage of prevalence and incidence.

  5. The scoring system is flawed in that it neglects to incorporate scores for symptom severity and instead chooses to use a quantitative value that misrepresents the complete story of what people with Long Covid are experiencing.  

    1. It is within reason that many symptoms mentioned would not occur in every Long Covid sufferers’ top twelve.

  6. As mentioned by the authors, the 12 point scorecard criteria does not emphasize or include the most common symptoms and though it is not intended to be used to diagnose Long Covid, the fact that the symptoms criteria omit large portions of people with Long Covid will be harmful to the community.

    1. According to the NIH press release, its application in clinical trials can now be used to influence the selection of patient groups; this would result in using public funds for scientific trials that are not representative of the Long Covid population.

    2. The scorecard also could be used to improperly diagnose or exclude Long Covid patients from enrollment in upcoming trials, diagnoses, delaying further testing, appropriate treatments, and access to disability.

    3. While the authors aim to find and highlight symptoms to differentiate Long Covid patients from others, the 12 point criteria system excludes a large portion of the Long Covid population without valid scientific reasoning. I strongly believe the next attempts at diagnostic criteria should take into account existing literature that shows more specifically defined symptoms for Long Covid, from objective findings. (E.g. PoTS, Vestibular issues, migraine, vs more vague symptoms like "headache" or "dizziness.) LCAP noticed that while PEM was used as a specific symptom with a high score to produce PASC-positive results, other suites of symptoms, like those in the neurologic category, could have produced an equal or higher score than PEM if questionnaires had not separated neuro-symptoms into multiple subtypes and reduced their total scores. This alone could have created a more scientifically accurate picture of the Long Covid population.

  7. If a 12 point symptom-based system is used to decide who’s included in clinical trials without including any biomarkers and imaging, an absence of care and screening for ongoing damage and impairment to the immune system, multiple body systems, and major organs, is now a grave concern for people with Long Covid. Biomarkers and imaging, while not always unique to Long Covid, have already been established in multiple studies to identify specific clinical subtypes instead of the one-point scoring assigned to a limited group of symptoms. It is possible that biomarkers and imaging will be more useful in choosing representative patient populations for trials without excluding people with Long Covid, but it is also possible that using biomarkers without knowing the pathophysiology of the disease could prove equally detrimental to the scoring system and not represent all people with Long Covid. Excluding patients who do not have access to adequate imaging and other tests, or who are chosen on the basis of a restrictive clinical case definition or score, will prevent their access to trials and/or diagnosis; it means life-saving findings could be missed. Symptoms by themselves are only part of the Long Covid picture. NIH RECOVER should look at the context as a whole of this disease in order to establish better treatments.

  8. Data integrity is of the utmost importance to finding the pathophysiology of Long Covid in the NIH RECOVER project. LCAP is sharing the following information and questions in the hopes to strengthen RECOVER data.  

    1. LCAP has learned from Public Herald, an investigative news agency, that verified post-vac sources are part of the NIH RECOVER Long Covid symptom data. It is within reason that compromised data could exist in the RECOVER program with people suffering from post-vac illness who are enrolled in both healthy controls without infection, and PASC cohorts. Although participants explained to NIH their post-vac symptoms prior to enrollment, they were still allowed to participate in Long Covid studies to share symptom data. Therefore, in these specific examples, the integrity of the data should be called into question. This would require new questionnaires to be delivered to all existing NIH participants with a separation of symptom data for post-vac cohorts collected and studied within the NIH RECOVER studies. Unreliable data which includes post-vac cases can result from various factors such as measurement errors, data collection flaws, data entry mistakes, sampling biases, or inadequate study design. It is crucial to identify and address these issues to ensure the integrity and quality of the study's findings.

As a person living with Long Covid or an advocate in the community, I feel invested in ensuring that an accurate study is shown to the public and healthcare professionals from the NIH RECOVER program. Therefore, I ask the NIH and the authors of the paper to seriously consider corrections or the possibility of a retraction for the aforementioned study and to publicly address the concerns outlined here. Retraction would provide an opportunity to reassess the findings and address our concerns with the aim of preventing any potential harm that may arise from misinterpretation or misapplication of the study's conclusions.

I would like the chance to have a constructive dialogue with the authors of the paper and the NIH to discuss these concerns in full. Together, we can work towards rectifying any potential harm caused by the publication and promoting an accurate understanding of Long Covid symptoms that is in everyone's interest.

It is crucial that the NIH RECOVER program publishes studies with a better understanding about its real impacts on people living with Long Covid — especially children who need immediate care after developing symptoms — and with data that the public can verify.

In the future, many of these concerns can be avoided if the NIH holds public meetings with the Long Covid community to maintain transparency and discuss the preprint of any paper scheduled for release.

It is also important for NIH RECOVER to prioritize studies focussing on pathophysiology and severity of Long Covid for research and enrollment in upcoming clinical trials.  

I hope that this letter will be taken seriously, and NIH will begin to address the reasons identified for prompt action. I believe that by working together, we can ensure the responsible flow of scientific research and protect the well-being of people with Long Covid.

Thank you for your attention to this matter. I look forward to your response.

Yours sincerely,

*This is the first in a series of letters from LCAP explaining why the “Development of a Definition of Postacute Sequelae of SARS-CoV-2 Infection” published in The Journal of the American Medical Association should be corrected and considered for retraction by NIH RECOVER.

Letter Campaign by
Long Covid  Action Project
Pittsburgh, Pennsylvania