woman with a thermometer up her head, with a question mark on it
Collage by Cathryn Virginia | Image from Getty

The Medical System Should Have Been Prepared for Long COVID

COVID long haulers have been breathlessly covered, but there’s nothing surprising about medically unexplained symptoms—or the reaction to them.

Last March, Lisa O’Brien began experiencing mild symptoms of what she thought might be COVID-19: scratchy throat, body aches, fatigue. But with no fever and just a hint of a cough, she decided to wait on a COVID test. Then new symptoms appeared, including chills and shortness of breath. After two weeks they were alarming enough for her to get tested: Negative.

But the negative result didn’t resolve her health problems. Four weeks out, O’Brien was barely able to breathe and she returned to the hospital. “I’d had anxiety before,” she told me. “But the shortness of breath was like nothing I’d ever experienced.”


Her visit followed a pattern, all too familiar to so-called COVID long haulers. “They did a bunch of tests. And at the end? The doctor asked me if I understood the power of suggestion.” 

“I had no proof that I’d had COVID,” O’Brien continued. “I felt kind of crazy already, and he was validating my feeling of being crazy.”

Nearly a year later, long haulers have been featured in nearly every prominent media outlet telling stories just like Lisa’s. In November, avid marathoner Nitza Rochez broke down in tears as she recounted her experience to 60 Minutes’ Anderson Cooper. Rochez had been suffering from months of brain fog, tremors, dizziness. One day her legs collapsed and she went to the ER. All the tests came back normal.

“The doctors were like, ‘You're fine. You're having anxiety attacks. You're just nervous. Breathe.’” 

“They thought it was in your head,” Cooper replied to Rochez.

“They thought it was in my head. And it's one of those moments that I'll never forget because how can I possibly be fine? And when I left the emergency room that day, I was like, ‘I'm just going home to die.’”

Long COVID is now widely accepted by scientific experts. Post-COVID clinics exist in 39 states, many housed at flagship medical institutions. But patients continue to be treated this way. (As this article was being prepared for publication, Lisa O’Brien reached out to tell me she had been referred to a psychotherapist who suggested her symptoms might be caused by generalized anxiety disorder and stress.)


“It’s a common refrain we hear again and again, symptoms were chalked up to stress and anxiety,” said Dr. Lekshmi Santhosh, a pulmonary and critical care specialist who runs UCSF’s multidisciplinary post-COVID clinic. Santhosh’s clinic is limited to patients who were hospitalized for acute COVID. That means the people she’s treating had severe, officially documented cases, and still received psychological diagnoses of persistent symptoms.

Dr. David Putrino, Director of Rehabilitation Innovation for the Mount Sinai Health System, grimly described his interactions with physicians for whom anxiety is still a default explanation of long haulers’ symptoms.

“Everyone should know that long COVID is real, but a lot of people are not there yet. I am having many conversations with [doctors] from states that are now where New York was in March/April in terms of COVID numbers. They’re coming to me completely fresh, completely naïve, saying symptoms of COVID aren’t supposed to last this long.”

COVID long haulers are frustrated enough that they face routine skepticism and dismissal. Then, in their debilitated condition, they feel they must also educate themselves through patient advocacy groups on how to demand follow-up tests and specialist screenings from reluctant physicians.

The problem, however, predates long COVID. As soon as COVID-19 hit, physicians and scientists should have been prepared for long haulers. When SARS—another coronavirus—hit Toronto in 2003, researchers documented multiple cases of persistent fatigue, pain, and sleep disorders that lasted up to 36 months, which they called “chronic post-SARS.”


And it’s not just coronaviruses. We knew about post-Ebola syndrome. Post-West Nile virus. Post-Epstein Barr (mononucleosis).

“Every virus has a post-viral syndrome,” said Putrino. “In a global pandemic we’re going to have a certain number of patients present with a post-viral syndrome, and shame on us for not being ready.”

How could top scientists and medical professionals suffer collective amnesia about this crucial piece of information? Why weren’t we being warned about two types of potential COVID complications, acute and chronic? 

The answer is simple: Our medical system is radically unequipped, practically and conceptually, to serve patients whose tests come back normal and whose chronic symptoms cannot be explained with a biological diagnosis or outsourced to a specialist. 

Long COVID patients are far from the only ones in this situation. Millions of people suffer from similar chronic symptoms, many of them too debilitated to work a job or even leave their bed. They, too, have been told their symptoms are psychogenic. Those I spoke with recounted how they watched in horror as the first reports of post-COVID began to surface. They saw what was coming, even if the doctors and scientists didn’t.


