A guest post by Dr. Keith Geraghty
Honorary Research Fellow at the University of Manchester, Centre for Primary Care, Division of Population Health and Health Services Research
The Advertising Standards Authority previously ruled that the Lightning Process (LP) should not be advertised as a treatment for CFS/ME. So how then, did LP end up getting tested as a treatment in a clinical trial involving adolescents with CFS/ME? Publication of the trial sparked controversy after it was claimed that LP, in addition to specialist medical care, out-performed specialist medical care alone. This blog attempts to shed light on just how a quack alternative online teaching programme, ended up in a costly clinical trial and discusses how the SMILE trial exemplifies all that is wrong with contemporary psycho-behavioural trials; that are clearly vulnerable to bias and spin.
The SMILE trial compared LP plus specialist medical care (SMC) to SMC alone (commonly a mix of cognitive behavioural therapy and graded exercise therapy). LP is a trademarked training programme created by Phil Parker from osteopathy, life coaching and neuro-linguistic programming. It costs over £600 and after assessment and telephone briefings, clients attend group sessions over three days. While there is much secrecy about what exactly these sessions involve, a cursory search online shows us that past clients were told to ‘block out all negative thoughts’ and to consider themselves well, not sick. A person with an illness is said to be ‘doing illness’ (LP spells doing as duing, to signify LP means more than just doing). LP appears to attempt to get a participant to ‘stop doing’ by blocking negative thoughts and making positive affirmations.
Leading psychologists have raised concerns. Professor James Coyne called LP “quackery” and said neuro-linguistic programming “…has been thoroughly debunked for its pseudoscience”. In an expert reaction to the SMILE trial for the Science Media Centre, Professor Dorothy Bishop of Oxford University stated: “the intervention that was assessed is commercial and associated with a number of warning signs. The Lightning Process appears based on neuro-linguistic programming, which, despite its scientific-sounding name, has long been recognised as pseudoscience“.
The first and most obvious question is why did the SMILE trial take place? Trial lead Professor Esther Crawley, who runs an NHS paediatric CFS/ME clinic, says she undertook the trial after many of her patients and their parents asked about LP. Patients with CFS/ME often report a lack of support from doctors and health care providers and some turn to the internet seeking help; some are drawn to try alternative approaches, such as LP. But is that justification enough for spending over £160,000 on testing LP on children? I think not. Should we test every quack approach peddled online: herbs, crystals, spiritual healing – particularly when funding in CFS/ME research is so limited currently? There must also be a compelling scientific plausibility to justify a trial. Simply wanting to see if something helps, does not merit adequate justification.
The SMILE trial has a fundamental design flaw. The trial compared specialist medical care alone (SMC) against SMC plus LP (SMC&LP). To the novice observer this may appear acceptable, but clinical trials are used to test item x against item y. For example, imagine trying to see which drug works better, drug A or drug B, you would not give drug A to one group and both drugs A and B to another group – yet this is exactly what happened in SMILE. In seeking to test LP, Prof. Crawley gave LP&SMC together – rendering any findings from this trial arm as pretty meaningless. The proper controls were missing. In addition, a trial of this magnitude would normally have a third arm, a do-nothing or usual care group, or another talk therapy control – yet such controls were missing.
Next we turn to the trial’s primary outcome measures. These were subjective self-reports of changes in physical function (using SF-36). Secondary outcomes were quality of life, anxiety and school attendance. These outcomes were assessed at 6 months with a follow-up at 12 months. It is reported that SMC+LP outperformed SMC alone on these measures at 6 and maintained at 12 months. However, there is no way to determine whether any claimed improvements came from LP alone, given LP was mixed with SMC. We could assume that LP+SMC meant more support, positive expectations and increased contact time. Here we see how farcical SMILE is as a trial. We have one group getting two treatments (possible double help) and one group getting one treatment (possible half help).
Of particular concern is how few of the available patients enrolled in and completed the trial: 637 children aged 12-18 attended screening or appointment at a specialist CFS/ME clinic; fewer than half (310) were deemed eligible; just 136 consented to receiving trial information and then only 100 were randomised (less than 1/3 of the eligible group). 49 had SMC and 51 had SMC+LP. Overall 207 patients either declined to participate or were not sufficiently interested to return the consent form. Were patients self-selecting? Were those less likely to respond to nonspecific factors choosing not to participate, and were we left with a group interested in LP – give Prof. Crawley said many patients asked about LP?
As the trial progressed, patients dropped out: of the 51 participants allocated to SMC+LP, only 39 received full SMC+LP. At 6-month assessment just 38 of the 48 allocated to SMC and 46 of the 51 in SMC+LP are fully recorded. At 12 months there are further losses to follow-up in both cohorts: 14% in LP and 24% in SMC. The reasons for participant loss are not fully clear, though the paper reports 5 adverse events (3 in the SMC+LP arm). It is worth noting that physical function at 6 months deteriorated in 9 participants (roughly 10% overall), 8 in the SMC arm, with 5 participants having a fall of ≤10 on the SF-36 physical function subscale (deemed not clinically important). Again questions are raised as to whether some degree of self-selection took place? The fact 3 of the participants assigned to SMC alone appear to have received LP reflects possible contamination of research cohorts that are meant to be kept apart.
Seven problems stand out in SMILE:
- The use of the SF-36 physical function test was questionable. This self-report instrument is not designed or adequately validated for use in children.
