There seems to be more and more systematic reviews and meta-analyses recently shared on social media that examine the effectiveness or efficacy of a treatment with a specific health problem. Last month, someone shared a Cochrane Review on Muscle Energy Technique (MET) on Facebook, which piqued my curiosity to see what it says.
MET is “a system of manual therapy for the treatment of movement impairments the combines the effectiveness, safety, and specificity of reeducation therapies and therapeutic exercise. The therapist localizes and controls the procedures, while the patient provides the corrective forces and energies for the treatment as instructed by the therapist.” (1) It was developed by an osteopath named Fred Mitchell, Sr., about 50 years ago, which was refined by his son, Fred Mitchell, Jr. “MET uses the patient’s own voluntary muscle contraction in a precisely controlled direction against an operator applied counter-force,” the researchers described in the latest Cochrane Review of MET’s effectiveness for nonspecific low back pain. (2)
However, the current philosophical underpinnings of MET is quite uncertain.
“This is a difficult question because I don’t think there is any more a common philosophy that underpins the application of MET,” Dr. Gary Frye explained, who is an associate professor at the College of Health and Biomedicine at Victoria University in Melbourne, Australia.
“Like the profession MET arose from — osteopathy — there are widely differing views on the principles of assessment and treatment, and different authors have emphasized different methods of application and uses of MET. The originator of MET, Fred Mitchell Sr., developed MET within a biomechanical framework using particular spinal and pelvic biomechanical models to guide assessment and treatment. My approach to the underlying principles and application of MET is influenced strongly by what appears supportable from modern research evidence.
“I have written and criticised the mechanistic approach commonly advocated in MET texts because I don’t think these principles are valid, and a purely biomechanical approach is out of touch with modern understanding of pain and disability.
“Rather than applying MET based solely on a biomechanical diagnosis, I think it is reasonable to instead consider the likely physiological actions of the technique, such as pain modulation, changes to motor control, reducing fear and gaining confidence in muscle contraction and movement, and apply it with these aims in mind.”
In an email correspondence, Dr. Fryer have had a long-standing interest in MET as a practitioner, educator, and researcher. “Anecdotally, I’ve found it to be very useful for patients, who generally like being treated with it, and I think it more effectively treats the neuromuscular influences in dysfunction more than passive manual techniques,” he said. “My colleagues and I became aware of a growing number of studies that examined the effectiveness of MET for LBP.
“Many of the studies reported positive outcomes for MET. We ended up including 12 studies in our review, but there were more potential studies that were excluded because of design or methodology issues, such as lack of proper randomisation or other aspects that did not meet the RCT criteria. Given the number of studies on this topic, we believed that a systematic review was required to appraise and summarise this evidence for clinicians.
“Unfortunately the conclusions of the review did not end up providing clear guidance to clinicians, but at least we know where we stand regarding the evidence and know that we need better clinical trials in this area.”
In the Review, the researchers reviewed 12 qualified studies with 14 comparisons with a total of 500 subjects. They excluded studies in which participates had specific cause of low back pain (e.g. fracture, tumor, infection) and pregnant women. Based on these findings, the concluded that MET “is not effective for patients with non-specific LBP,” based on the all but one low-quality studies. “There is not sufficient evidence to confidently determine whether MET is likely to be effective in practice and large, methodologically sound studies are necessary to investigate this question. Given this, no implications for practice can be made at this stage.” (2)
The randomized-controlled trials included compared four factors:
1. MET plus any intervention versus that same intervention alone;
2. MET versus no treatment;
3. MET versus sham MET;
4. MET versus all other therapies. (2)
And here’s what they found:
1. Six low-quality studies found “no difference regarding pain and functional status” between MET + other intervention vs. other intervention alone for chronic nonspecific low back pain. For the acute version, one small study found no difference in pain intensity either.
2. There is “no clinically relevant difference in pain intensity between MET and sham MET,” based on one study with 20 subjects with non-specific acute low back pain.
