I am preparing a keynote address, Mindfulness Training for Physical Health Problems, to deliver at the World Congress of Behavioural and Cognitive Therapies in Melbourne on Friday, June 24, 2016. I have been despairing about the quality of both the clinical trials and systematic reviews of mindfulness treatments that I have been encountering.
Mindfulness training, mindfulness-based cognitive therapy, and mindfulness-based stress reduction (MBSR) are hot topics. That means that studies get published with obvious methodological problems ignored, and premature and exaggerated claims are rewarded. Articles are prone to spin and confirmation bias, not only in the reporting the results of a particular study, but also in what past studies get cited or are buried, depending on whether they support all the enthusiasm.
It is difficult to get a fair evidence-based appraisal of MBSR for clinicians, patients, and policy makers. There is a bandwagon rushing far ahead of what best evidence supports. Aside from all the other problems, the literature is being hijacked by enthusiastic promoters with undisclosed conflicts of interest who hype what they don’t tell us they are offering for sale elsewhere. Clear declarations of conflicts of interest, please.
We know that MBSR is better than no treatment. But there is only weak and inconsistent evidence telling us whether MBSR is better than other active treatments delivered with the same intensity and the same positive expectations.
Most often, MBSR is compared to a waitlist control or treatment as usual. Depending on the context, we don’t know if these control groups are actually the opposite of placebos, nocebos Patients agreed to participate in a study that gave them a chance to get MBSR. Left in the (unblinded) control condition, they got nothing except having to be assessed repeatedly. They are going to be disappointed and this reaction is going to register in the self-report outcome data received from them.
Also, the ill-described routine care or treatment as usual as being provided may be so inadequate that we are only witnessing MBSR compensating for poor quality treatment, rather than making an active contribution of its own. This was particularly true when mindfulness training was used to taper patients from antidepressants. Patients receiving MBSR and tapering were compared to patients remaining in the routine care in primary care in which they had placed on antidepressants some time ago. At the time of recruitment, many patients were simply being ignored with minimal or no monitoring of theyor whe were taking their medication or they were even still depressed. It’s not clear whether reassessing whether the medication is still being of any benefit and providing support for tapering would’ve accomplished as much or than MBSR accomplished, without requiring daily practice or a full day retreat.
Data describing treatment as usual or routine care control conditions are readily available, but almost never reported in studies of evaluating MSB are.
To take another example, when patients with chronic back pain are being recruited from primary care, their long term routine care eventually lacks support, positive expectations, or encouragement and may become iatrogenic because guidelines requiring escalating futile interventions. Here too, putting back in some support and realistic expectations may work as well as more complicated n interventions.
Some members of the audience in Melbourne surely anticipate a relentlessly critical perspective on mindfulness from me. They will be surprised when I present limitations of current literature, but also positive recommendations for how future studies can be improved.
We need less research evaluating MBSR, but of a better quality.
There is far too much bad mindfulness research being done and uncritically cited and being put into systematic reviews. A meta-analysis cannot overcome the limitations of individual trials, if the bulk of the studies being integrated share the same problems. Garbage in, garbage out is a bit too harsh, but communicates a valid concern.
I think it is very important that meta analysis with a hot topic like MBSR not become overly focused on summary effect sizes. Such effect sizes are inevitably inflated because of a dependence on at best a few small studies with a high risk of bias, which includes the allegiance of overenthusiastic investigators. These effect sizes are best ignored. It is better instead to identify the gaps and limitations in the existing literature, and how they can be corrected.
Stumbling on a quality review of MBSR for fibromyalgia.
I was quite pleased to stumble upon a review and meta-analysis of MBSR for fibromyalgia. Although it is published in a pay walled journal, a PDF is available at ResearchGate.
Lauche R, Cramer H, Dobos G, Langhorst J, Schmidt S. A systematic review and meta-analysis of mindfulness-based stress reduction for the fibromyalgia syndrome. Journal of Psychosomatic Research. 2013 Dec 31;75(6):500-10.
Here’s the abstract:
Objectives: This paper presents a systematic review and meta-analysis of the effectiveness of mindfulness-based stress reduction (MBSR) for FMS.
Methods: The PubMed/MEDLINE, Cochrane Library, EMBASE, PsychINFO and CAMBASE databases were screened in September 2013 to identify randomized and non-randomized controlled trials comparing MBSR to control interventions. Major outcome measures were quality of life and pain; secondary outcomes included sleep quality, fatigue, depression and safety. Standardized mean differences and 95% confidence intervals were calculated.
Results: Six trials were located with a total of 674 FMS patients. Analyses revealed low quality evidence for shortterm improvement of quality of life (SMD=−0.35; 95% CI−0.57 to−0.12; P=0.002) and pain (SMD=−0.23; 95% CI −0.46 to −0.01; P=0.04) after MBSR, when compared to usual care; and for short-term improvement of quality of life (SMD=−0.32; 95% CI −0.59 to −0.04; P=0.02) and pain (SMD=−0.44; 95% CI −0.73 to −0.16; P=0.002) after MBSR, when compared to active control interventions. Effects were not robust against bias. No evidence was further found for secondary outcomes or long-term effects of MBSR. Safety data were not reported in any trial.
Conclusions: This systematic review found that MBSR might be a useful approach for FMS patients. According to the quality of evidence only a weak recommendation for MBSR can be made at this point. Further high quality RCTs are required for a conclusive judgment of its effects.
I will be blogging about MBSR for fibromyalgia in the future, but now I simply want to show off the systematic review and meta-analysis and point to some of its unusual strengths.
