By David Tuller, DrPH
The Switch is a New Zealand variant of the Lighting Process, the so-called “mind-body” intervention created by a Phil Parker, a British former faith healer who once declared that he had “this ability to step into other people’s bodies… to assist them in their healing with amazing results.” I don’t know whether Mel Abbott, founder of The Switch, possesses this unusual skill. Perhaps not. But she makes big promises for The Switch—as Parker does for the Lightning Process.
The website for The Switch includes this tagline in small capital letters: “CHRONIC ILLNESS RECOVERY IS MY SPECIALITY!” The intervention is said to treat the following list of conditions: anxiety, depression, chronic fatigue, Long COVID, chronic pain, chronic regional pain syndrome, autoimmune disorders, digestive disorders, neurological disorders, insomnia, allergies and intolerances, eating and weight issues, low self-esteem, and reproductive and sexual issues.” Wow!
Like Parker, Abbott is making an effort to get her program accepted as a viable medical intervention. Most recently, she and a colleague published a paper called “Initial evidence for mind body retraining intervention ‘The Switch’ for those with idiopathic chronic fatigue” in the journal Fatigue: Biomedicine, Health & Behavior. Suffice it to say that the results are largely uninterpretable and meaningless.
Abbott’s bio indicates that she has a bachelor of science degree in psychology and is trained in neurolinguistic programming (NLP), which is generally regarded as pseudoscience. The paper describes The Switch in language similar to that used for other mind-body interventions:
“The Switch is a structured 4-day small group programme whose key premise is that chronic illnesses occur due to widespread dysregulation of the nervous, immune, digestive, and hormone systems, brought on by states of high and prolonged physiological stress. Before the body can heal, this dysregulation must be calmed, and body system functions restored. The Switch uses education, self-reflection, and structured techniques to help participants understand the wider framework of their illness, recognise and resolve the original stress, and modify current thinking, behaviour and belief patterns that might maintain an illness.”
The paper notes that up to 30% of the content of The Switch overlaps with the LP. The four sessions, each lasting six hours, include elements of, among other approaches, NLP, cognitive behavior therapy, and “inner child work” to resolve past trauma and “to clear long-held beliefs that are impacting health.”
The corresponding author is Suzanne Barker-Collo, who produced the paper based on data collected and provided to her by Abbott. Barker-Collo is a psychology professor at the University of Auckland. According to the paper, Abbott “had no input into the coding of the data, conducting the statistical analyses, or the interpretation.”
The research is described as a “feasibility” study designed to explore the possibility of conducting a randomized clinical trial (RCT)—the so-called “gold-standard” level of evidence. In the discussion section, Barker-Collo introduces some welcome cautionary notes. But she also—bizarrely—suggests that the data from this study are sufficiently robust to recommend The Switch even without such bothersome complications as having to conduct a bonafide RCT.
The “research,” such as it is, suffers from a host of weirdnesses.
For one, it’s unclear who the participants are. They are identified as “individuals reporting idiopathic chronic fatigue (self-reported diagnosis of Myalgic encephalomyelitis (ME)/chronic fatigue syndrome (CFS) or long COVID).” Given that self-diagnoses are not considered reliable, did the participants have idiopathic chronic fatigue or did they have ME/CFS or Long COVID? Who knows? (Certainly not the researchers.) Dumping them all together in one bucket makes it hard to tell what is being studied and whether the findings can be extrapolated to any other populations.
The study is not a clinical trial, and there is no comparison arm. That means the findings represent associations or correlations—not causal relationships. All of the 104 participants had already enrolled in The Switch before signing on to the study; in fact, having completed at least one of the four six-hour sessions was one of the entry criteria. According to the procedures for doing The Switch outlined in the paper, that means the participants had been pre-interviewed by Abbott to ensure that they were in an appropriate state of “readiness for recovery.”
**********
What does “readiness for recovery” mean?
The paper does not explain what Abbott means by “readiness for recovery” or how she determines if this quality is present. This procedure appears similar to what LP participants have reported—that they must agree in advance that the intervention works as advertised. Under such circumstances, findings cannot be extrapolated to anyone who does not first affirm that The Switch is as effective as claimed.
Requiring people to agree beforehand that an intervention can cure them clearly generous enormous issues in interpreting any findings. It would be hard, if not impossible, to measure the level of bias that such a requirement would introduce.
The outcomes were all subjective and self-reported—making it impossible to determine whether positive results had anything to do with the intervention or were generated solely by the huge amount of bias created by hopes, expectations, therapeutic allegiance, and other factors. This is a routine problem in so-called “biopsychosocial” studies of ME/CFS, Long COVID, and related illnesses, given the challenges in blinding these interventions.
The main instrument was the SF-12, a “short form” health survey with 12 questions condensed down from the 36 questions in the full version of the questionnaire. Like the longer version, the survey produces eight domain scores: general health, physical functioning, role-physical, pain, role-emotional, emotional well-being, energy/vitality, and social functioning.
However, four of the domains in the SF-12 are based on a single item, and the other four are based on only two items. It is therefore generally advised that the best use of the SF-12 is to calculate two separate composite scores rather than relying on eight individual domain scores. The Switch paper ignores that advice and bases its conclusions on the eight separate scores. This decision is not explained and is hard to justify, given the known limitations of the SF-12.
The study also includes responses to visual analogue scales for both physical and mental health. These too are subjective and so should be treated cautiously, given that everyone knew they were taking an intervention that would purportedly lead to recovery.
Despite the multitude of statistical tests conducted in the various analyses, the paper did not feature a standard strategy to minimize the risk that some findings might have occurred by chance. In these cases, researchers generally include an statistical adjustment for multiple comparisons to try to rule out positive findings that might have occurred by chance.
Barker-Collo recognizes some of the study’s enormous limitations. The paper includes a host of appropriate and necessary caveats:
“It was not possible to check the veracity of the source data for bias. A further limitation is that diagnosis was self-reported. While only seven (6.7%) participants did not specifically mention seeing a doctor for diagnosis, the diagnoses here were self-reported and therefore the participants were referred to as having idiopathic chronic fatigue rather than ascribing a specific diagnosis. Diagnoses would need to be medically verified in any future trials. Any future examinations of The Switch for this population should ideally include not only verification from existing records but also recognized formal assessments to verify the fatigue diagnosis (e.g. Institute of Medicine)…In addition, the sample was self-selected, and it must also be noted that only one therapist (MA) provided the intervention so the findings could reflect a clinician effect.”
The paper also noted that future studies should “include not only a control condition to tease out placebo and practice effects, but to include objective assessments of outcomes (e.g. 6-min walk test or actigraphy).” Good idea!
Unfortunately, in this field, such objective measures almost always fail to support the much rosier results on subjective measures. The leading experts have routinely dismissed objective measures that fail to match subjective responses or, in more recent studies, have decided to not include any objective measures at all. Much better not to have pesky objective measures out there if they might contradict your subjective ones!
Nonetheless, despite having outlined all of these shortcomings, here is Barker-Collo’s conclusion: “However, while this is a one therapist trial and some effect could be attributed to placebo, it could be argued that the benefits are sufficiently high to not require an RCT.”
I suppose that “could be argued,” as the paper notes. But, based on these data, it would be a stupid and unconvincing argument that fails to acknowledge the obvious—these self-reported findings are so fraught with bias and ambiguity that they really mean nothing at all.