And so, to understand the catastrophic lack of preparation for post-COVID syndrome and the treatment of those who suffer from it, we must turn to the question of why so many before them have been treated in precisely the same way.

In 1988, an essay in the British Medical Journal coined the term “heartsink patients” to identify patients who evoke in doctors “an overwhelming mixture of exasperation, defeat and sometimes plain dislike that causes the heart to sink when they consult.”

Among the characteristics associated with heartsink patients are thick clinical records, being female (in the original essay, 22 of those described were women and 6 were men), repeated visits to healthcare providers, and, most importantly, the technical term “medically unexplained symptoms” (MUS).

MUS might appear self-explanatory: medically unexplained means… medically unexplained! After examining the patient and running appropriate tests, the doctor simply doesn’t know what’s causing the symptoms—exactly the situation that early COVID long haulers found themselves in. 

But as with the term UFO, the literal uncertainty of MUS has turned into practical certainty. Instead of aliens, the explanation of medically unexplained symptoms is almost always identified as psychology. 


One of the world’s foremost experts on the problems with MUS is the philosopher and bioethicist Diane O’Leary. Studies of MUS put its prevalence at anywhere between 10 and 86 percent of outpatient visits; O’Leary splits the difference at 50 percent, still a remarkably high number.

So how are these enormous numbers of patients viewed by clinicians? In a scholarly article, O’Leary gives the example of a hypothetical general practitioner who looks up the term in Macleod’s 2013 Clinical Examination Handbook and finds the following entry: “MUS cause similar levels of disability to those resulting from disease and are often associated with emotional distress. ”

“Every virus has a post-viral syndrome. In a global pandemic we’re going to have a certain number of patients present with a post-viral syndrome, and shame on us for not being ready.”

The implication here is that MUS do not result from disease, and therefore their explanation must be psychosocial—an implication consistent, O’Leary notes, with the classification of MUS as “psychiatric” in the Oxford Textbook of Medicine.

“In medicine, the leap between ‘doctor doesn’t know’ and ‘problem is in the mind’ is instant,” O’Leary told me. “They are the same thing.”


The medical literature offers endless evidence to support her assertion. Last May, for instance, while COVID long haulers were struggling with dismissive doctors, a case report was published with the title “Medically Unexplained Symptoms in the Times of COVID-19 Pandemic.” The authors proceed by assuming MUS means psychosocial, and caution that “extreme anxiety reactions to COVID-19 may elicit unexplained medical symptoms in predisposed individuals.”

The classic modern example of medically unexplained symptoms leading to mistaken psychogenic diagnoses is myalgic encephalomyelitis or “chronic fatigue syndrome.” According to the CDC, an estimated 836,000 to 2.5 million Americans have ME/CFS, 90 percent of whom haven’t been diagnosed. ME/CFS is tremendously disabling, leaving people incapable of doing simple tasks like brushing their teeth, and renders around 1 in 4 sufferers bedridden or housebound for extended periods of time. (Like heartsink patients and long COVID sufferers, women are overrepresented.)

ME/CFS qualified as medically unexplained because no biological cause had been conclusively identified and it had no standard diagnostic tests. Consequently, “the doctor doesn’t know” became “the problem is in the mind.” Patients begged to be taken seriously while physicians and scientists scoffed at the “yuppie flu” and psychologized their symptoms.


In Osler’s Web, the definitive book on the emergence of ME/CFS in the 80s and 90s, the journalist Hillary Johnson excoriates the scientific community’s lack of interest in the condition. As a point of reference she observes that in “the early 1990s, the NIH was spending $75,000 in research for every HIV-infected American per year.” By comparison, NIH yearly spending on ME/CFS research has hovered at around $2-6 per patient.

Those who saw Dr. Anthony Fauci recently acknowledge that COVID can lead to a post-viral fatigue syndrome resembling ME might be surprised to read Johnson’s account of how, as head of the National Institutes of Allergy and Infectious Disease in the 90s, Fauci ascribed ME/CFS to psychiatric issues and relegated its study to the fringes of the NIH.

