Are we exaggerating the frequency of long Covid?
US authors worry that long Covid's prevalence is exaggerated owing to vague definitions, methodological flaws, and lack of control groups
Our understanding of Covid is constantly evolving, much as the virus itself. For example, new evidence shows that anti-viral drugs aiming to treat the disease may in fact be fuelling its evolution.
Scientists said antiviral use increases the genetic diversity of the virus in the wild, and provided more options for future evolution.
Another area is long Covid. Some studies estimate the prevalence of post-Covid symptoms is 25-45 per cent. However, scientists have cast doubt on these studies.
An opinion piece in BMJ Evidence-Based Medicine claims these studies have been misleading and vastly overestimate the prevalence of long Covid.
"Overly broad definitions", "methodological flaws", "striking absence of control groups": these are just some of the issues of long COVID research studies, according to scientists writing in an influential medical journal.
The US authors believe poorly conducted studies and lack of clarity around how to diagnose long Covid has led "undue concern and anxiety" in the public. They claim that treatable conditions are commonly misdiagnosed as long Covid, which diverts attention and funds away from people who truly suffer from conditions not caused by Covid-19.
"I think this publication is important because it gives food for thought. The first thesis put forward by the authors is that the frequency of long Covid is overestimated. I partially agree with that," said Andreas Stallmach, head of the long Covid centre, Jena University Hospital, Germany.
The researchers call for clearer definitions and diagnostic criteria for long Covid, and the use of longitudinal studies with appropriate controls to better monitor the illness.
One of the major issues that the piece highlights is the lack of consistent terminology around long Covid and related post-Covid syndromes and symptoms.
None of the definitions provided by health bodies like the World Health Organization (WHO) and the US Centers for Disease Control and Prevention (CD) requires a causal link between the Covid-19 virus and a long Covid diagnosis.
The piece highlights several studies which failed to confirm prior Covid-19 infection in patients included in long Covid studies.
"Many studies include a broad range of symptoms without any evidence of a causal link to SARS-CoV-2 infection," the authors write.
"Over 200 symptoms are now associated with Covid-19. All are largely unspecific and without a causal connection to the disease," Peter Berlit, secretary general, German Society for Neurology (DGN), told media outlets.
According to Berlit, the only hallmark symptoms of Covid-19 are impaired sense of smell and taste. "There are also quite specific patterns of neurocognitive disorders that can last longer. But very few long-lasting symptoms can be attributed solely to Covid-19," he said.
The authors also claim many studies lack control groups with people who didn't test positive for Covid-19, highlight a recent systemic review which identified control groups were only included in 11 per cent of long Covid studies.
"I agree with the shortcomings described in the BMJ analysis. For example, the initial lack of control groups, the lack of a clear definition of the disease. Such deficiencies can distort the risk of disease." said Clara Leman, an infectious disease specialist at Cologne University Hospital.
Berlit added that inadequate control groups in many studies have certainly contributed to the disease risk being overestimated today.
"Even in the opinion of many doctors, the number of long COVID patients is far too high," said Berlit.
The piece highlights several, rigorous, long Covid studies which the authors believe give an accurate, "more reassuring” picture of long Covid prevalence.
For example, the study would refute the WHO's claims that 36 million people in Europe are suffering from long Covid.
Again, a UK study found that 12 common symptoms of long Covid were prevalent in 5 per cent of people 12-16 weeks after a Covid-19 infection. However, the same symptoms were found in 3.4 per cent of people who did not have a Covid-19 infection, showing only a 1.4 per cent difference.
These figures also show how relatively common symptoms are in the population at any given time, say the authors.
The study calls for a new definition of long Covid which should include continuous symptoms after a confirmed Covid-19 infection, and take into account physical and mental health, which may contribute to an individual's post-Covid experience.
"The fundamental problem is that there are currently no clear biomarkers or radiological findings that allow a clear diagnosis of long COVID," said Lehmann.
The fact remains that many people are still struggling with debilitating long Covid symptoms, with women less likely to recover from long Covid.
While the authors strongly criticised the state of long Covid research, they hope their recommendations for research of long Covid will improve standards of medicine for those who do have long Covid.
Berlit agrees, saying we need to know more about patients' history and social situation, as many previous physical and mental illnesses lead to an increased risk of long Covid.
"I'm not talking about long Covid being an imaginary illness, but depending on your previous illness and your life situation, just dealing with the potentially serious infection can lead to long-lasting symptoms," said Berlit.
Researchers have identified a number of risk factors, but are still trying to work out what exactly causes long Covid in some people, and not in others.