A literature search and analysis of 295 long-COVID studies finds that only 35% used agency-established definitions of the condition, which the authors said complicates comparison of findings, translation into clinical management, and cost-effectiveness assessment.
The research was published yesterday in JAMA Network Open.
A growing body of research
A team led by University College London researchers searched PubMed to determine how many of 295 included studies used the long-COVID definitions published from December 2020 through October 2021 by the US Centers for Disease Control and Prevention (CDC), the World Health Organization (WHO), or the UK National Institute for Health and Care Excellence (NICE).
"Despite the growing volume of research on lasting symptoms of COVID-19, the definition has not been universally agreed on," the authors noted.
The studies were published from February 1, 2020, to October 26, 2022, and included two randomized clinical trials (0.7%), 134 cohort studies (45.4%), 66 cross-sectional studies (22.4%), 13 case-control studies (4.4%), 45 case reports or case series (15.3%), and 35 studies using other designs (11.9%). A total of 56.6% were performed in Europe.
33% didn't define condition at all
Only 102 studies (34.6%) used one of the three established long-COVID definitions (NICE, 56; WHO, 31; and CDC, 15), and 193 studies (65.4%) did not use any of the definitions. Six studies were submitted for publication before NICE published its definition on December 18, 2020.
Among the 193 studies that did not follow any agency-established long-COVID definition, 129 (66.8%) used their own definitions (eg, chronic symptoms lasting longer than 5 months or 2 weeks after infection), while 64 studies (33.2%) did not define the condition at all.
"This study highlights major issues in comparing interventions and outcomes between these reported studies in PCC [post-COVID condition] due to differences in definition," the researchers wrote. "The differences also result in considerable variation when translating findings into clinical management and cost-effectiveness assessments of interventions in patients with PCC."
The authors called for a clearer definition of long COVID so that evidence from clinical trials can be reliably applied to evidence-based clinical management of the condition to improve the well-being of affected patients.
One name, one definition
In a related commentary, Daniel Pan, MBBS, and Manish Pareek, MBBS, PhD, both of the University of Leicester, said that using different definitions leads to inclusion of different patients, symptoms, and time intervals. "Hence, any estimated prevalence or incidence of PCC will vary widely between studies, even if conducted in the same time period in the same population," they wrote.
This study highlights major issues in comparing interventions and outcomes between these reported studies in PCC [post-COVID condition] due to differences in definition.
To eliminate the confusion, Pan and Pareek suggested use of a universal definition, namely, "Signs and symptoms following initial SARS-CoV-2 infection, that persist for more than one month (in mild cases), and more than three months (in cases severe enough to warrant oxygen support), which have a disproportionately severe effect on a patient's quality of life, far beyond what is expected from their initial infection."
This definition, they said, differs from those of NICE, the CDC, and the WHO in that it considers infection severity and thus acknowledges that long COVID is different from the prolonged recovery from, say, straightforward pneumonia. "Our suggested definition also acknowledges that PCC can be present even in those whose persistent symptoms can be explained, since the explanation may involve a long-term consequence of acute infection," they wrote.
Pan and Pareek also said they also would prefer to refer to the condition as "long COVID" rather than the WHO- and JAMA journals-preferred "post–COVID-19 condition. "[Long COVID] was the term originally developed by patients, not clinicians, to raise concerns about the impact of their symptoms on their well-being," they wrote. "Even if a patient's symptoms following infection can be explained, it should still be included in the definition so that treatments can be investigated."
