lørdag 30. april 2011

The Back Letter and the Alexander Technique

- the hazards of misinterpreting a clinical trial

In August of 2008 The British Medical Journal published a «Randomised controlled trial of Alexander technique lessons, exercise, and massage (ATEAM) for chronic and recurrent back pain».(1) The publication was accompanied by extensive media coverage and even made it onto the BBC news.

About a month later an American medical journal, The Back Letter, published an article criticising the study, and the media hype. Norwegian back-care specialists have used The Back Letter article as a reference when assessing the ATEAM-trial, and being an Alexander Technique teacher in Norway I have been interested in reading the article for quite some time. The Back Letter is unfortunately not freely available on the internet and it was only recently I got around to purchasing and reading it.

The article, titled «The Alexander Technique for Low Back Pain: The Hazards of Overinterpreting a Single Clinical Trial»(2), raises some very pertinent questions about the ATEAM-study, but expecting something amounting to a scientific standard I was surprised by its low quality.

Media coverage
STAT, The Society of Teachers of the Alexander Technique, engaged a PR company some years ago to help raise the profile of the Alexander Technique. When the ATEAM-trial was published in the BMJ they succeeded in having the results reported in major British news media. Unfortunately words like «cure», «treatment» and «therapy» were used, words that I as an Alexander Technique teacher never will use myself. The media coverage reinforced the misconception that the Alexander Technique is a form of therapy.
The Back Letter studied 30 media features on the ATEAM-trial. None of them expressed any sceptic views on the study. One has to conclude that the media coverage at the time was unbalanced. This kind of publicity will in the long run be detrimental to the credibility of the Alexander Technique profession.

But what did the Back Letter say about the ATEAM-trial itself?

24 versus 6
The Back Letter article claims that the trial compares 24 lessons of Alexander Technique with 6 sessions of massage. This is not correct. 6 sessions of massage were compared with 6 lessons in the technique. The reason for doing this was to investigate whether positive outcomes from Alexander Technique lessons could be due to placebo-effects related to touch.

«We particularly wanted to assess massage as it provides no long term educational element, in contrast with lessons in the Alexander technique.» (1)

«That six sessions of massage were much less effective at one year than at three months whereas six lessons in the Alexander technique retained effectiveness at one year shows that the long term benefit of Alexander technique lessons is unlikely to result from non-specific placebo effects of attention and touch.» (1)

In the tables representing the results of the study the 6 sessions of massage is placed next to the 6 lessons in the Alexander Technique. The fact that The Back Letter makes this claim about 24 versus 6 indicates that they have not understood the basics of the design of the study.

Exercise intervention
I was surprised to see The Back Letter deem the exercise intervention of the ATEAM-trial to be «unproven». According to the authors of the ATEAM-trial their exercise intervention is comparable with the one in the 2004 UK BEAM-study(3):

«Comparison with the United Kingdom back pain exercise and manipulation trial suggests the benefits are similar to a supervised exercise scheme in the short term, and potentially greater in the long term, since the effect of supervised schemes in that trial was no longer apparent by 12 months.» (1)

The Back Letter would have liked the ATEAM-study to have included «a proven exercise intervention of similar duration and intensity, involving similar personal attention and face-time with providers» as the 24 lessons of Alexander Technique. This seems like a good idea as it could have prevented misrepresentation of the trial outcome, but reveals that the authors regards the Alexander Technique as an alternative to exercise, which it is not.
The design of the ATEAM-trial means that the type or intensity of exercise intervention is not important as long as it has en effect, and the four exercise groups have the same rate of adherence.

Numbers of lessons and assessor blinding
The participants having massage or Alexander Technique lessons could pay for additional sessions/lessons at their own expense. The Back Letter article questions the lack of information about these possible additional lessons/sessions. I myself can't find any information about how this possible extra intervention was treated during the analysis.

One of the interviewees in the Back Letter article, Roger Chou, MD, of Oregon Health and Science University adds the question whether outcomes assessors were blinded to intervention status as the study only says that data entry was blind to study group.

These are potentially very important issues.(4) I say 'potentially' because The Back letter does nothing to clarify this. As a layperson I find this attitude from a scientific journal peculiar. They find something that could possibly give reason to question the outcomes of the trial they are reviewing, but do nothing to get the facts on the table. An email to the ATEAM-trial authors may have been sufficient.

Timing of the 6 sessions of massage
Concerning the 6 sessions of massage The Back Letter article says that: «The authors didn’t specify when those sessions were delivered». It is natural to assume that those were given in parallel with the six lessons of Alexander Technique and if given later it would only have favoured massage over Alexander Technique lessons at 3 months. Table 1 in the trial paper lists the interventions and give information on how they were delivered. The massage sessions were given «one session a week for six weeks». (5)

The back pain patient population
The Back Letter article claims that «the study results are not clearly generalizable to other clinical populations». The reason given is the fact that the participants where former back pain patients having contacted the health service because of back pain from three months and up to five years previously. «These people were not seeking care and may not be comparable to those who visit their primary care doctor for treatment of lower back pain», the Back Letter continues.

