Evidence base for acupuncture: part 1

Healing Hands: Exploring Acupuncture
23 followers
1 Video View·May 23, 2025
Hello, my name is Ian Appleyard. I am the Research and Policy Manager at the British
Acupuncture Council. Welcome to this presentation on the evidence base for acupuncture.
This is an introductory presentation and I'll just be looking at some key pieces of evidence.
There are controversies with how acupuncture clinical trials ought to be interpreted
and I'll be dealing with those in a separate presentation,
so you can have a look at that if you are interested.
Now you may have seen or heard a sentence similar to this one,
there is no evidence that acupuncture works. This sentence actually has two potential meanings.
The first one is that clinical trials have been conducted and the results were negative.
That means that evidence shows that acupuncture is not effective for that particular condition,
and I believe many people when they see or hear this sentence believe that to be the case.
However, the sentence could also mean that there is simply not enough clinical evidence,
either no clinical trials have been conducted or those clinical trials are too small or poor
quality, therefore scientists cannot make a firm conclusion, and in both cases we can say
there is no evidence that acupuncture works, but obviously the meaning is quite different.
In reality, it is commonly the case that there are no insufficient good quality clinical trials,
so we are often talking about the story so far. So today we are going to look at the story,
the evidence for acupuncture for chronic pain where we have quite strong evidence,
for depression which we have a reasonable amount of evidence, and also have a quick look at other
conditions, many of which there simply isn't sufficient evidence for scientists to say one
way or the other. So let's start by looking at chronic pain. This really is the area where we
have the most high quality clinical trials. In my opinion, this is the highest quality
evidence regarding acupuncture for any condition, Vickers et al. 2018, which looks at chronic pain.
It is a meta-analysis. Now a meta-analysis combines the results of a number of different
clinical trials. So why is this particularly high quality? The first thing is that all the
clinical trials are high quality, which is only something that we can really do for chronic pain
because a lot more clinical studies have been carried out for chronic pain. So the study
combined the results from 39 clinical trials. That means the number of participants in the
clinical trials is also high, over 20,000 people, which is also very good. And finally,
the clinical trials were conducted in a number of different countries, including the UK,
so we have a good understanding of how effective acupuncture is for the UK population.
Chronic pain can come from a number of different conditions, so we need to be clear about which
conditions the Vickers et al. study actually looked at. So these were headache, which was
both migraine and tension type headaches, osteoarthritis, and all the studies were
osteoarthritis of the knee, shoulder pain, and then non-specific musculoskeletal pain,
basically lower back pain and neck pain. Before we look at the results, it's helpful
if we have a closer examination of what was being compared within the study.
So in the Vickers et al. study, there were two main comparisons. One is acupuncture versus placebo,
also known as sham acupuncture, and the other one was acupuncture versus no acupuncture control.
Now the sham placebo debate is quite complicated, and we'll deal with that in another presentation,
but for today, it's good to have a little look at what is meant by no acupuncture control.
For many people, they may believe that no acupuncture control means that there's no
therapy given at all, which is actually not the case. There were different forms.
The low-intensity form is where people, participants in the study, were able to
continue taking their standard medication, normally non-steroidal anti-inflammatory drugs,
but none of the studies actually forbid people from doing this. Then there was a more moderate
control, moderate intensity, and that would be basically anyone, the participants could continue
having whatever therapy it was their doctor was prescribing for them in their usual care.
Finally, there were some trials where they had an active comparison, high intensity,
and this is something like physiotherapy. All in all, these things together are the
no acupuncture controls. In none of the cases does it actually mean nothing was happening.
In fact, in some of them, patients, participants were actually having quite a high intensity,
extra physiotherapy, and things like that. Here is the result written out in a scientific format.
The simple answer is this shows that acupuncture does work for chronic pain.
However, to understand the sentence in a bit more detail, I picked out a few words. We have a P
value, standard measure, which indicates whether or not the result could have happened by chance.
Very low P value suggests that it is extremely unlikely to have happened by chance,
which is the case here. We also want to know how big a difference there is. Is acupuncture a lot
better than no acupuncture control, or is it a little bit better? This is known as the effect
size. The measure here for effect size are the standard deviations. We have an effect size of 0.5
compared to the no acupuncture controls. That by convention is considered a moderate effect size.