When funding is secured to study the condition, psychology almost always plays a central part in the proposal. This reinforces the idea that the underlying problem is mental, not physical. In 2005, British researchers were awarded £5 million, the most ever for research on ME/CFS, to test the efficacy of Cognitive Behavioral Therapy and Graded Exercise Therapy as treatment. Dubbed the PACE trial, this controversial study essentially sought to vindicate what was already the status quo for patients—a combination of therapies meant to decondition their fear of fatigue. Fear, not physiology, was assumed to be the underlying cause. 


The supposedly successful trial results were published in The Lancet in 2010 over the vehement objections of patient advocates. They complained that CBT was ineffective and GET could actually be harmful. But the results fit too neatly with existing dogma about ME/CFS. Despite letters and commentaries (and even scientific studies) the PACE trial remained the gold standard for treatment, endorsed by every major medical organization. Finally, in 2015, the journalist and public health expert David Tuller published a 15,000 word critique of the study, and the tide began to turn. Now the CDC no longer recommends CBT or GET for symptom management, and new UK guidelines have followed suit.

But physicians don’t keep up on every change in the medical literature, which means new recommendations don’t filter down to new patients. This is especially concerning given that long COVID is leading to symptoms that appear to line up with ME/CFS. 


“I still have doctors who are prescribing me exercise as a person with long COVID,” said Dr. Margot Gage, an epidemiologist at Lamar University in Texas who has struggled for months with having her symptoms properly addressed. “That’s the worst thing in the world you could tell somebody that’s in my condition.” 

Just as a disease doesn’t exist if you can’t test for it or observe it, a problem doesn’t exist if you can’t quantify it. Experiences like Gage’s and O’Brien’s are anecdotally common, but no one is keeping track of how often long COVID is being dismissed as anxiety or addressed with faulty treatment protocols—no one, at least, outside of patient advocacy groups like the one founded by Karyn Bishof, a former firefighter and paramedic from Florida whose life was derailed by long COVID. She used Facebook to survey group members about their experiences, including “the response of medical practitioners with longhaulers.” 

Although a significant number of the respondents indicated their doctors were sympathetic and attentive, others were less enthusiastic. The group has fewer than 3,000 members and only a small portion respond to surveys. Even so there were 231 votes for my practitioner said ‘just give it time you will be fine’; 157 for my practitioner said my issues are from anxiety; and 130 for my practitioner dismissed my concerned [sic], didn’t listen to me and didn’t believe me


“They love to say it’s anxiety, or it’s psychosomatic, or you’re just doing too much,” Bishof told me. “But this is leading to chronic conditions that will be with us for the rest of our lives. And these things that we’re getting have no cure, and they’re the same illnesses and diseases that they’ve been ignoring for decades.”

Illnesses and diseases. The way Bishof used the words, the way most of us use them, they’re interchangeable. Illness. Disease. Choosing between the two amounts to trivial semantics.

But in the world of medical terminology they mean vastly different things, with non-trivial practical consequences. You have illness with disease, that is, symptoms with corresponding physiological abnormalities, and you have illness without disease, the experience of symptoms unvalidated by biological evidence—aka medically unexplained symptoms.

There’s a bit of “if a tree falls in the forest but no one hears it” to the distinction. If someone has a disease we haven’t discovered yet—think celiac disease before 1953, or multiple sclerosis before 1868—then is it a disease or an illness? Shouldn’t we have a third category of diseases yet to be discovered? 

Science can’t answer these questions, and when O’Leary took a closer look at how medicine was dealing with terminological challenges, she found a situation that was hopelessly confused. 


“In medicine, the leap between ‘doctor doesn’t know’ and ‘problem is in the mind’ is instant,” O’Leary told me. “They are the same thing.”

Tracing the history of terms like “medically unexplained symptoms,” O’Leary showed how they have been deliberately chosen to mislead patients about the nature of their diagnosis as it is understood by the physician. These terms—call them ambidextrous—are used to soothe patients. They don’t sound like a diagnosis of psychogenic illness. But at the same time, they communicate exactly that diagnosis to other medical professionals.

Consider this passage from “To Tell or Not to Tell: The Problem of MUS,” written by the British psychiatrist Simon Wessely (who consulted on the PACE trial):

“ ‘So it is all in my mind, is it doctor?’ says the patient threateningly. The correct answer from our truth telling neurologist would of course be ‘yes’, followed by a plaintive ‘but psychiatric disorders really are genuine illnesses’, but by that time… the patient may well have left in disgust … There are insuperable objections to the neurologist ‘telling it as he sees it’.”

