Paradigm Lost: Lessons For Long COVID-19 From A Changing Approach To Chronic Fatigue Syndrome
For decades, patients with chronic fatigue syndrome, the poorly understood illness also known as myalgic encephalomyelitis/chronic fatigue syndrome or ME/CFS, have experienced neglect from the medical establishment and dismissal from society at large. In recent months, however, the condition has gained remarkable visibility, since many survivors of COVID-19 are reporting the sort of profound exhaustion, cognitive impairment, orthostatic intolerance, recurrent relapses, and other medical complaints that characterize ME/CFS. The New York Times, The Guardian, the BBC’s “Newsnight,” and other major news outlets have extensively covered the apparent overlap between ME/CFS and what has been dubbed Long COVID or, more formally, post-acute COVID-19 syndrome.
As with Long COVID, a large subset of ME/CFS patients reports that the condition began with an acute viral infection from which they never seemed to recover. Many remain severely impaired for years, even decades. Anthony Fauci, the country’s top infectious disease expert, has drawn attention to the similarities, observing that the pattern of Long COVID symptoms is “highly suggestive of myalgic encephalomyelitis and chronic fatigue syndrome.” It is still unknown what proportion of Long COVID patients will ultimately recover on their own. Despite the fact that periods of post-viral fatigue and other symptoms are actually quite common after a variety of acute infections, the phenomenon is surprisingly understudied.
Because Long COVID research is in its infancy, the current rehabilitation strategies being offered to patients are, understandably, often based on assumptions and theories drawing from analogous contexts rather than from specific insights into the new syndrome itself. Among the theories being debated is a notion that long dominated research and treatment in the field of ME/CFS—that patients’ persistent symptoms are driven by their “unhelpful cognitions.” In believing they have an ongoing disease, the theory posits, patients avoid exertion out of fear of getting worse, and thus engage in a counter-productive pattern of sedentary behaviour that renders them severely deconditioned. Therefore, the argument goes, patients merely need either a program of graded exercise therapy (GET) to help them regain physical strength or cognitive behavior therapy (CBT) to correct their false beliefs—or quite typically a combination of the two.
This therapeutic logic, deeply entrenched for decades, has lately been seriously challenged. In 2015, following an assessment of thousands of studies, a report from the Institute of Medicine (now the National Academy of Medicine) declared ME/CFS to be a complex physiological disease and called for research to focus on biological pathways, with no mention of “unhelpful cognitions” as a cause. More recently, the evidence base for the GET/CBT approach has faced high-profile critiques, including from the authors of this post, that have drawn attention to serious methodological flaws.
Processes of change in medicine and other disciplines, during which the authority of die-hard practitioners is supplanted by new insights, are frequently referred to as “paradigm shifts,” in reference to Thomas Kuhn’s classic analysis of scientific revolutions. By the time the coronavirus pandemic arrived last year, a paradigm shift in the ME/CFS field was already well underway, with the psychogenic view having steadily lost ground in the debate.
Nonetheless, proponents of the long-standing paradigm have appeared undeterred. They have seized upon the comparisons between ME/CFS and the current reports of persistent post-viral symptoms to quickly push exercise and psychotherapy as go-to treatments for people experiencing Long COVID. A recent Wall Street Journal opinion piece, for example, presented Long COVID as essentially a mental illness—and cited deeply flawed ME/CFS research as evidence.
Analysis, Reanalysis
The GET/CBT framework for ME/CFS first emerged in Britain, where it has retained the most support among medical professionals. Last November, however, the approach was dealt a serious blow when a major public health agency, the National Institute for Health and Care Excellence (NICE), concluded that the empirical evidence for psychological and behavioral therapies for ME/CFS was weak to non-existent. This conclusion was rendered through the publication of a draft of new clinical guidelines for ME/CFS, which were developed under NICE’s purview and based on an extensive review of the research.
The previous NICE guidelines, issued in 2007, had enshrined the two rehabilitative interventions—GET and a bespoke form of CBT—as best practice in the treatment of ME/CFS. That view was reinforced in 2011, when The Lancet published the first results from a major study known as the PACE trial. The reported findings of efficacy helped support claims that the use of both GET and CBT for ME/CFS was “evidence-based.” Britain’s National Health Service structured care provision through a countrywide network of dedicated ME/CFS clinics. In the US, the Centers for Disease Control and Prevention as well as major medical organizations, such as the Mayo Clinic, made similar recommendations, guided by the PACE results and related research.
But many patients who underwent GET or CBT found the treatments useless or even harmful. Those with some understanding of research design began raising serious reservations about the PACE study and were eventually joined by concerned scientists and academics from around the world. Among multiple issues, the apparent treatment efficacy was linked to self-reported data only, with objective data on physical function and fitness showing no meaningful improvement. Moreover, the trial was not only unblinded, but participants in the GET and CBT arms were informed that the interventions had already been shown to be effective, thereby maximizing the risk of confirmation bias effects. Most damningly, the PACE authors had chosen to weaken their criteria for improvement and recovery after collecting their data.
In response to these legitimate methodological concerns, the PACE authors and their allies portrayed the critics as vexatious and potentially dangerous extremists. Major medical journals, public health agencies, and mainstream news outlets leaned into the establishment narrative, defaulting to the study authors’ purported expertise and impartiality. The one-sided media account of this conflict was a vivid demonstration of how, in assessing the merit of competing medical assertions, status can often be accorded a more important role than the science itself—a realpolitik that might be considered a form of “eminence-based medicine.” But in 2018, when independent investigators conducted a reanalysis of the PACE trial after the release of some previously unpublished data, they discovered that the claimed treatment effects had been greatly overstated and were largely illusory. In other words, as patients had long insisted, the GET and CBT interventions did not work as advertised.