So acupuncture compared to a group of things, as we discussed in the previous slide,
like the medication or physio, all combined on average has an effect size of 0.5, which is
considered to be moderate. The effect size compared to the sham is 0.2, which is considered to be
small, but it is a real difference because that's what the P value tells us. The results can also
be described in terms of responder rates. A responder is somebody whose pain score reduces by
more than 50%. So if a participant's pain score was 6 out of 10 at the beginning of a clinical trial
and then dropped to 3 or less, then they count as a responder. Now across those 20,000 people,
we know that for those who had acupuncture, 50% of them were responders. For those who didn't
have acupuncture, but had continued with their usual care, etc, the no acupuncture controls,
30% of them were responders. This is another way of looking at the results in terms of a
plain language explanation. So we imagine that there is a pain score between 0 and 100, where
100 is as painful as it can possibly be. At baseline, at the beginning of the clinical trials,
the average pain score was 60, and we find that the no acupuncture control groups together on
average, their pain would be dropping to 43. Those who had sham acupuncture, their pain score
would drop to 35, and those who had acupuncture, their pain would drop to 30. So we can see
there's progressive improvement as you go from no acupuncture control to acupuncture.
The Vickers et al study also looked at whether the benefits of acupuncture lasted over time,
and the results show that the benefits do indeed last for at least one year. We don't know what
happens beyond one year, because the clinical trials were not following the participants for
that length of time. The other interesting thing, or the thing that we can note, is for many people
the idea of placebo is that you're made to feel better, and therefore there can be a short-term
benefit. The fact that these reductions in pain are seen to last pretty much for one year,
indicates that acupuncture is not a placebo. Okay, as a summary, the largest high-quality
systematic review for any condition looking at acupuncture is the Vickers et al study.
It shows that acupuncture is beneficial for those with chronic pain, and that those
reductions in pain are seen to last for at least one year. And finally, it also demonstrates that
acupuncture is not a placebo. Now that we have more high-quality clinical trials that have
looked into acupuncture for chronic pain, we find that acupuncture is being included into
national guidelines. So currently, the National Institute for Health and Care Excellence,
NICE, include acupuncture within two of their guidelines, one for chronic primary pain,
and the other which is for headaches. Secondly, we also have the Scottish Intercollegiate
Guidelines Network, which recommend acupuncture for osteoarthritis and chronic lower back pain.
Also, if we look internationally, we find acupuncture included in guidelines in the
United States, the American College of Rheumatology, recommending acupuncture for knee,
hip, and hand osteoarthritis. And also in Germany, it is recommended for the
German health insurance companies for low back pain and knee osteoarthritis.
Okay, so let's move away from chronic pain and have a look at the evidence for mental health
condition. And probably the best condition to choose in terms of we have more evidence or
better quality evidence is depression. So as we did for chronic pain, what we want to do is find
the best high-quality systematic review with a meta-analysis. So for depression, we have one
that was published in 2019 by Armour et al. Now this study had 29 trials and just over 2,000
patients. So we can compare it a little bit to the chronic pain study that we looked at earlier.
That study, the chronic pain, had over 20,000 patients, and here we have just over 2,000.
So you can see that with chronic pain, we generally have stronger evidence, and that's a
question of basically having good numbers. So here we have the results for the Armour et al. study.
As you can see, there were a number of different comparisons, acupuncture versus usual care,
acupuncture versus sham, acupuncture versus SSRIs, that's selective serotonin reuptake
inhibitors, and acupuncture versus psychological interventions. Now for the first three, we can
see that there is an effect size, as in that it was shown that acupuncture is more beneficial
than what it was being compared to. There's a slightly different measure of effect size here,
it's called Hedges-G, but we just really have to consider the same standard assessment of effect
size. So that is 0.5, around 0.5 is moderate, 0.2 would be small, and actually 0.8 is a large effect
size. So there are clear differences that show that acupuncture is beneficial in these conditions.
Acupuncture is also shown to be pretty much the same as psychological interventions. There is a
small effect size, as you can see, 0.12, but actually that is not a statistical significance.
