What the Long Covid Numbers Aren’t Telling Us

Nearly two years ago, as the first confirmed Covid-19 cases in the United State began trickling in, before scientists understood how SARS-CoV-2 affected and infected humans, numbers were all we had. Our grasp on the magnitude of what was becoming a global pandemic came from state and national statistics, like the numbers of new Covid-19 cases, hospitalizations, and deaths. 

Not long after, in spring of 2020, some of the first people diagnosed with Covid-19 began speaking publicly about how their symptoms lasted longer than the two weeks public health experts said was the standard duration of infection. Since then, they — along with estimates of millions of other Americans living with Long Covid — have been met with disbelief from family members, employers, and the medical profession, and constant requests to prove that they are legitimately sick. 

Because numbers have been integral to people’s understanding of the severity of the pandemic since it began, it’s not surprising that Long Covid skeptics often ask questions about how many people “actually” have the condition, and what percentage of people with Covid-19 end up with Long Covid. 

At the moment, there are statistics that appear to answer these questions, most of which indicate that between 10 and 30 percent of people who get Covid-19 end up with Long Covid. But these are only estimates emerging from academic research and referenced in news coverage — not the kind of population-level data that exists for other conditions, and that people expect to be readily available for Long Covid. In fact, those statistics are likely still years away.

One of the main reasons coming up with definitive Long Covid statistics is such a lengthy process is the same reason the current estimates are frequently misunderstood: Long Covid has yet to be adequately defined. And while the broad, vague definitions from organizations like the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) were a necessary step in the recognition of Long Covid as a legitimate medical condition, in reality, Long Covid encompasses multiple categories of patients with vastly different symptoms, experiences, and outcomes. 

Here’s what to keep in mind about the current numbers — including why estimates range so widely — as well as what needs to be done in order to get population-level statistics.

The type of Long Covid statistics people expect are the kind that result from large-scale studies conducted by the National Institutes of Health (NIH) that follow a group of people over a long period of time, says Natalie Lambert, PhD, a biostatistician and associate research professor of biostatistics and health data sciences at the Indiana University of School of Medicine. In this case, she says, that would involve enrolling participants when they test positive for Covid-19 and keeping track of their health long after — even if they don’t immediately develop Long Covid. “That’s the gold standard, but those tests take a very long time,” Lambert tells Rolling Stone. 

As new research on Long Covid is published, it’s important to keep in mind that the studies were conducted with the populations and knowledge that were available at a particular time — and not necessarily with the aim of coming up with national Long Covid statistics. Early on, physicians and researchers were trying to learn whatever they could about this new type of post-viral illness, and at times, that meant working with a narrow sample of participants — like ones who had been hospitalized, for example. 

“But that’s leaving out the majority of people [with Long Covid] who don’t end up in the hospital,” Lambert explains. “We could try to do better by trying to follow up with people who went to their primary care provider and seeing how many of them became long-haulers, but most of them aren’t going to the doctor.”

Although the current definitions of Long Covid are incredibly broad, before they existed, there was even less certainty about what “counted” as Long Covid among patients, researchers, and medical professionals. “Long Covid means different things to different people,” says Jaime Seltzer, director of medical and scientific outreach for #MEAction, an advocacy organization for people living with myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) — including the estimated 10 to 12 percent of those with Long Covid who will develop the condition. “Studies are measuring Long Covid at different intervals and capturing different populations within Long Covid, which may explain the discrepancy in the statistics.” 

For example, a study that includes a large number of people who were hospitalized, may be capturing a segment of people who are dealing with post-intensive care syndrome, Seltzer explains. Or, data gathered soon after a person’s acute Covid-19 infection may identify people with post-viral fatigue that is recoverable, while missing people who develop ME/CFS months later. 

Lambert estimates that it will take years to get accurate and reliable population-level data on Long Covid, but notes that “the more funding there is, and the more researchers that can be involved, the faster we’ll have the answers.” And before anything else, she says, Long Covid needs to be defined.

Nearly a year into the pandemic, David C. Lee, MD, began noticing a new subset of patients in the emergency department of NYU Langone Health. Their symptoms ranged from chest pain with shortness of breath, to abdominal pain, to one or more abnormal skin problems, and either started, or never went away after they had Covid-19. But in early 2021, what stood out the most to Lee, who is also an assistant professor at NYU School of Medicine, was that “the overwhelming majority of patients had already seen several doctors, often including multiple specialists,” he tells Rolling Stone.

