ACUFLASH in a pan

The journal Menopause has just published a randomised controlled trial of acupuncture for hot flushes in menopausal women (ACUFLASH).

The authors make some key points in their introduction which are worth highlighting:

The effects of acupuncture treatment are those of a treatment package including the therapeutic relationship and expectation, and acupuncture is available for women seeking an alternative to pharmacological treatment

I’m glad the authors have pointed out that any acupuncture intervention is not simply about the act of sticking a needle in someone. I thought this was revealing:

the treatment is individually tailored and compromises both lifestyle advice and needling in selected acupuncture points

There are major issues with recruitment in this trial. After all, if part of your intervention is based on ‘expectation’ then it is going to matter where you select your patients. Women were recruited by newspaper advertisements and media coverage. They weren’t recruited in a systematic way, say from consecutive patients attending gynaecology appointments or in a GP surgery. Of these 535 volunteers only a further 428 were deemed eligible to get a diary and then only 399 completed and returned it.

A rather large number of women (80) didn’t enter the study because they wanted immediate treatment. In the end, only 267 entered the randomisation phase and 2 of these withdrew immediately when they didn’t get acupuncture. Another 14 in the selfcare arm dropped out at 4 weeks because they ‘wanted active treatment’. The total number of women analysed was 248.

Now apparently Norway is a popular place for acupuncture. The paper quotes two recent surveys that showed 28% reported lifetime use and 10.8% reported use within the previous year. The baseline characteristics make eye-watering reading if one is considering the quality of the ‘sample’. Previous use of acupuncture came in at 64% in both acupuncture and self-care group. A whopping 60% of the intervention group expected acupuncture to relieve their hot flushes and 51% of the non-acupuncture self-care group.

Clearly, the women were aware of the treatment they were given. A far more interesting study would have been to give 10 sessions of further ‘lifestyle advice’ to an additional arm (or instead of the self-care group) and assess if there was a significant difference between the groups.

My next major concern is with the length of the study. Why twelve weeks? This seems an incredibly short period of time for a problem which can affect a third of women for up to 5 years. My inner skeptic tell me this is because it is unlikely that the benefit will be sustained. Or it could be because the costs of giving weekly acupuncture for 5 years for minimal or no benefit would make an MP weep. We’ll never know from this trial anyway.

And I would point out that if you gave HRT for 12 weeks the risks of adverse effects would be absolutely minimal.  I recognise the women may have been recruited on the basis that they didn’t want drugs but it would be an interesting comparison.

It’s worth a look at some of the ‘significant’ results. These women started with an average of 12.6 hot flushes per day. Acupuncture reduced the number of hot flushes by an average of 2.1 per day. This may be statistically significant but is clinically deeply unimpressive. Mean hours of sleep increased by a statistically significant 0.28 hours – that’s just under 17 minutes to you and me.

We can then add these dubious clinical benefits to the massive weaknesses in this study – a flawed recruitment process, an unfair comparison without placebo and a trial period which is clinically irrelevant.

Another comment from the authors:

This study does not allow us to estimate what proportion of the clinical benefit was due to the effects of needling itself and what was due to other factors, such as the patient-provider interaction.

Of course not. If I may re-phrase this slightly – it actually doesn’t allow us to interpret what proportion could be an enhanced placebo effect. The authors claim at the very end that acupuncture treatment is comparable to the effect of SSRIs on hot flushes. I will wait for the trial that puts them head to head before I go for that line.

They claim that acupuncture can contribute to a clinically relevant reduction in hot flushes. I think this is an optimistic interpretation given the absolute reductions in flushes and the absolute improvements in sleep for the limited period of 12 weeks. Given the costs involved I won’t be rushing off to the local commissioners to demand more acupuncture on the NHS.

Many of the aspects of this trial superficially ticks the boxes for a decent quality methodological study. Several don’t; but then that is not unusual in any study. My main gripe is with the suggestion that this is some pragmatic trial and not a method, rather coyly, to enhance the evidence base for acupuncture. Unfortunately, they also fall into the trap of using the term ‘pragmatic’ to excuse methodological flaws. They might as well call it a pragmatic pilot and be done with it. It’s not reasonable to compare acupuncture with ‘normal self care’ (ie no treatment) because it’s been done and the placebo effect is a nightmare. Placebo may be difficult with acupuncture but that doesn’t make research without decent comparisons acceptable. It might be highly convenient to compare acupuncture for just 12 weeks for a condition that continues for months and months but it isn’t that useful. Pragmatism is just a smoke screen for a self-serving slice of CAM dogma. 

The main researchers are based at the National Research Centre in Alternative and Complementary Medicine in Norway and this makes one wonder about the flaws in this study. Many would claim that acupuncture is an elaborate and theatrical placebo and this paper really doesn’t move us on past that position. I have no doubt they will be calling for more research…

Einer Kristian Borud, Terje Alraek, Adrian White, Vinjar Fonnebo, Anne Elise Eggen, Mats Hammar, Lotta Lindh Astrand, Elvar Theodorsson, & Sameline Grimsgaard (2009). The Acupuncture on Hot Flushes Among Menopausal Women (ACUFLASH) study, a randomized controlled trial Menopause: The Journal of The North American Menopause Society, 16 (3), 484-493 DOI: 10.1097/gme.0b013e31818c02ad