So we don't know whether the difference is due to pure chance. Many of the clinical trials
included in the Armour et al. study were conducted in China. These tend to show acupuncture as being
more beneficial than clinical trials conducted in the West. So we need to think about why this
would be. Firstly, some people have suggested that as acupuncture is part of the culture in
China, it may generate a stronger placebo effect, and that accounts for the difference. However,
the clinical trials in China typically provide more acupuncture at a higher frequency. So a
Chinese trial may provide 40 acupuncture treatments, which are given three times or five times a week,
whereas in the UK, a Western trial, you might have acupuncture once a week, and that you would
have a course of 10 treatments over 10 weeks. So you can see that the intensity and the overall
number of treatments is much higher in China, and that may explain the difference. However,
whatever the reason, we still need to consider whether acupuncture is beneficial for people
living in the UK. So fortunately, we do have a good quality clinical trial that was conducted
by Hugh McPherson and others and published in 2013. The acupuncture delivered in this trial
was provided by British Acupuncture Council members. The trial compared acupuncture and
usual care versus counselling and usual care versus usual care alone. This is a pragmatic
trial. So basically, it's asking a question, which people would ask themselves. You have a choice,
do you want to have acupuncture, counselling, or just continue with the usual care?
This graph illustrates the results from the McPherson et al study. As you can see,
there's a strong similarity between the acupuncture and the counselling. There is
no statistical difference between the two. But what we can say is both of these lead to improvements
when compared to just continuing with usual care alone, especially at three months where there is
a statistical significance. The difference is not maintained at 12 months. However, for acupuncture
in actual practice, what we would normally be expecting is a slightly different pattern of
acupuncture compared to what was delivered in the clinical trial. In the clinical trial,
the course of treatment was given within those three months at an interval of once a week and
then treatment stopped. However, in common everyday practice, what we'd typically be suggesting is
have an intense course of treatment once a week at the beginning and then space it out. So maybe
once a month, once every two months. So if that were done, you might find that there's a stronger
long-term effect. So here is the conclusion that McPherson et al made from their results.
Basically, the acupuncture and counselling were both associated with significantly reduced
depression at three months when compared to usual care alone. In a similar way to as we saw with
chronic pain, as evidence gets stronger like better quality evidence regarding acupuncture
for depression, we find that acupuncture is being included in clinical guidelines. So these two
guidelines include acupuncture for depression. The first one is the American Psychological
Association and the second one, the Canadian Network for Mood and Anxiety Treatments.
So now let's think about the evidence base for acupuncture in the treatment of other conditions.
I started this presentation talking about the pitfalls of the sentence,
there is no evidence that acupuncture works for such and such a condition. Unfortunately,
for many conditions, we do not have the same high-quality, large-scale clinical trials
that we do have for chronic pain. This means we can be less sure of the actual result.
The results, however, sometimes are positive. It's just that we need to consider the evidence
is weak because we need, at the moment, more clinical trials to be conducted.
So this means we can think back over 20 years as well. 20 years ago, had somebody gone to their
acupuncturist for osteoarthritis of the knee or lower back pain, the acupuncturist would
not be able to say to them with certainty that the clinical evidence shows that acupuncture works
for those conditions. But we now know that it does. Because the evidence has improved,
acupuncture has stayed the same. The fancy sentence for this phenomenon is absence of evidence is not
the evidence of absence. There are controversies regarding how the results of acupuncture clinical
trials should be interpreted. Most of these debates centre on the placebo or sham acupuncture
controls. I will be talking about that in more detail in the second presentation of this series.
However, it fundamentally means that researchers and scientists disagree at which point we can say
that the evidence is strong enough to indicate that acupuncture works for a particular condition.
However, according to John Hopkins Medicine, studies have shown that acupuncture is effective
for these conditions. Many of these conditions you will have seen were included in Novick's et al study
such as osteoarthritis, lower back pain and headaches. But there are others such as menstrual
cramps, asthma and fibromyalgia. If you would like to know more about the key pieces of research that
I have been discussing in this presentation, fortunately they are open access which means
that you can go online and download a copy of the original paper and it is free. Similarly,
the guidelines for pain conditions that include acupuncture can be accessed freely online.
The guidelines which include acupuncture for depression are also freely available online.
This publication by McDonald and Jans provides a good overview of a whole variety of conditions.
It assesses the quality of the evidence and it also gives an indication of whether the results
as they are today are positive or negative. Thank you very much for listening to this
presentation. You can find more information about acupuncture and the evidence base for
a variety of conditions on our website. So, please visit the British Acupuncture Council
website www.acupuncture.org.uk