Treating Long Covid patients who came to the ER, Lee also saw that the chronic condition isn’t equally debilitating for everyone. “There is varying severity of these persistent symptoms from person-to-person,” he explains. “Some people are completely disabled and have been sick since 2020, but then there are a lot of people who don’t have it as badly.”

And while the lack of a clear definition of Long Covid makes getting accurate statistics a challenge, Lee points out that’s to be expected when dealing with an illness that’s been around for fewer than two years, and isn’t something that should be rushed. “For a new condition like Long Covid, we need to be careful about deciding what it is or isn’t before we have all the evidence worked out,” he says.

For context, the WHO’s definition of Long Covid (announced in October 2021) — which the organization refers to as “Post-Covid-19 Condition” — notes that it typically begins around three months after the onset of a Covid-19 infection, lasts at least two months, and can’t be explained by another diagnosis. And while the CDC provides a similarly broad introduction to Long Covid on its website, it goes on to identify three types of “post-Covid conditions”: new or ongoing symptoms; multi-organ effects of Covid-19; and effects of Covid-19 illness or hospitalization. 

While breaking Long Covid down into categories, rather than approaching it as a catch-all condition, is an important (and necessary) step towards defining the condition — and ultimately, getting closer to accurate Long Covid statistics — it has yet to be seen whether this will change the underwhelming approach the CDC and government have taken so far.

When Lambert partnered with Survivor Corps, an online support group for people with Long Covid, in April 2020 and began researching Long Covid, she and her colleagues essentially had to start from scratch. “At the very beginning, we were interested in talking to people that had Covid, and two months later still had not recovered,” she explains. “We asked them about their symptoms to get a basic understanding. However, not recovering from something isn’t a good definition of a disease.”

This is something that as a biostatistician, Lambert knows from experience. “To do a rigorous scientific study where you’re studying a population of people who have a disease, you first have to define the population,” she explains. “And you can’t say how many people have Long Covid unless you know that whatever sample of people you ask about their recovery or non-recovery after infection is representative of everyone who got Covid.”

In order to identify someone who has a disease (as well as someone who doesn’t), scientists first need to fully understand that condition, its characteristics, and the mechanisms of how it occurs, Lambert says. But much of the research into the various symptoms that can continue or begin after a Covid-19 infection is either underway, or has yet to start — meaning that the deep understanding of Long Covid necessary to define it is still a way off. For now, though, thinking of Long Covid split into distinct categories, like the ones researchers have proposed and the CDC lists, can help provide a clearer picture of what the chronic condition entails, and how it affects people differently.

And unlike many diseases, there’s no clear signal, like a diagnostic lab test, that is seen as sufficient evidence that a person has Long Covid. Plus, as Seltzer explains, when there aren’t clear diagnostic tools and/or adequate medical knowledge on a condition, “stigma takes a stranglehold on moving the field forward,” she notes. “Instead of medical providers showing uncertainty, they misdiagnose the patient as having a psychiatric disorder like depression, anxiety or conversion disorder, or dismiss the patient entirely as attention-seeking.” 

Given that, as Lambert points out, “our cultural mindset about health is that we have to prove we’re sick,” not having statistics on how many people are living with Long Covid makes the task of getting people to believe the condition exists even harder.

Long Covid has a numbers problem — but not necessarily the one most people think. For those who have been following the trajectory and gauging the severity of the Covid-19 pandemic through daily reports of the new cases and infection rate in their area, not having a definitive Long Covid statistic may be frustrating, or make it easy to simply dismiss the condition.

But what if part of the problem is the weight given to Long Covid numbers, and the pervasive idea that population-level statistics are the only legitimate way not only to study a condition, but even prove that it exists? 

“I think we need to go to the other end of the spectrum, talking directly with as many patients as possible to learn about their range of experiences in detail, and using that [knowledge] to help  inform clinical studies,” Lambert explains, noting that this would ensure that researchers are asking the right questions, as well as considering factors like stress levels, environment, and other social determinants of health. “There are many different fields of research that do incredible work that is valid and trustworthy, and many methods of inquiry that can find really important patterns about the world that are very useful to us. And we need all hands on deck.”

This isn’t to say that having accurate, reliable Long Covid statistics resulting from robust, longitudinal clinical studies isn’t important. But a lack of these numbers shouldn’t be used as an excuse not to take Long Covid — and the people living with the condition — seriously. 

“At the end of the day, we need to listen to patients better, help them in the ways that we can, and keep searching for better treatment options,” Lee says. “Each person affected [by Long Covid] is meaningful, and maybe the statistics don’t matter as much as the individuals do.”