I can envisage that patients contacting their GP's because of back pain would be in more acute pain than patients not actively seeking medical assistance at that particular moment in time. But these people still have the same diagnosis meaning that they in any case would have been treated according to the same guidelines for intervention.
The Back Letter article does not substantiate why there should be any difference, or why the participants were not representative of a «predominantly chronic, severely affected, and currently ineffectively managed population». (1)

What did the control group do?
The control group received «normal care» and The Back Letter article notes that: «It is not clear whether this involved some type of medical care or was simply a continuation of the self-care the subjects engaged in prior to recruitment». Unfortunately this piece of information is lacking in the published ATEAM paper.

The information given by STAT to their members at the time of publishing was that this group was to continue whatever medical intervention they were receiving at the time and seek further assistance when needed. It would have been interesting to know to what extent this group received medical intervention. The fact that the score is more or less unchanged from baseline to 12 months goes to show that the description of a «predominantly chronic, severely affected, and currently ineffectively managed population» is fitting.

It is a sad fact that if you come to your doctor with chronic back pain there is little he or she can offer that has long term effects. Exercise intervention has some effect, but the effect is low. (6)

What boosting what?
The Back Letter article claims the exercise «boosted» the effect of Alexander Technique lessons. This is not necessarily correct. The Back Letter authors use data from two different sources. Looking at the outcome scores from the same table, table 5 (7), the adding of exercise shows insignificant change on the Roland-Morris disability score (-0.08), and no change in number of days in pain (-20). The combination of 6 lessons of Alexander Technique and exercise gives a reduction on the Roland disability score (-1.44 to -2.98), but no change in the number of days in pain (-13). In total only one out of four outcome numbers are significantly affected.

The Alexander Technique aims at helping people to «avoid poor habits affecting postural tone and neuromuscular coordination»(2). This improves the quality of movement, which is why the Alexander Technique is widely used by musicians and actors. It is more plausible that Alexander Technique lessons boosted the effects of exercise, and not the other way around.

Moderate outcome on disability
According to the Back Letter the outcome levels on the Roland-Morris disability score was «moderate», but fails to put the findings into context. 24 lessons of Alexander Technique gave only a 3.4 advantage on Roland-Morris disability. This is still higher than most of the scores by conventional methods in other randomised controlled trials on back pain that I've had the opportunity and time to check up on.

Traditional Alexander Technique teaching focuses mainly on basic activities like standing, sitting and walking. I would suggest that taking a more dynamic approach by dedicating time to specifically target activities the individual has difficulties performing may increase the impact on disability score. This approach would not be different from the way an Alexander Technique teacher for instance helps a musician solve technical challenges in playing a musical instrument.

6 lessons in the Alexander Technique resulted in a change of only 1.4 on the Roland-Morris Disability Questionnaire. Although the result is not clinically significant, I find it quite remarkable that only six lessons still had an effect after 12 months. 20 to 30 lessons are generally regarded as necessary for learning the Alexander Technique and secure lasting change.

But what about the back pain?
Even more remarkable was that only six lessons in the Alexander Technique led to a 10 days reduction in days of pain compared to control (during the previous 4 weeks) at 12 months.

The Back Letter article says about the results on reduction in pain that: «It is more difficult to interpret the improvement on the other primary outcome measure, i.e. 'days in pain over the previous four weeks' —since 'days in pain' refers to pain of any intensity and functional impact».
The article then goes on to list the results, claiming that exercise gave no reduction in days in pain compared to the control group. Again, this is not correct, albeit the advantage of exercise regarding days in pain was small.(8)

Bearing in mind that this was a randomised controlled trial on back pain one should think that the results concerning pain was rather important. According to research patients tend to rate the reduction in pain as more important than reduction in disability when recovering from back pain.(9) Yet, the Back Letter article gives no further comments on the results on pain reduction.

24 Alexander Technique lessons resulted in 18 fewer days in pain compared to control. This is a reduction of almost 86 percent. Should the Back Letter article authors going to scrutinise any aspect of the ATEAM-trial it must be this result. It is almost too good to be true. Have any other intervention had the same success rate on chronic and recurrent back pain?

Further research
The Back Letter has nothing much to say about future research, only asking for a «level playing field». The Alexander Technique is a skill, and the experience from a hundred years of teaching it tells us that 20-30 lessons is necessary for learning and implementing the Alexander Technique in daily life. The 24 lessons were necessary to give the Alexander Technique access to the playing field.
Presumably supervised exercise interventions of intensity and duration comparable to the 24 lessons have already been tested. If not, the Back Letter would have done well to point this out.

