By David Tuller, DrPH
The Lancet has just published an anguished whine de coeur from supporters of the graded exercise therapy/cognitive behavior therapy/ [GET/CBT] approach to ME/CFS. (Or CFS/ME, as these authors insist on calling the illness in what those familiar with the debate will recognize as a childish fit of pique.) These impassioned members of the GET/CBT ideological brigades, all from northern European countries, are metaphorically stomping their little feet over the new ME/CFS clinical guideline from the UK’s National Institute for Health and Care Excellence.
Per the new guideline, which was published in October, NICE no longer recommends GET and only advises CBT as supportive care—not as a treatment for the illness itself. The US Centers for Disease Control and Prevention made a similar decision several years ago and removed from its site multiple references to the infamous PACE trial—a key plank of support for the GET/CBT approach. During NICE’s own extensive review process, it determined that the quality of the evidence to support the effectiveness of these interventions, including from the PACE trial, was either “very low” or merely “low.”
The five authors of this Lancet comment are (in order of authorship) Signe Flottorp, Kjetil Brurberg, Per Fink, Hans Knoop, and Vegard Wyller. Flottorp, Brurberg, and Wyller are all Norwegian; Fink is Danish, and Knoop is Dutch. All are well-known players in this field. Like their fellow GET/CBT brigadiers in the UK and elsewhere, they have made some dubious statements when it comes to ME/CFS. They really don’t like me.
(A related aside: In a 2020 letter recently posted on Twitter, Flottorp described me as an “ME activist.” I don’t personally consider the word “activist” as dismissive unless the speaker or writer is using it in a dismissive manner, as Flottorp did in her letter. Whatever my “activist” status, I am a public health academic at one of the world’s great research institutions–and it is inappropriate for that qualification to be omitted when my work is being discussed in a professional context. In reality, I am an “activist” in terms of advocating for my views just as Flottorp and her colleagues are “activists” in their advocacy for their own.)
In their Lancet comment, the authors raise specious arguments in asserting that NICE got things wrong. First, they seem baffled at the very notion that guidelines might change, even after more than a decade. They appear to believe that re-affirming an existing guideline is the main point of reassessing the evidence base, as when they write this:
“Although we applaud guideline efforts, it is remarkable that recommendations in the 2021 NICE guideline differ substantially from the previous one, and do not include CBT and GET as means to treat CFS/ME. The new guideline presents strengthened evidence, but a major shift in interpretation. How could this happen?”
Of course, the goal of a guideline revision is to evaluate the available data and reach updated conclusions—not just to parrot old ones. The complaint here seems to be that since 2007, when the previous guideline was published, some “strengthened evidence” for the CBT/GET approach has emerged, and that the committee has inexplicably rejected these findings.
For this lot, the release of the main results of the PACE trial in 2011 in this very journal was a major event—a key part of that purported “strengthened evidence.” The authors cite the PACE trial multiple times–as if it still retains widespread credibility beyond its die-hard defenders. They don’t cite the 2018 paper in BMC Psychology that effectively refuted the PACE findings. (I was a co-author of that paper.)
If NICE is wrong in its evaluation of PACE, then so were the 100+ experts from Harvard, Columbia, University College London, Stanford, Yale, Berkeley, etc., who signed an open letter to The Lancet lambasting the trial. The letter noted PACE’s “unacceptable methodological lapses” and asked for an independent investigation. At Berkeley, my epidemiology colleagues have used PACE in seminars as a pedagogical tool–a case study in how NOT to conduct a clinical trial. In short, it is a methodological and ethical disaster.
So Flottorp and her colleagues are free to cite the PACE trial as often as they want. Unfortunately for them, their reliance on it undermines their position. It also raises questions about their scientific judgment and acumen.
The Lancet comment authors argue that the NICE guideline is internally inconsistent because it defines ME/CFS by self-reported symptoms but then downgrades trials with self-reported outcomes. The most valid way to assess the results of interventions, they insist, is by outcomes assessing the same self-reported symptoms—not by objective markers.
This is sophistry. The PACE investigators have mounted the same silly point. If this were so, why did they bother to include four objective outcomes in their so-called “definitive” study of GET and CBT? In the end, they rejected the objectivity of their own objective outcomes only after these results failed to match the positive reports from their subjective measures.
One of the comment’s co-authors, Hans Knoop, has himself used an objective measure of activity–actigraphy, in which participants wear a sensor that monitors their movement over a period of time. In some studies of psycho-behavioral treatments for ME/CFS, this measure has yielded null results despite positive reports on the subjective measures. In a recent interview about a controversial ongoing trial of CBT in long Covid, Knoop suggested that objective outcomes in these sorts of studies are deficient because the results do not support the positive reports from subjective measures.
Here’s what he stated: “We have only subjective primary and secondary outcome parameters [in the current study]…In previous research, we have shown that there is no correlation between the objective and subjective outcome parameters.” Abandoning objective outcomes because they don’t “correlate” with subjective ones is an odd methodological strategy. In the absence of evidence that the objective outcomes are deficient in ways that go beyond their non-concordance with subjective ones, it is a hard strategy to defend.
Moreover, the concern about subjective measures is not about subjective measures per se. The concern is specifically about the bias inherent in unblinded trials that rely solely on subjective outcomes for reports of success. When used in blinded studies or in combination with objective outcomes, subjective measures can obviously have an important role.
The comment authors also express surprise that some NICE committee members had “negative opinions regarding the interventions considered,” citing a thread on the online forum Science For ME (of which I’m a member). Apparently they are blind to the irony in this statement. For many years, GET/CBT proponents have held strong and repeatedly expressed positive opinions “regarding the interventions considered.” The 21-member NICE committee included many such proponents. While three members decided to remove themselves from the committee near the end, most did not—and presumably those that remained, including but not limited to the five patient representatives, supported the final version of the guideline.
The comment authors appear to believe that a minority on the NICE committee managed to browbeat professors and highly credentialed clinicians to ignore viable evidence and agree to unjustified recommendations. Most likely, the arguments offered to support the CBT/GET approach were along the unconvincing lines as those outlined here and were therefore properly rejected. The authors repeatedly reference submissions made to NICE during the guideline consultation process. Presumably the NICE committee’s overall judgment was that these submissions were not persuasive.
The fact that NICE did not adopt the authors’ perspective on the matters under dispute does not mean that NICE ignored these claims. Rather, NICE has rejected them.
The authors conclude that the various alleged “shortcomings in the guideline process suggest that it was not driven by science but by ideology—that is, socially determined perceptions of reality characterised by, for example, insufficient corrective effects from experiences and inherent defence mechanisms against criticism.” This statement demonstrates an impressive lack of self-awareness. Can I point out again that irony is not a strong suit among this group?
The entire article reads like collective psychological projection. The authors accuse others of doing what they themselves have done. Cherry-picking from the literature? Bingo! Favoring ideology over science? Absolutely! Their position is hypocritical and untenable. They are on the wrong side of a crumbling paradigm, and they sense it, and in this tendentious document they come across as dismayed, confused, and even scared.