The evidence on complementary medicine – what works?

View the slides from Dr MacPherson's presentation here.

‘The evidence is beginning to come through for complementary medicine,’ said Dr Hugh MacPherson, senior research fellow in health sciences at the University of York, ‘even at the stringent levels’ demanded by the evidence-based medicine movement.New guidelines from the National Institute for Health and Clinical Excellence (NICE) on low back pain say clinicians should ‘consider offering’ acupuncture, osteopathy and chiropractic, the audience heard. The guidelines were published on Wednesday 27 May.
While the standard allegation thrown at supporters of integrated health is that there is little hard evidence for complementary medicine, the same applies to much conventional practice, Dr MacPherson said.

He cited a study published in the British Medical Journal’s Clinical Evidence Online, showing there was ‘inconclusive evidence’ for 46% of conventional treatments carried out by the NHS, out of 2,500 reviewed. Only 13% were found to be beneficial.

Dr MacPherson, who also practises in acupuncture and Chinese herbal medicine, said the field of integrated health had a ‘huge research agenda ahead,’ in order to build an evidence base.

‘We are starting from a very low level and we have a climate of negativity towards complementary medicine at the moment.’

The current hierarchy of research, with systematic reviews of many trials and randomised controlled trials (RCTs) at the top, emphasises external evidence, and crowds out patient values and clinical expertise, he added.

Yet for the ‘big five’ therapies – acupuncture, homeopathy, herbal medicine, chiropractic and osteopathy – the evidence was much more positive than is generally recognised.

* Out of six reviews of homeopathy trials by the well-regarded, independent Cochrane Collaboration, two were cautiously positive and four inconclusive.

* Eight out of 23 systematic reviews quoted by the Faculty of Homeopathy were positive, the rest inconclusive.

* For acupuncture, nine out of 28 Cochrane reviews were positive or cautiously positive, the rest inconclusive.

* Ten of 28 Cochrane reviews into herbal medicine were positive or cautiously positive, seven inconclusive.

Acupuncture, chiropractic and osteopathy had been shown to be both clinically and cost-effective, judged by the standard measure used by NICE, he added. ‘If it’s cost-effective, it should be introduced into the NHS, says NICE.’

Chiropractic and osteopathy cost just £4,200 for each year of quality life (QALY) gained per patient with low back pain treated, according to a respected RCT, the UK BEAM trial (Back pain Exercise and Manipulation), published in the British Medical Journal. Acupuncture cost just £3,800 on the same measure, in a separate trial also published in the BMJ.

NICE recommends treatments that cost up to £20,000 to £30,000 per QALY.

However, much of the research that currently exists for complementary medicine was conducted outside the UK. ‘We need to look at more and better research, and go beyond just RCTs,’ said Dr MacPherson.

Dr Catherine Zollman, a GP in Bristol and a Foundation Fellow, said RCTs are ‘influencing health policy and spending’. But they do ‘not have all the answers to help with decisions about clinical patients’.

Factors such as the individual patient and practitioner, the effect of the intervention and of being involved in a research trial were ‘different in an RCT to normal clinical practice’, she added.

Dr Zollman, who is also a former Director of Medical Education for the Research Council for Complementary Medicine, asked ‘Are many patients similar to those in a trial? Patients are variable.’

RCTs were a ‘reductive model’, testing one intervention and assuming cause and effect. Complementary therapies often involved complex lifestyle issues, such as changes in nutrition, stress management and exercise, which could not be measured as easily as taking a single pill.

Calling for researchers to use a ‘broader range of research and evaluation techniques’, she said: ‘We should develop open-minded, patient-related outcome measures.’

The current assumed hierarchy of research evidence
1. Systematic reviews and meta-analyses of many trials
2. Randomized controlled trials (RCTs)
3. Cohort studies
4. Case studies.