The call for a «level playing field» misses, however, two important points.

First, the Alexander Technique is not a competitor to other interventions, and in particular not to exercise. It is a fellow player. The Alexander Technique may improve the performance of exercises thereby enhancing treatment effects.
No one learns much from six lessons in the Alexander Technique, it is comparable to having six piano lessons. But I would have liked to see another trial testing six lessons of Alexander Technique and an exercise intervention, using two different teaching strategies. In the ATEAM-trial no particular short or long course was devised.(10) If one intervention group were given six lessons where the main aim is to enable the participants to employ the 'constructive thinking' of the Alexander Technique to the performance of the exercises this may have an impact on the outcome score.
Such a trial would also give the opportunity to test whether having six lessons only is likely to give any effect after 12 months.

Second, back pain is a complex issue demanding a multidisciplinary approach. There is, and never will be, one intervention that outmatch the other ones. Not even the Alexander Technique. The ATEAM-trial adds important contributions going beyond proving the effects of the Alexander Technique. It shows that unlearning bad movement habits is likely to be effective against back pain. This vital aspect is completely lost on The Back Letter, but there are people within the the medical profession who have come to realise the impact of habitual movement strategies.

In a television programme produced by the Norwegian Broadcasting Company on occasion of the recent National Cross-Disciplinary Neck- and Back-Pain Congress (11) we saw a doctor instructing a small group of patients. The aim was to see whether unlearning bad habits and 'moving normally' would aid recovery. Simple habits related to the avoidance of pain, like leaning on the hand to relieve the back, was focused on. The patients had a few weeks of group instruction and three months later five out of six patients reported an improved condition.(12) The Alexander Technique offers a much more advanced approach to unlearning habits of movement, but it is in this context the technique must be viewed.

Seeing that educational approaches is gaining increased attention in the treatment of back pain I'm convinced that the medical profession is bound to seek the expert assistance from the Alexander Technique profession in the future.

Randomised controlled trial of Alexander technique lessons, exercise, and massage (ATEAM) for chronic and recurrent back pain

The Alexander Technique for Low Back Pain: The Hazards of Overinterpreting a Single Clinical Trial

United Kingdom back pain exercise and manipulation (UK BEAM) randomised trial: effectiveness of physical treatments for back pain in primary care.

Reporting of Outcomes in Orthopaedic Randomized Trials: Does Blinding of Outcome Assessors Matter? Poolman et al. (2007)
“Unblinded outcomes assessment was associated with a potential for exaggeration of the benefit of the effectiveness of a treatment in our cohort of studies.
Conclusions: In future orthopaedic randomized controlled trials, emphasis should be placed on detailed reporting of outcome measures to facilitate generalization and the outcome assessors should be blinded, when possible, to limit bias.”

ATEAM-trial. Table 1, trial groups for patients with chronic or recurrent back pain

Exercise therapy for chronic nonspecific low-back pain.
van Middelkoop et al. (2010)
Conclusions: The authors conclude that evidence from randomised controlled trials demonstrated that exercise therapy is effective at reducing pain and function in the treatment of chronic low back pain. There is no evidence that one particular type of exercise therapy is clearly more effective than others. However, effects are small and it remains unclear which subgroups of patients benefit most from a specific type of treatment.

ATEAM-trial. Table 5, individual groups one year after randomisation.

ATEAM-trial. Table 4, outcomes at one year after randomisation: mean difference compared with control group.

How little pain and disability do patients with low back pain have to experience to feel that they have recovered? Kamper et al. (2010)
“Subjects consider pain more than disability when determining their recovery status.” (abstract)

ATEAM-trial. Appendix. Trial interventions

The 2. National Interdisciplinary Neck- and Back-Pain Congress

NRK, «Når ryggen krangler!» (When the Back Complains)
You see the patient group at about 1 minute and at 24 minutes into the programme. This programme is in Norwegian only and may not be accessible outside of Norway.

2 kommentarer:

  1. The fact that the number of additional lessons was not documented is a strong critique. I'd call it a minor blunder on the part of the researchers. We are left to question whether AT is more effective, or just more addictive. From my experience it is very addictive. Lessons are mysterious and fascinating and make you want to learn more and more. Hopefully follow up studies will resolve this question.

  2. Thanks for comment. Concerning number of extra lessons The Back Letter omits the fact that a number of participants did not actually do all of the six or twenty-four lessons. I also presume that the British Medical Journal would not have accepted the paper if there were any serious doubts about the number crunching.
    The effectiveness of a regular course (20 lessons or more) of Alexander technique is clear anyway. What might be in doubt is whether only six lessons will be shown as effective after 12 months in another trial.