Variability in Long COVID Definitions and Validation of Published Prevalence Rates, 2025, Wisk et al

Wyva

Senior Member (Voting Rights)
Key Points

Question
What is the prevalence of long COVID in the Innovative Support for Patients With SARS-CoV-2 Infections Registry (INSPIRE) cohort based on definitions in published literature?

Findings In this prospective cohort study of 4575 INSPIRE participants, when replicating long COVID definitions used in the 5 published studies with the largest sample size of reviewed studies, the prevalence of long COVID among participants who were COVID-19 positive varied at 3 months from 32% to 42% and at 6 months from 15% to 22%.

Meaning The findings suggest that long COVID prevalence estimates vary as a function of disparate definitions used.




Abstract

Importance
Long COVID definitions vary widely, and no consensus exists on how to accurately measure its prevalence, complicating both clinical care and research.

Objective To assess long COVID prevalence using various definitions from published literature.

Design, Setting, and Participants This prospective, multicenter cohort study used data from the longitudinal Innovative Support for Patients With SARS-CoV-2 Infections Registry (INSPIRE). Participants aged 18 years or older with symptoms suggestive of COVID-19 illness at the time of their index SARS-CoV-2 test enrolled at 8 sites across the US from December 11, 2020, through August 29, 2022, with follow-up surveys collected through February 28, 2023.

Exposure Positive or negative SARS-CoV-2 test result at the time of acute symptoms.

Main Outcomes and Measures Long COVID prevalence among INSPIRE participants with a positive vs negative index SARS-CoV-2 test, based on long COVID definitions in published literature. Secondary outcomes were sensitivity and specificity of published definitions compared with self-reported long COVID.

Results A total of 4575 INSPIRE participants were included (mean [SD] age, 40.40 [14.58] years). Most were female (3013 of 4448 [67.7%]) and aged 18 to 49 years (3338 of 4541 [73.5%]). Applying 5 published definitions for long COVID yielded a prevalence that ranged from 30.84% (95% CI, 29.33%-32.40%) to 42.01% (95% CI, 40.37%-43.66%) at 3 months and 14.23% (95% CI, 13.01%-15.55%) to 21.94% (95% CI, 20.47%-23.47%) at 6 months postinfection; in the 5 comparator studies, reported prevalence of long COVID at 1 to 5 months postinfection ranged from 2.6% (≥84 days) to 47.4% (3-5 months) and at 6 or more months postinfection ranged from 10.0% (95% CI, 8.8%-11.0%) to 61.9% (6-11 months). Using participants’ self-reported long COVID as a criterion standard, existing published definitions had low-to-moderate sensitivity (up to 66.32% [95% CI, 62.59%-69.90%] at 3 months and 45.53% [95% CI, 41.51%-49.60%] at 6 months) and high specificity (up to 81.29% [95% CI, 79.32%-83.15%] at 3 months and 94.26% [95% CI, 92.98%-95.37%]) at 6 months.

Conclusions and Relevance In this cohort study, variability in long COVID prevalence across published definitions highlights the need for a standardized, validated definition to improve clinical recognition and research comparability, ultimately guiding more accurate diagnosis and treatment strategies.

Open access: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2837446
 

News Release 12-Aug-2025

What exactly is Long COVID? New UCLA research shows the answer depends on whom you ask​

Lack of a standard definition leads to widely varying estimates, complicating care and research

Peer-Reviewed Publication
University of California - Los Angeles Health Sciences


What is Long COVID? Despite hundreds of published studies and millions affected worldwide, the medical field still lacks a clear answer.

New research from UCLA finds that the definition of Long COVID varies so widely across published studies that the percentage of people identified as having the condition can differ dramatically, making it harder to treat patients and advance research.

The study, published August 12 in JAMA Network Open, highlights just how much the lack of a standard definition is clouding our understanding of Long COVID.

“The findings highlight the need for a standard definition for Long COVID,” said study lead Lauren Wisk, an assistant professor of medicine in the division of general internal medicine and health services research at the David Geffen School of Medicine at UCLA. “Up to one third of the variation in the published studies may stem from the fact that they use different definitions for Long COVID.”

Using data from 4,700 U.S. patients in the CDC-funded INSPIRE cohort, UCLA researchers applied five published Long COVID definitions from studies conducted in the US, U.K., Netherlands, Sweden, and Puerto Rico. The result? Depending on which definition was used, the estimated prevalence of Long COVID ranged from 15% to 42% - all in the same group of patients.

The five definitions assessed in the study differed by symptom duration (4 weeks to 6 months) and number of symptoms considered (9 to 44), resulting in reported prevalence of Long COVID that ranged from 2.6% to 61.9% in the original studies. When these different published definitions were applied to the INSPIRE cohort, these differences also created significant variation in who was labeled as having Long COVID.

Researchers also found that published definitions had only moderate sensitivity –the ability to correctly identify those with the disease— compared to participants’ own self-reported experience of Long COVID. While specificity of these definitions was better, none of the definitions reached an ideal level to be considered the optimal identification test.

“If every study on Long COVID uses a different definition for identifying who has it, the scientific conclusions become harder to compare across studies and may lead to delays in our understanding of it,” Wisk said. “In the absence of an objective measure, like a blood test, or a uniform standard for measuring Long COVID, researchers and clinicians will need to decide which definition is best suited for their scientific question and be more transparent about the potential limitations of using a more vs less restrictive definition.”

These differences may lead researchers and physicians to miss some legitimate Long COVID cases and to label some patients as having Long COVID when they don’t, said Dr. Joann Elmore, professor of medicine in the division of general internal medicine and health services research at the David Geffen School of Medicine at UCLA and the study’s senior author.

“Without a clinically usable and standardized research definition of Long COVID, it’s like every study is using a different measuring stick,” she said. “That makes it hard to compare results, develop treatments, or track progress. Without a shared definition, we risk mislabeling patients and misguiding care. This is more than an academic debate- it affects real people.”

Study limitations include potential recall bias among participants; participant demographics that may have influenced prevalence were unexamined, and results may have differed had the researchers included more than the five published studies in their analysis.

Study co-authors are Michelle L’Hommedieu and Kate Diaz Roldan of UCLA, and others from the INSPIRE Group.

INSPIRE was funded by the National Center of Immunization and Respiratory Diseases (75D30120C08008) in the Centers for Disease Control and Prevention.




Journal​

JAMA Network Open

DOI​

10.1001/jamanetworkopen.2025.26506

Method of Research​

Literature review

Subject of Research​

People

Article Title​

Variability in Long COVID Definitions and Validation of Published Prevalence Rates

Article Publication Date​

12-Aug-2025

 
Back
Top Bottom