Another popular ambidextrous term is “functional,” which appears most often in neurology (“functional neurological disorder”) and gastroenterology (“functional bowel disorder”). The term is only applied to symptoms that have no clear biological origin, and was originally coined by neurologists and psychiatrists as a replacement for “conversion disorder,” itself a euphemism for what used to be known as hysteria. 


From hysteria to conversion to medically unexplained to functional—it’s a ride on what Steven Pinker calls the “euphemism treadmill,” in which the meaning remains constant but the term changes so as not to give offense. 

For long COVID advocates, the worst case scenario is for their condition to be categorized using one of these nebulous terms. Patients wouldn’t realize they were being diagnosed with something psychological, but members of the medical community would know the truth and treat them accordingly. 

O’Leary sees this kind of categorization practice as unethical deception. But physicians like Dr. Jon Stone, a specialist in functional neurological disorders, sees such terms as a regrettable necessity in a culture that stigmatizes mental illness and trivializes symptoms as fake or manufactured when they don’t have a clear physiological origin.

Stone told me his colleagues sometimes say “real seizures” when they mean epileptic seizures, even though he has warned them against it. The implied contrast is between “real seizures” and functional seizures. “But they’re all real seizures,” insisted Stone, “just with different causes.”

In light of medical history, his colleagues’ mistake is understandable. Before functional, the term was “pseudoseizure,” which has the connotation of fakeness built right in. Seizures with a biological etiology like epilepsy are real; seizures without an identified biological etiology are “pseudo.” This, of course, is false: seizures that have a non-biological origin are still quite real, just like vomiting from nervousness is not “pseudo vomiting.


Making things trickier is that erring on the side of biological diagnoses has its own problems. The neurologists I corresponded with for this article all emphasized that discouraging psychogenic and functional diagnoses could lead to mistaken biological diagnoses, with devastating results. “Doctors generally over-investigate for conventional pathophysiologic causes of symptoms and avoid diagnosing and treating functional disorders even when these are clearly the most likely possibility on the differential diagnosis,” Dr. Ben Tolchin told me by email. “In the realm of functional seizures, 80 percent of patients were previously misdiagnosed with epileptic seizures and treated with anti-seizure medications (which have significant adverse effects).”

But Tolchin’s 80 percent brings us back to one of the central issues faced by long COVID patients and the millions who came before them. Even if we accept that number for functional neurological disorders, the vast majority of chronic and rare diseases lack any reliable data on psychogenic misdiagnoses. As with long haulers, no one is keeping track of how many ME/CFS patients were once told they had anxiety. In addition, patients whose symptoms get psychologized are likely to have fragmented medical records from switching doctors and seeking out far-flung specialists. Should they receive a biological diagnosis 10 years later, it’s extremely difficult to trace it all the way back to some primary care physician who said ‘relax.’


Nevertheless, we have some idea of what the numbers look like. Informal patient surveys paint a grim picture: In 2014 the Autoimmune Diseases Association found that 51 percent of patients were told “their disease was imagined or that they were overly concerned.” It’s entirely possible these patients misremember what they were actually told—just like the long COVID patients testifying to similar treatment might be misremembering, exaggerating or anomalous. Without good studies, however, it’s impossible to know the truth. And until those good studies appear, the ethical approach is to err on the side of believing patients about their experiences, and look closely at what might tempt physicians to jump from “I don’t know” to a psychological diagnosis.

How long does it take for a general practitioner to diagnose medically unexplained symptoms? Like everything related to diagnostic uncertainty the question is understudied. The first quantitative study addressing it was published in January of 2020, just as the WHO began to warn about coronavirus. Researchers from the Netherlands recorded consultations and then played them for general practitioners, asking them to pinpoint when they made the determination of MUS. Since MUS is a diagnosis of exclusion—that is, all other possible etiologies must be eliminated—it ought to take longer to arrive at than other diagnoses, and only after taking a thorough patient history and running tests. This is especially true for general practitioners who are not trained in psychology. 


And yet: The average time to an MUS diagnosis for a patient presenting with new symptoms was 4 minutes and 25 seconds—and the median time, which is more representative because it eliminates the effect of outliers, was 1 minute and 47 seconds.

1 minute and 47 seconds. It seems impossibly fast until you see the median time for patients who had presented symptoms at a prior visit: 12 seconds.

