By David Tuller, DrPH
I have slammed a recent Dutch study, Kuut et al, that investigated CBT for fatigue after an acute bout of COVID-19. The study, “Efficacy of Cognitive-Behavioral Therapy Targeting Severe Fatigue Following Coronavirus Disease 2019: Results of a Randomized Controlled Trial,” nicknamed ReCOVer, was published in Clinical Infectious Diseases, a well-regarded journal, and proclaimed itself a success. The intervention, called Fit after COVID, was based on the premise that dysfunctional beliefs about illness and problematic behavior—that is, remaining sedentary because of a purportedly irrational fear of activity and related cognitions—prevent patients from getting better. (I have written about this study here, here, and here.)
Among other flaws, the unblinded study relied solely on subjective outcomes for its claims that CBT is efficacious—a set-up guaranteed to generate an unknown amount of response bias. In a disturbing omission, the study did not mention that the one objective measure—how much people moved, as assessed by actigraphy—yielded null results. In my view, this disturbing omission is a clear example of research misconduct, and arguably qualifies the study as fraudulent. All standard ethics guidelines include provisions barring researchers from withholding salient data that would undermine the reported conclusions.
Such provisions are obviously relevant to this case. Yet the investigators advanced ridiculous justifications for their problematic decision—what I referred to in a previous post as “dog-ate-my-data” excuses. The most laughable was the argument that levels of physical activity are completely unrelated to fatigue. Huh?? As I previously reported, this self-serving argument has been contradicted in previous research by none other than Professor Hans Knoop, the senior author of Kuut et al. It is astonishing that leading investigators would stoop to this level of double-talk—or it would be astonishing if Professor Knoop and his colleagues hadn’t pulled similar stunts in previous research.
I engaged in a spirited exchange on this issue with Professor Chantal Rovers, a physician and infectious disease expert as well as a co-author of the study, on X, the social media platform formerly known as Twitter. I wrote that the arguments in support of witchcraft in Arthur Miller’s classic The Crucible, which I’d just seen in a terrific London production, were more convincing than the explanations offered for the decision to not include the null objective findings in the paper. (I don’t think she appreciated that comparison.)
Now an interesting split has emerged between some of the researchers themselves. In an interview published online, Tanja Kuut, the study’s lead author, stated the obvious in noting the following: “We think that behavior and views prevent some people from recovering.” The senior author, Professor Hans Knoop, was also interviewed and uttered related musings.
Nothing about these statements could be considered remotely unexpected, given the study itself and previous research and pronouncements from Professor Knoop. Yet Professor Rovers, a long-time colleague of Professor Knoop, has now professed surprise that her co-authors have made this public declaration. When confronted on X with the argument promoted by her co-authors, Professor Rovers responded thus: “I didn’t know this part. I certainly do not support the comment ‘we think that behavior and views prevent some people from recovering’. This was not looked at at all in this study and we even included a sentence about it in the discussion in the article.” (She wrote this in Dutch; the translation is X’s, so I cannot account for its accuracy.)
It is commendable that Professor Rovers is publicly contradicting bullshit. At the same time, it is hard—for me, at least—to take her statement with anything other than a massive grain of salt. The entire premise of the CBT/GET approach to ME/CFS, and now long Covid, is that patients harbor unhelpful beliefs about their illness and pursue strategies that keep them sick—in other words, they don’t engage in enough activity and remain severely deconditioned. The PACE trial itself posited that all the symptoms could be interpreted as being caused by these factors. So it is certainly perplexing that Professor Rovers now insists she had no idea about this frame for the study to which she has contributed. Her response echoes Casablanca’s crafty Captain Renault, played by the brilliant Claude Rains, who was “shocked, shocked” to find gambling taking place at Rick’s Café.
It is simply not true that “this was not looked at at all in this study.” Of course it was! This hypothesis is the basis of the CBT/GET approach to ME/CFS over the last three decades. And it is the basis of the effort to deploy this strategy for long Covid. The intervention included modules specifically addressing these purportedly counterproductive factors. Why did Fit after COVID target supposedly problematic thoughts and sedentary behavior if these are not hypothesized as being implicated in preventing patients from getting better? I mean, what did Professor Rovers think they were studying? It’s a mystery.
In countering that the study included language essentially rebutting her colleagues’ statements, Professor Rovers was presumably referring to this passage at the very end: “Of note, applying a cognitive-behavioral approach to the treatment of post-COVID-19 fatigue neither implies that its cause is psychological, nor does it negate a possible somatic cause. We encourage research into its underlying (neuro)biological mechanisms.”
It should be noted that the discredited PACE trial also included a throwaway sentence suggesting that the biopsychosocial approach it embodied should not be taken to mean the illness is psychological. Such concessions are essentially meaningless when a study is clearly designed to demonstrate that psycho-behavioral interventions can lead to recovery. If that was not the intent, then the authors need to explain why they dubbed their study “ReCOVer.” In this context, the call for research into biological mechanisms comes across largely as a sop meant to satisfy critics, not as a genuine appeal. Perhaps Professor Rovers was indeed sincere on this point; if so, she has chosen to collaborate with investigators who do not appear to share her views.
I have heard from more than one person that Professor Rovers is a sympathetic clinician who genuinely cares for her patients. I have no reason to question that. But if that is the case, and if she truly disagrees agree with the recent remarks from Dr Kuut and Professor Knoop, she needs to make a much more forthright public statement of protest than a mild pushback in response to a question on a dying social media platform like X. Beyond that, she also should acknowledge that the failure to publish the null objective findings was a flagrant misstep rather than parroting preposterous justifications for this serious ethical and methodological lapse.
(Originally posted on Virology Blog.)