3. Based on two high-risk of bias studies on the effects of nonspecific acute low back pain, there was “no clinically relevant difference between MET and other therapies” for pain intensity, and “no difference between MET and other therapies” for functional status.”
4. No surprise! There were no difference of pain intensity reported between interventions. (4)
When asked if MET is concurrent with the modern understanding of pain, Dr. Fryer responded, “It can and should be congruent with the current understanding of pain, but I suspect that it is often practiced in a way that is not congruent.”
He cautioned that if MET is practiced in a conceptual framework that is not consistent with pain science, the treatment could potentially reinforce pain and disability to patients.
“Traditional books on MET emphasise biomechanical assessment, structural spinal and pelvic dysfunctions, and biomechanical treatment effects. This may lead the practitioner to overemphasize any possible biomechanical factors in pain and in treatment, and inadvertently ignore important influences of pain mechanisms and psychosocial factors.
“Worse, the practitioner may reinforce the belief of a patient in a biomedical or structural cause for their pain when one might not exist.”
“Consider the case of a patient with persistent low back pain where they have little tissue injury or dysfunction, and their pain may be largely be attributed to abnormal pain processing (central sensitisation). If a practitioner overlooks the clues to sensitisation (chronicity, widespread pain, hyperalgesia, allodynia) and is focused on a biomechanical model of assessment and treatment, they may end up ‘diagnosing’ and treating a number of structural ‘lesions’, such as ‘flexed and rotated vertebra’, ‘rotated innominate bones’, and ‘torsioned sacrums’.
“These dysfunctions are theoretical and likely erroneous, but the description of such to a patient will very likely reinforce the ‘bone out of place’ misconception, encouraging the patient to think that the cause of their pain is a twisted pelvis or the like, leading to catastrophising, fear avoidance behaviour, and dependence on passive treatment in order to ‘keep putting the bone back in place’.”
Dr. Fryer explains a more plausible mechanism of how MET and other types of manual therapies might work.
“The therapeutic mechanisms underpinning most manual techniques, including MET, are still largely speculative, but is it likely that these techniques have an influence on pain processing (descending inhibition) and reducing the fear of movement and pain (in combination with reassurance and education). There is limited evidence on the effect of MET on spinal pain and what evidence exists is of poor quality, but the hypothesis that isometric muscle contraction influences descending pain inhibitory systems in the short-term is certainly plausible.
“Isometric contraction has been seen to affect pain perception during stretching, such as the many contract-relax techniques producing a greater tolerance of stretch. In addition to short-term pain modulation, I think there are important potential benefits of guided voluntary contraction and relaxation – together with pain education and reassurance – in improving confidence and reducing fear avoidance behaviour associated with pain.”
Since most of the trials were classified as low quality, could such trials be better?
“Higher quality randomised controlled trials (RCTs) on MET for spinal pain can certainly be made and there are many guidelines that are available to ensure rigorous quality, such as the CONSORT statement,” Dr. Fryer explained. “Ideally trials should involve a larger number of participants than previous trials, have clear randomisation and blinding (where possible) procedures, limit the co-interventions, and have a clear description of the treatment. A pragmatic comparative effective design is probably most suitable, where a MET approach might be compared to another manual therapy and to a comparison group consisting of usual care.
“But I suspect there is little will from researchers to do high quality RCTs on MET. Larger RCTs are very expensive and I think researchers are most likely to test pragmatic treatment approaches – eclectic manual therapy treatments that include but are not restricted to MET – in any large trial for a better chance of a positive treatment outcome and to be generalizable to the real world practice setting.
“Applying a single technique such as MET does not represent the typical treatment approach of most practitioners (who instead typically use a range of different techniques), so we are unlikely to see large trials that examine just MET. So I think we may see more small ‘proof or principle’ efficacy trials – hopefully of high quality – that examine whether MET influences parameters such as pain and movement in tightly controlled situations, but not larger clinical trials using just this intervention.”
One statement in the paper was puzzling: “This suggests that MET is not effective for the treatment of LBP, but, given the low quality of the evidence, no conclusions can be made unless large high-quality studies are available.”