The trigger for these decisions varied. Sometimes patients themselves would suggest the symptoms were related to stress. But MUS—universally understood by GPs in the study as synonymous with psychosomatic—was also diagnosed in situations that could easily describe the typical presentation of long COVID in a clinical encounter:

“She presented with a verbal flood of symptoms, and they weren’t very structured either, jumping from one thing to another and shifting from one physical symptom to the next.”

“The pattern wasn’t that logical and so on. I can understand you getting out of breath at the top of the stairs, but of course if you also get short of breath regularly when at rest, that’s a weird pattern.”

“I think he is a real MUS patient because he never has a concrete story. Of course that’s not very nice of me and I realise that.”

Especially influential was a previous diagnosis of MUS or psychogenic symptoms, which means that once one doctor says it’s anxiety, even if they are wrong, others are more likely to do the same, a phenomenon known as “anchoring bias.” O’Leary points out the anchor is much heavier when the misdiagnosis is psychogenic, leading to “diagnostic delays at least 2.5 times as long, and up to 7 times as long, as those caused by mistaken medical diagnosis.”


Or, in the words of one Dutch practitioner:

“That is simply the medical history you have. I’ve seen her a few times and the first time, well, you get a shock because she really comes across as very neurotic and crazy.”

The authors of the Dutch study conclude that “non-analytical reasoning was a central component in [the GPs] thought process.”

Doctors themselves admit as much. Dr. Ethan Weiss, a cardiologist at UCSF, was frank with me in his evaluation of the situation.

“These people don’t get better,” he said. “We all want to feel good at our jobs, and it is very frustrating not to be able to make progress. And these people are often very frustrated, they’ve been dismissed by multiple previous doctors, and so there’s a level of contentiousness when they walk in the door.”

Most crucial is the question of limited time, which came up in every one of my conversations. “If you are being rushed to see someone with a complex issue in 20 minutes, establishing a good relationship, doing an exam, a history—it’s laughable,” said Dr. Kate Chebly, chief resident in internal medicine and primary care at NYU’s School of Medicine. “PCPs [primary care physicians] would agree with that, patients would agree with that.”

Symptoms that fit a logical pattern are quickly dealt with. So are patients whose condition clearly fits a medical specialty. ME/CFS, rare diseases, long COVID do not fit a pattern or specialty. With 30 patients in the waiting room, a difficult person in the exam room, and billing policies that incentivize quick decisions, MUS provides an easy out.


A culture of medicine where certainty is the norm does not help. “It’s an understatement to say that a lot of physicians have God complexes,” said Dr. Allen Hwang, a gastroenterologist who frequently deals with difficult to diagnose symptoms. “This is becoming less of a problem with the current generation of residents and fellows,” he continued, but acknowledged that the situation was far from fixed.

Jen Brea, a patient advocate whose award-winning documentary follows her own journey dealing with psychogenic misdiagnosis of rare conditions, agrees with Hwang. 

“I think medical students are being sold a falsehood. You come, you go to school, we train you in this incredible body of knowledge, and that is the world. Anything that is beyond your training that you cannot recognize isn’t real, it’s psychological,” she said. “There’s many different ways to externalize the patient that doesn’t fit the pattern. Doctors are trained to recognize patterns, and if you don’t fit a pattern then you don’t exist. Where is the new pattern factory?”

Every physician I interviewed emphasized that standard patterns of psychosocial etiology can and do successfully identify problems. Anxiety and depression really do cause physical symptoms. Patients are poorly served by doctors who ignore non-biological causes and prescribe unnecessary drugs, perform needless surgeries, or order endless tests, all of which are costly and have potentially severe side effects. 


Many invoked the “biopsychosocial” model of health, which refuses to separate biology, psychology, and social context, instead seeing them as overlapping categories. “I don’t like the distinction between medical treatment and lifestyle treatment,” said Weiss, the cardiologist. If you have a physiological heart problem it can be exacerbated by stress, which means counseling can be an appropriate intervention. Comorbidity between chronic disease and mental illness is common; physiological disease can lead to psychological issues and vice versa.

However, overlapping categories do not entail interchangeable categories. As a diagnostic tool, “the biopsychosocial model is a disaster,” Diane O’Leary told me. “There is such a thing as biological disease and it is different from mental illness.”

Yes, psychosocial factors can cause physical symptoms, but that doesn’t obviate the difference between an actual heart attack and racing heart due to believing you’re having a heart attack—or having long COVID symptoms due to damage caused by the virus and believing your symptoms are due to the virus, when in fact they’re a product of your mistaken beliefs.