For its 2020 review of the research for the GET/CBT approach, NICE rated the quality of evidence from dozens of studies, including the PACE trial, through a standardized methodology. This rigorous assessment found most of the evidence to be of “very low quality” and the rest of “low quality.” The most common methodological problem identified was reliance on a combination of unblinded study arms and subjectively assessed outcomes—a recipe for bias. As a result, the NICE draft asserts that any interventions offered to patients should not be framed as “a treatment or cure for ME/CFS,” and that CBT should only be offered as a way “to reduce psychological distress” for those patients who “would like” this type of support.
Beyond that, the draft notes widespread reports that GET appears to actually harm a large subset of ME/CFS patients. As it explains, a characteristic aspect of ME/CFS is a serious deterioration after minimal activity called “post-exertional malaise” or “post-exertional symptom exacerbation”—a phenomenon also reported by patients with Long COVID. The clear import of the new guidelines is that GET, in particular, could be contra-indicated and that the 2007 guidelines were not merely erroneous but potentially harmful for patients.
Sporadic And Imperfect Progress
Paradigm shifts are sometimes depicted as heroic events in which entire fields are transformed in brisk order by victories of scientific rigour over human frailty. But in Kuhn’s model, progress is rarely smooth-running, linear, inevitable, or occurring everywhere simultaneously. New paradigms are forged when observed anomalies create an unignorable crisis of confidence among the producers and consumers of science. The next generation of researchers—who are naturally more inclined to embrace new thinking and discard old ideas—will then gradually emerge as the new center of intellectual gravity.
Kuhn’s is a theory of sporadic and imperfect progress characterized by pragmatic consensus and professional jockeying between rival schools of thought and practice. With health care practice, the medical hierarchies in some regions may be early adopters of new clinical strategies while others can persist with older treatments that are already well-established, profitable, and easy to deliver within existing systems. Unexpected events—such as pandemics—can exert their own influence on these processes. And Kuhn’s model is cyclical: Paradigm shifts take science from one period of “normal” science to another, and then the process leading slowly to the next paradigm shift begins.
Any paradigm shift creates winners and losers. In Kuhnian terms, the losers are usually those who persist with a flawed model despite the emergence of empirical reasons to doubt it. The shifting of the ME/CFS paradigm significantly undercuts those prominent psychiatrists, psychologists, and other health professionals who chose to invest significant reputational capital, over multiple decades, in what is a now a collapsing theoretical construct. And so accordingly, we have entered the Kuhnian stage of crisis.
Confronted with the unpleasant prospect of NICE withdrawing its approval for their work—and no doubt alarmed at the implications these developments have for resource allocation, training courses, and other practicalities of service delivery in their domains—proponents of the dissolving paradigm immediately sought to strike back. On the same November day that NICE published its draft guidelines, a major science communications agency released prepared statements from a core group of prominent figures closely associated with the psychogenic approach to ME/CFS, including all three of the PACE trial’s lead investigators. These experts continued to describe GET and CBT as “evidence-based” treatments, despite the fact that NICE had just rated the purported underlying “evidence” as so low quality as to be rightfully ignored. NICE is expected to release a final version in August.
The impulse to defend the status quo reflects a recurring experience in the history of psychological medicine. At various times in the past, members of this discipline have similarly defended the “refrigerator mother” theory of autism, the use of tooth extractions to treat schizophrenia, and claims that AIDS is caused by depression rooted in “group fantasy” among gay men. Given this history, the use of press releases to rebut scientific documents such as the NICE draft and the accompanying evidence review is unlikely to prove convincing. The absence of peer review removes an important safeguard against self-serving bias, thereby facilitating the endorsement of claims that have already been authoritatively discredited.
While the NICE draft presents a significant milestone, the emergence of Long COVID has thrown a new controversy into the mix. Undaunted by NICE’s verdict on the state of the evidence, some proponents of the psychogenic view of ME/CFS have promoted what to them is a logical inference—namely, that symptoms of Long COVID resembling ME/CFS must be psychogenic too. Long COVID patients without identifiable organ damage continue to report that physicians are attributing their complaints to mental health factors, advising them to increase their exercise levels or seek psychotherapy.
The problem is that the use of these rehabilitative approaches for treating Long COVID has not been supported by enduring empirical research. And there is some reason to doubt their safety and effectiveness, given that a significant number of Long COVID patients are experiencing recurrent relapses and post-exertional malaise, the symptom characteristic of ME/CFS. The current lack of proven options does not justify the promotion of questionable treatments derived from low-quality studies, much less in a novel context.
In developing its draft guidelines for ME/CFS, NICE applied rigorous standards in assessing the evidence behind the various treatment options. Going forward, similarly rigorous standards should be applied when evaluating studies of treatments for patients experiencing Long COVID. For a condition that could potentially afflict millions of people, young and old, all over the world, therapeutic options should be approached very cautiously—and on the basis of carefully studied data rather than outmoded clinical lore.
Contrary to models that posit an inevitability to scientific progress, paradigm shifts are never guaranteed to succeed. If researchers and clinicians fail to learn from the example of the ME/CFS controversy, and if proponents of the psychogenic theory simply refit their preferred paradigm to a different and fast-evolving context, then the growing cohort of post-COVID-19 “long haulers” could be the latest casualties of resistance to this long-overdue revolution.
Authors’ Note
David Tuller’s position is supported by crowdfunding campaigns on the University of California, Berkeley’s in-house platform for campus projects. Myalgic encephalomyelitis/chronic fatigue syndrome patients and advocates donate to these campaigns. Stephen Lubet been living with with ME/CFS since 2006.
