(White, 2000) and two-thirds of those receiving treatment for anxiety or depression
report using CAM in the US (Bassman and Uellendahl, 2003). In Europe, 75 per cent
of the French population and 50 per cent of the UK population report having used
CAM (Murcott, 2006). Half of UK general practitioners (family doctors) provide CAM
and more than 80 per cent of Australian general practitioners have referred patients
for CAM (Murcott, 2006). Consequently, CAM has an important role in health care
and many patients seen by psychologists will be using CAM.
The UK House of Lords Select Committee on Science and Technology (House
of Lords, 2000) published a report on CAM, which distinguished between therapies
that do and do not provide diagnoses. The report found that those that do include
osteopathy, chiropractic (which are regulated by UK Acts of Parliament), acupuncture,
herbal medicine and homeopathy; and those that do not include aromatherapy,
hypnotherapy, reflexology and shiatsu. The Select Committee made a series of recom -
mendations including (1) therapies that claim to treat specific conditions should have
evidence of being able to do above and beyond the placebo effect (see below); (2) if
a therapy does gain a critical mass of evidence to support its efficacy the UK National
Health Service should provide access to it; (3) training in anatomy, physiology,
biochemistry and pharmacology should be included within the education of CAM
practitioners likely to offer diagnostic information; (4) CAM therapists should be trained
in research methodology and have a clear understanding of the principles of evidence-
based medicine; and (5) CAM therapists should encourage patients to see traditional
health care professionals. These recommendations highlight the challenges of, on the
one hand, providing access to effective treatments based on alternative models of health
and, on the other, ensuring that such treatments are indeed effective.
Placebo effects
Placebo effects refer to health or well-being gains observed following administration
of pharmacologically inert interventions such as saline injections or sugar pills (see
Kaptchuk, 2009 for a brief and interesting historical perspective). In 1955, Beecher
found that, across 15 clinical trials, 35 per cent of patients showed health gains in
placebo conditions. In contrast, across 114 trails, Hróbjartsson and Gøtzsche (2001)
found that when improvement was measured using a binary outcome (e.g. cured not
cured) placebo treatments had no significant effect on outcome. Placebo conditions
also showed no benefit when assessed using objective clinical outcomes. Placebo effects
were observed on subjective, continuous outcome measures and in 27 trials assessing
pain. Average placebo pain reduction was found to be equivalent to a 6.5 millimetre
reduction on a self-report visual-analogue scale, measured on a 100 millimetre line.
One weakness of such a review is that, if placebo effects are limited to particular types
of health gain (such as reduced pain), these effects may not be evident when trials are
pooled across conditions (Stewart-Williams, 2004).
Main effects of adherence have also been found in placebo conditions (Epstein
1984). For example, in a trial of beta blockers for women who had had a heart attack,
Gallagher, Viscoli and Horwitz (1993) found that 5.6 per cent of those who took 75
per cent of the medication died within 26 months but 13.6 per cent of those who
took less than 75 per cent died in the same period. Remarkably, this difference was
not noticeably diminished in the placebo condition. In this case taking beta blockers
244 RELATING TO PATIENTS