“It’s an understatement to say that a lot of physicians have God complexes.”

“A part of it is practical,” O’Leary said. “They’d be fighting all day with patients so now they say mind-body is a continuum. It’s just bullshit, using pseudo-philosophy to avoid a fight.”


As I was writing this, legal action forced the release of video footage that shows a man crawling out of Toronto’s Humber River Hospital. His excruciating leg pain had been chalked up to a previous diagnosis of bipolar affective disorder, when in fact he had a rare disorder known as Guillain-Barre Syndrome. Anxiety was the dominant symptom noted in his medical records. No one even helped him out the doors. 

“We are failing this population miserably,” Dr. Vicky Stergiopoulos, psychiatrist and physician-in-chief at the Centre for Addiction and Mental Health (CAMH) in Toronto, told the CBC. Then she proceeded to invoke the same common-sense distinction between mind and body as O’Leary did: “They go in for a broken leg and get sent to psychiatry to check their head.”

Unlike ME/CFS and long COVID, for which effective treatments are still elusive, the failure of our medical system to deal well with diagnostic uncertainty—the same failure that caused us to be unprepared for long COVID—has a number of clear solutions. Given the number of people who are already affected by these failures, and who will be affected by these same failures in the future, any further delay in implementing them is unconscionable.

The first step in reforming how medicine deals with uncertainty is understanding the scope of the problem. Studies of psychogenic misdiagnosis rates are rare. The few that exist focus on specialties like neurology and suffer from serious flaws. In a study of misdiagnosed “functional” disorder, co-authored by Dr. Jon Stone, researchers looked at the frequency that patients with functional diagnoses received a new “organic” diagnosis (note the distinction between the two categories) that better explained their symptoms within 18 months. They found rates of misdiagnosis were unbelievably low, just 0.4 percent, far lower even than typical rates of physiological misdiagnosis.


However, patients with rare diseases wait much longer than 18 months for their eventual diagnoses. Ehlers Danlos syndrome, for example, has a median time to diagnosis of ten years. Autoimmune diseases frequently take over four years to diagnose, and even rheumatoid arthritis can take up to three. In addition, these studies only look at official diagnostic reversals that appear in medical records, failing to capture the myriad unofficial ways people’s symptoms are dismissed as psychogenic by health care practitioners.

Then there is the question of harm. Feeling “crazy,” as Lisa O’Brien did when her long COVID symptoms were dismissed, is incredibly damaging. A physiological misdiagnosis—say, confusing epilepsy with multiple sclerosis—can result in physical harm. But a psychological misdiagnosis can cause patients to question their own sanity and blame themselves for their pain. 

Harm can also result from how others perceive you. The science journalist Julie Rehmeyer described the despair she felt upon hearing the first reports of long haulers. As an ME/CFS sufferer herself, she knew what was coming.

“They would face skepticism from friends and family, not everyone, but as a general pattern. I knew that when they described their situation to healthy people, healthy people by and large wouldn’t get it,” Rehmeyer told me. “The hardest part is the feeling of having fallen off the edge of the earth. All the usual ways you expect the world to work just stop working. You can’t describe your experience and have people relate. You feel like you’ve lost your citizenship in the world and now you’re a refugee.”

Physiological harms that result from physiological misdiagnosis are well documented. The psychosocial harms that result from psychogenic misdiagnoses are not, and they need to be. 

Meanwhile, allocation of research resources remains too low for conditions that have a history of being dismissed as psychological. We need to increase our research of chronic post-viral syndromes, chronic fatigue syndrome, and related conditions that are poorly understood such as endometriosis and postural orthostatic tachycardia syndrome (POTS).

Patient advocates are already using the plight of long haulers to push for more ME/CFS funding from the NIH. Figures have not improved very much since the 90s when Osler’s Web was written. In 2019 the (potentially inflated) amount was $15 million—just $15 million to study a debilitating condition that afflicts as many as 2.5 million Americans. 

By comparison, $13 million was allocated to Charcot-Marie-Tooth disease, a rare nerve disorder affecting 126,000 Americans—the same amount, incidentally, that was spent on endometriosis research, which affects 2 to 10 percent of all women, and is also often misdiagnosed as psychological. Observing the pattern of lopsided funding it becomes immediately apparent that sexism plays a major role, an egregious ethical lapse that has been amply documented without being rectified.

