By David Tuller, DrPH
A new piece in The Conversation shows just how problematic it is when poorly done biopsychosocial studies claim to have documented that cognitive and/or behavioral therapies are effective—and when these questionable findings are published in high-impact journals. The headline of the article: “Success in treating persistent pain now offers hope for those with Long COVID.”
My response to that: “No it doesn’t.”
The Conversation is a venue in which academics and experts present information designed to be consumed by a lay audience. The co-authors of this piece are a GP and an orthopedic surgeon, both at the University of Otago, in New Zealand. They see no reason why insights into the neuroscience of pain should not be applied wholesale to the realm of “fatigue”—by which they apparently mean Long Covid and, along with that, the illness they call “CFS/ME.” (What is the current standard nomenclature in New Zealand?)
One excellent reason to challenge their arguments is the low quality of the research cited.
As they write, “Research has shown how appropriately trained general practitioners can provide explanations that aid recovery for a wide variety of persistent symptoms, including fatigue and pain.” The link for this claim is a study recently published by The Lancet of an intervention for a heterogeneous grab-bag of people with persistent physical symptoms (PPS). The lead author, Chris Burton, is a professor of primary medical care at the University of Sheffield. A long-time biopsychosocialist, he served on the committee that developed the new ME/CFS guidelines for the National Institute for Health and Care Excellence (NICE).
The NICE guidelines, which were issued in 2021, rescinded the recommendation for graded exercise therapy and cognitive behavior therapy as curative treatments. Even though this study is about PPS and not specifically ME/CFS, its approach nonetheless appears to undermine at least the spirit of the guidelines.
As is so often the case with these studies, this one was unblinded and relied solely on subjective outcomes. Because such studies are likely to produce an unknown amount of bias, the findings are generally hard to interpret or to take seriously as meaning much of anything—beyond, that is, proving that unblinded studies relying on subjective outcomes tend to produce at least modestly positive results. As Professor Burton well knows, NICE downgraded all such studies in its ratings of quality. The same judgement should be applied to Professor Burton’s new trial.
The intervention involved up to four sessions with GPs, who provided patients with “rational explanations” for their symptoms. These explanations were not identified, but presumably they focused on psychological, behavioral and related factors rather than “organic” causes, since the investigators specifically excluded patients with “organic” conditions from their study cohort.
The primary outcome was an instrument measuring the severity of somatic symptoms, the PHQ-15. All participants reported symptoms of moderate severity at the start of the trial. But the benefits for the primary outcome, while statistically significant, fell below the threshold considered to be the minimal clinically important difference (MCID) for the measure. In other words, the improvement was so marginal that it would not have been expected to be noticed or experienced by patients—and this was in a study already designed to generate positive bias.
And yet the investigators—and the journal—treated this finding as a success. As for the fact that the results for the primary outcome were not clinically significant, the investigators relegated this highly salient point to the deep bowels of the text. They also made efforts to downplay this very embarrassing detail by suggesting it didn’t matter much because at least some people seemed to have been helped by the intervention.
And now The Conversation has further disseminated their claims without qualification. Did the co-authors of the piece read Professor Burton’s study, or did they just review the headline or abstract? Did they know that the reported benefits did not meet the MCID and therefore had no practical impact on patient health? And if they did know, would they have cared or thought it was a problem worth highlighting?
********
Long Covid and ME/CFS are not the same as “persistent fatigue”
And here’s more: “Recent insights from the burgeoning field of neuroscience now guide clinical management of chronic pain and may offer hope for people living with persistent fatigue.”
The word “may” is doing a lot of work in that sentence. In fact, this perspective is fraught with problems that render it suspect from the get-go.
For one thing, Long Covid is not “fatigue,” and it is not “persistent fatigue.” Nor is ME/CFS. The U.S. Centers for Disease Control inflicted huge damage on the patient community in 1988 when it dubbed the illness investigated in Lake Tahoe “the chronic fatigue syndrome.” Profound and—yes—persistent fatigue is a common experience among patients. But placing that word front and center ends up turning a distinct clinical entity into a single symptom.
The core characteristic of the illness is post-exertional malaise—or what is called post-exertional neuroimmune exhaustion (PENE) in the International Consensus Criteria for ME. The biopsychosocial crowd doesn’t seem to quite know what to make of PEM. They prefer to overlook the many scientific findings indicating that it involves pathophysiological responses to activity. That helps them believe it is amenable to a strategy focused on “desensitization,” as if it were some sort of phobia.
After presenting their theory that advances in pain neuroscience could help people with Long Covid, the investigators suggest this is already proving to be so. Unfortunately, the Dutch study they cite as evidence, published by the journal Clinical Infectious Diseases, is yet another in which the investigators claimed success on questionable grounds. The trial investigated the use of cognitive behavior therapy to address prolonged fatigue after acute cases of Covid-19. Like Professor Burton’s study, it was unblinded and relied on subjective outcomes for its reports of success. As expected with this design, the subjective findings were modestly positive. However, the investigators failed to report that their one objective outcome—how far people moved, as measured objectively by actigraphy—had null results.
The failure of this one objective outcome to match the modestly positive subjective results, as well as the decision by the investigators to engage in apparent research misconduct by not reporting those results in the trial paper, mean it is hard to take the study at face value. As I wrote at the time, the investigators offered “dog-ate-my-data” excuses on social media and in correspondence for not reporting these objective data. (Other posts on this Dutch study are here and here.)
Oh, and did I mention that The Conversation piece also highlights the self-serving screed from the so-called “Oslo Chronic Fatigue Consortium,” which includes key proponents of the Lightning Process, the woo-woo program developed by self-proclaimed faith healer Phil Parker?
In other words, the two New Zealand “experts” who wrote this piece for The Conversation either don’t understand the limitations of the literature, or they don’t care. Whichever it is, nothing they say here is worth paying attention to.