A digression: What is fibromyalgia?
Fibromyalgia syndrome is a common and chronic disorder characterized by widespread pain, diffuse tenderness, and a number of other symptoms. The word “fibromyalgia” comes from the Latin term for fibrous tissue (fibro) and the Greek ones for muscle (myo) and pain (algia).
Although fibromyalgia is often considered an arthritis-related condition, it is not truly a form of arthritis (a disease of the joints) because it does not cause inflammation or damage to the joints, muscles, or other tissues. Like arthritis, however, fibromyalgia can cause significant pain and fatigue, and it can interfere with a person’s ability to carry on daily activities. Also like arthritis, fibromyalgia is considered a rheumatic condition, a medical condition that impairs the joints and/or soft tissues and causes chronic pain.
You can find out more about fibromyalgia from a fact sheet from the US National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMSD) that is only minimally contaminated by outdated notions of fibromyalgia being a psychosomatic condition, i.e., all in the head, or recommendations for unproven complementary and alternative medicines.
What I like about this systematic review and meta-analysis.
The authors convey familiarity with the standards for conducting and reporting systematic reviews and meta-analyses, recommendations for the grading of evidence, and guidelines specific to the particular topic, fibromyalgia. They also admit that they had not registered their protocol. No one is perfect, and it is important for authors to indicate that they are aware of standards, even when they do not meet them. Readers can decide for themselves how to take this into account.
This review was planned and conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. guidelines (PRISMA) , the recommendations of the Cochrane Musculoskeletal Group [16,17] and the GRADE recommendations (Grading of Recommendations Assessment, Development and Evaluation) . The protocol was not registered in any database.
The authors also laid out key features of systematic review and meta-analyses where you would expect to find them with explicit headings: eligibility criteria, search strategy, study selection and data collection including risk of bias in individual studies, etc.
Designation of primary and secondary outcomes.
Fibromyalgia causes pain and fatigue, disrupting quality of life. These are the outcomes in which patients and their healthcare providers will be most interested. Improvement in pain should be given the priority. However, in clinical trials of MBSR for fibromyalgia ,investigators often administer a full battery of measures, and select the ones that are positive, even if they’re not the outcomes that will be most important to patients and providers. For instance, the first report from one trial focused on depressive symptoms . Designating depressive symptoms as the primary outcome ignored that not all patients with fibromyalgia have heightened depressive symptoms, and depression is not their primary concern. Moreover, the paper reporting this clinical trial is inconsistent with its regiistration, where a full range of other outcomes were designated as primary. Ugh, such papers defeat the purpose of having their protocols registered.
In the review under discussion, depressive symptoms were designated as a secondary outcome, along with sleep and fatigue.
Compared to what?
The review clearly distinguished waitlist/routine care from active comparison treatments and provide separate effect sizes.
The review also indicated whether the patients had been randomized to MBSR versus comparison treatment, and explicitly indicated that any significant effects for MBSR are disappeared when only randomized trials were considered.
Strength of recommendation.
The review took into account the small number of studies (4 randomized and 2 non-randomized trials with a total of 674 patients) and the low quality of evidence in grading the recommendation that it was making:
According to GRADE, only a weak recommendation could be made for the use of MBSR for FMS, mainly due to the small number of studies and low quality of evidence.
Summary of main results.
The article produces a series of forest plots [How to read one ] that graphically display the unambiguous results showing weak effects of mindfulness in the short-term but none in the long term. For instance:
This meta-analysis found low quality evidence for small effects of MBSR on quality of life and pain intensity in patients with fibromyalgia syndrome, when compared to usual care control groups or active control groups. Effects however were not robust against bias. Finally, data on safety were not reported in any study.
Agreements and disagreements with other systematic reviews.
The few other reviews of MBSR fibromyalgia are of poor quality. So, the authors of this review discusses results in the context of the larger literature of MBSR for physical health problems.
Implication for further research.
Too often, reviews of fashionable psychological interventions for health problems end with an obligatory positive assessment and, of course, “further research is needed.”
Enthusiasts assume MSBR is that it is good for whatever ails you. MBSR training can help you cope, if it doesn’t actually address your physical health problem. I really liked that this review gave pause and reflected on why MBSR should be expected to be the treatment of choice and to make sure that relevant process and outcome variables are being assessed.
Patients with fibromyalgia are seek to relieve their debilitating pain and accompanying fatigue, or at least resume some semblance of a normal life they have lost their condition. It is important that results of MBSR research allow informed decisions about whether it is worth the effort to patients and providers to get involved in MBSR or whether it would simply be more more burden with uncertain results.
One major implication for future research is that researchers should bear in mind that MBSR primarily aims to establish a mindful and accepting pain coping style rather than to reduce the intensity of pain or other complaints. Therefore researchers are encouraged to select custom outcomes such as awareness, acceptance or coping rather than intensity of symptom which might not reflect the intention of the intervention. Only two trials measured coping, however, only one of them actually reported results and the other one  did not provid data but stated that besides catastrophizing there were no significant group differences. Results of the trial by Grossmann et al.  on the other hand indicated significant improvements on several subscales, which could be worth further investigations.
Further high quality RCTs comparing MBSR to established therapies
(e.g. defined drug treatment, cognitive behavioral therapy) are also required
for the conclusive judgment.
This systematic review found low quality evidence for a small short term improvement of pain and quality of life after MBSR for fibromyalgia, when compared to usual care or active control interventions. No evidence was found for long-term effects.
Not much spin here or basis for yet recommending MBSR for fibromyalgia as ready for implementing in routine care.