However, research is only half of the problem. Increased scientific knowledge does not address one of the major causes of misdiagnosis: overworked primary care practitioners who don’t have enough time to spend on patients with complicated, long-term conditions.

One striking feature of Jen Brea’s documentary about her struggles with chronic illness is the amount she had to learn. “For people who are chronically ill there’s this idea that the primary care practitioner is the quarterback of your team,” said Brea. “But in reality, for people with complex cases, you’re the quarterback or there isn’t one.” 

When your disease fits into a clear specialty the quarterback becomes your specialist. In the case of cancer it’s your oncologist, for arthritis your rheumatologist. But when there’s no specialty for your disease, you get bounced from one specialist to the next, each focusing only on their area. Without Brea’s encyclopedic knowledge of multiple medical specialties there would have been no one to synthesize all the information she was getting from specialists and her own research. It is almost certain she never would have found the neurosurgeon who repaired her tethered spinal cord.

Gastroenterologist Allen Hwang has seen this in his own practice. “The patients who arrive at their diagnoses most quickly and accurately are advocates for themselves or have a PCP who is an advocate for them,” he said, recognizing that quality medical care all too often requires the privilege of having time and money to secure that diagnosis. “But in the American healthcare system the specialist has taken the front seat—patients have a cardiologist and a pulmonologist and the primary care doctor is bogged down.”

Virtually every patient and physician I spoke to told me something similar. “The sad trend in America is that we do not have enough primary care physicians for those who need them, and the way physicians are reimbursed for their work is political in nature,” said Kate Chebly, the chief resident of NYU’s primary care program. “This is a product of the constraints of the current healthcare system.”

The failure of that system forces patients with complex medical conditions into the welcoming arms of charlatans whose main virtue is the time they spend listening. They already prey on ME/CFS patients, and COVID long haulers who do not receive adequate care and support will be next. “You’re going to have people that pop up, it’s going to be the fringe, naturopaths or chiropractors, they’ll either see dollar signs, cha-ching, or they’ll say that’s my wheelhouse, that’s what I deal with,” said Clay Jones, a pediatrician who writes about pseudoscience for the popular blog Science-Based Medicine. He worries that costly, unproven therapies will be sold to a vulnerable population for whom hope and validation is in short supply.

Lastly, there is the issue of medical education. Given that medically unexplained symptoms account for at least a quarter of all clinical encounters, it is noteworthy that not a single physician I spoke with reported significant education about them or training in how to deal with them. 

“Recognizing personality disorders and how they operate takes a lot of training and supervision that most doctors can’t do,” said Dr. Emily Deans, a private-practice psychiatrist affiliated who lectures at Brigham and Woman’s Hospital in Massachusetts. 

I asked her about how she was able to distinguish between someone suffering from a psychiatric condition and someone with an underlying physiological problem. “It’s hard for me to say,” she said. “I can spot that in four minutes but that’s what I do. Some in primary care and family are naturally good at this, and some people aren’t, and that frustrates them and they start labeling everybody.”

Although it is hard, there are certain features of chronic fatigue syndrome that make it easy to distinguish from anxiety or depression. The term fatigue is actually deceptively vague, confusing clinicians and patients alike. This is why official medical descriptions of ME/CFS have switched to “post-exertional malaise.” While fatigue could describe a range of feelings including general tiredness, post-exertional malaise refers specifically to a crash triggered by physical exertion, which Body Politic, a patient-led research group for COVID-19, has identified as a key symptom experienced by long haulers in a preprint study.

Likewise, there are important differences between people who are inactive because of depression and those who are inactive because of chronic fatigue. “There’s definitely crossover between depression and chronic fatigue,” said David Tuller, the journalist and public health expert whose work on the PACE trial for ME/CFS helped bring to light its flaws. “But if you ask someone with depression, ‘What would you do after your depression?’ they’d say ‘Oh, I don’t, know.’ Ask someone with CFS and they’ll say go on a hike, get back to my job!”

And when physicians have run all the tests, consulted all of their medical knowledge, and still can’t tell what’s going on? Instead of jumping from “we can’t see it on the tests” to “it’s all in your head,” there must be more emphasis on the importance—the duty—of saying “I don’t know.” 

Or, as Diane O’Leary put it: “The best path to improving care in this area is truthful humility about diagnostic uncertainty.” Chronic medical conditions like long COVID may not be preventable, but chronic amnesia about their existence and neglect of their causes certainly is.

Follow Alan Levinovitz on Twitter.