- Many of the participants appear to have had symptoms of anxiety and depression at the start of the trial. SMILE defined anxiety and depression as a score of ≥12 out of 22 on the self-report HADS. Usually a score of 8 or above is considered positive for mild anxiety and depression, and of above 12 for moderate anxiety and depression. The average mean HADS score at trial entry was 9.6 (meaning using standard cut-offs, most participants met a criteria for anxiety and depression). On the Spence Anxiety Scale (SCAS) the average entry score was 35, with above 33 indicative of anxiety in this age group. Such mild to moderate elevations in depression and anxiety symptoms are very responsive to nonspecific support.
- There is an anomaly in the data on improvement: in the physical function test, the average base level of the children at entry into the trial was 54.5 (n=99), considered severely physically impaired. Only 52.5% of participants had been able to attend at least 3 days of school in the week prior to their entry into the study. Yet those assigned to SMC+LP were well enough to attend 3 consecutive days of sessions lasting 4 hours. The reports of severe physical disablement do not match the capabilities of those who participated in the course. Were the children’s self-reported poor physical abilities exaggerated to justify enrolment in the trial? Were the children’s elevated depression and anxiety symptoms responsive to the nonspecific elements in extra time of being assigned to LP plus standard care?
- If the subjective self-report is accepted as a recovery criterion, in LP, just 12 hours of talk therapy, added to SMC would cure the majority of children with CFS. Such an effect would be astonishing, if true. In randomized controlled trials in adults with CFS/ME, such dramatic restoration of physical function (a wholesale return to near normal) is universally not seen. The SMILE Trial is clearly unbelievable.
- SMILE’s reliance on the broad NICE criteria means there is a clear risk patients were included in the trial who would not have met stricter definitions of the illness. There is a growing concern that loose entry criteria in clinical trials in ME/CFS allow enrolments of many participants who do not in fact have ME/CFS. A detailed study of CFS prevalence found many children are wrongly diagnosed with CFS, when they may just be suffering from general fatigue and/or mental health complaints (Jones et al., 2004). SMILE uses NICE guidelines to diagnose CFS: fatigue must be present for at least 3 months with one or more of four other symptoms, which can be as general as sleep disturbance. In contrast, Jones et al. showed that using the Centre for Disease Control criteria of at least four specific symptoms alongside detailed clinical examination, many children believed to have CFS are diagnosed with other exclusionary disorders, often general fatigue, mental health complaints, drug and alcohol abuse or eating disorders (that are often not readily disclosed to parents or doctors).
- LP involves attempting to coerce clients into thinking that they have control over their symptoms and to block out symptoms. This alone would distort any response by a participant in a follow-on questionnaire about symptoms.
- LP was delivered by people from the Lightning Process Company. Phil Parker and his employees held a clear financial interest in a positive outcome in SMILE. Such an obvious conflict of interest is hard to disentangle and totally nullifies any outcomes from this trial.
The SMILE trial holds many anomalies and leaves us with more questions than answers.
It is not clear whether the children enrolled in the trial, diagnosed with CFS using NICE criteria, might of been deemed non-CFS using more stringent clinical screening (e.g. CDC or IOM Criteria).
There is no way of determining whether any effect following SMC+LP was anything more than the result of non-specific factors, psychological tricks and persuasion.
The fact LP+SMC appears to have cured the majority of participants with as little as 12 hours talk therapy is a big flashing red light that this trial is clearly fundamentally flawed.
There is a very real danger of promoting LP as a treatment for CFS/ME: The UK ME Association conducted a survey of members (4,217 members) and found that 20% of those who tried LP reported feeling worse (7.9% slightly worse,12.9% much worse). SMILE cannot be, and should not be, used to justify LP as a treatment for CFS/ME.
The Lightning Process has no scientific credibility and this trial highlights a fundamental flaw in contemporary clinical trials: they are susceptible to suggestion, bias and spin. The SMILE trial appears to draw paediatric CFS/ME clinical care for children into a swamp of pseudoscience and mysticism. This is a clear step backward. There is little to smile about after reviewing the SMILE trial.
Dr. Geraghty is currently an Honorary Research Fellow within the Centre for Primary Care, Division of Population Health and Health Services Research at the University of Manchester. He previously worked as a research associate at Cardiff University and Imperial College London. He left a career in clinical medicine after becoming ill with ME/CFS. The main themes of his work are doctor-patient relationships, medically unexplained symptoms, quality and safety in health care delivery, physician well-being and evidence-based medicine. He has a special interest in medically unexplained symptoms (MUS), and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome.
Although only recently published, his recent ‘PACE-Gate’: When clinical trial evidence meets open data access is already ranked #2 out of 1,350 papers in altmetics in Journal of Health Psychology.
A recent Times article cited Dr Geraghty on reasons why NICE need to update their recommendations for ME/CFS
Special thanks to John Peters and David Marks for their feedback.
Coyne, J. (2017) Mind the Brain Blog, https://www.coyneoftherealm.com/blogs/mind-the-brain/embargo-broken-bristol-university-professor-to-discuss-trial-of-quack-chronic-fatigue-syndrome-treatment
Dorothy Bishop andExpert Commentary to the SMC (2017) http://www.sciencemediacentre.org/expert-reaction-to-controversial-treatment-for-cfsme/
1. Crawley, E., et al., Chronic disabling fatigue at age 13 and association with family adversity. Pediatrics, 2012. 130(1): p. e71-e79.
2. Crawley, E.M., et al., Clinical and cost-effectiveness of the Lightning Process in addition to specialist medical care for paediatric chronic fatigue syndrome: randomised controlled trial. Archives of Disease in Childhood, 2017.
3. Jones, J.F., et al., Chronic fatigue syndrome and other fatiguing illnesses in adolescents: a population-based study. Journal of Adolescent Health, 2004. 35(1): p. 34-40.