Conventionally, the best evidence for an intervention is that for which there is scientific support, which means that evidence of its efficacy and safety is derived from well-controlled clinical trials. Using such evidence as the basis for making rational decisions about medicine use is logical and seems straightforward. Yet, there are several problems with relying on formal evidence alone.

Problems with evidence as a basis for rational use

• Evidence may not be available. In reality, much of what is done in medicine is not based on evidence, simply because it is not available. This is often the case for cheaper non-drug interventions. Most clinical trial ‘evidence’ relates to drugs and is mainly generated by pharmaceutical companies (e.g. there are thousands of published clinical trials of nonsteroidal anti-inflammatory drugs (NSAIDs) in osteoarthritis but hardly any on the use of walking sticks).

• Clinical trials are often focused on proving efficacy for drug registration instead of drug effectiveness for actual ‘real life’ use (see Chapter 8).

• There may be poor access to evidence and a lack of openness from regulatory authorities and pharmaceutical companies who hold such information.

• Promotion of medicines to health professionals and consumers by pharmaceutical companies is often misleading.

• Patients included in trials that make up the evidence may not be representative of ‘real’ patients, for example, because the trials were carried out in a different country, or the patients in the trial had no co-morbidities.

• Health professionals and patients do not necessarily behave like those involved in the trials. For example, they might be less likely to give patients detailed and motivating instructions, or to perform the specified regular checks on whether the treatment is being effective (e.g. blood pressure measurements) or causing adverse effects (e.g. liver function tests).