So what can clinicians make use of the existing evidence?
“Clinicians should be aware that the current evidence for the effectiveness of MET is of poor quality and does not provide guidance as to its effectiveness for LBP,” Dr. Fryer said. ”There is some ‘proof of principle’ evidence that MET may improve pain and range of movement, and so it may be a useful addition to an evidence-based treatment protocol, but caution is needed given the limitations of this evidence and the lack of clinical studies.”
To those who are familiar with statistics and methodology, however, a closer examination of the Cochrane Review raised a few eyebrows.
Questioning the Review Itself
While the Cochrane Review doesn’t shy away from a negative report, the review itself may have its own reporting biases. If this systematic review is up to Cochrane’s standards, then it should be critically appraised for possible pitfalls. “Using GRADE does not judge the quality of the study itself, but the quality of the evidence for the chosen outcome of interest, which were pain intensity and function primarily,” Monica Noy identified, who is a practicing osteopath and a continuing education provider for registered massage therapists in Toronto, Ontario.
“A Cochrane review would have accepted far less studies to include because the criteria for inclusion usually also come with some sort of randomized-controlled trial critical appraisal, like the PEDro scale or the JADAD score for example of just two. These assess for internal validity of the study, like was it a good study? Low internal validity means a low quality study, which can still be included, but the score means something about outcome.”
And so, this Cochrane Review risks bias assessment and the quality of evidence for the outcome, not the quality of the study itself. “If the trials were assessed for internal validity and found wanting, then you could not really present the low quality evidence argument. You can’t assert that the evidence is low quality if the study itself has significant problems. The evidence in that case would be of little value as any type of quality,” Noy added.
“The low-quality evidence argument provides a case for setting about to get high-quality evidence, but the argument would be different if the quality of the studies themselves was provided.” Noy cited this passage in the review:
“Studies conducted to date generally provide low-quality evidence that MET is not effective for patients with non-specific LBP. There is not sufficient evidence to confidently determine whether MET is likely to be effective in practice and large, methodologically sound studies are necessary to investigate this question. Given this, no implications for practice can be made at this stage.”
“What this tells me is that this review is being done to provide evidence that there is not enough ‘quality’ evidence to make an assessment for clinical use (construct and reporting bias),” Noy continued. “MET is applied specifically and according to MET assessment, which is really question number one. What is the plausibility of providing a technique, no matter how specific, if the assessment criteria is questionable and remains questionable? There are enough studies out there on lack of specificity of manual assessment that any technique based on and derived from manual assessment ‘findings’ is dodgy to begin with.”
One of the last paragraphs raised Noy’s eyebrow:
“A number of clinical trials have examined the effect of osteopathic management on the treatment of LBP where MET has been a component of the treatment. Many of these studies have reported favourable results, but it is not possible to determine the influence of MET in the treatment package. Several systematic reviews have been performed using these studies to determine the effect of osteopathic management for LBP or musculoskeletal pain. The conclusions of theses reviews have differed from generally favourable outcomes to inconclusive outcomes due the lack of available high-quality studies.”
Noy mentioned that this passage is evidence of low-quality evidence for osteopathy or any type of manual therapy or technique, which is a plead for “higher quality of evidence.”
“This is not a call to strengthen the research itself, but to strengthen the quality of evidence,” she stated.
While Noy acknowledged this review is informative to clinicians about the evidence behind MET for low back pain, the quality of the evidence lacks context if the study does not also provide information on the methodological quality of the research itself.
“The researchers cannot really be faulted for that, as it fulfilled the study’s purpose,” Noy added.
1. Mitchell, Jr F, Mitchell PKG. Muscle Energy Manual Volume 1. MET Press; 1995.
2. Franke H, Fryer G, Ostelo R, Kamper S. Muscle energy technique for non-specific low-back pain. A Cochrane systematic review. International Journal of Osteopathic Medicine , Volume 20 , 41 – 52.