Widening Research Methods

We must widen research methods: BMJ editorial

The gold standard scientific research tool to test how drugs affect a single, well defined disease is the randomised control trial (or RCT). But this has worrying limitations, not least because patients are often excluded from RCT trials if they have multiple problems or are already using other treatments. These limitations especially apply to people with long term conditions.

Professor Peters and his co-authors, Dr Hugh MacPherson and Dr Catherine Zollman – all Fellows of the Prince's Foundation for Integrated Health – say that there is often insufficient research evidence available to support every day treatment decisions, particularly where integrated medicine and complementary therapies are involved. Yet these approaches may prove to be particularly relevant in treating chronic diseases, which account for 78% of NHS costs.

The answer is to encourage other techniques. Randomisation and control groups can still provide rigour and reduce bias, even when collecting evidence from typical populations receiving treatment in ways that reflect normal medical practice. Observational studies and basic research into mechanisms of action are valuable approaches too. Narrative research and patient reported outcome measures take what patients have to say about their health seriously. The notion of researching with people rather than on them has much to recommend it.

That view was recently backed by NICE Chairman Professor St Michael Rawlings. Talking about hierarchies of evidence – where the classic randomised trial is at the top and reviews of groups of such trials even higher – he said that they were "an attempt to replace judgement with oversimplistic, pseudo-quantitative assessment of the quality of evidence".

Closing the evidence gap in integrative medicine, BMJ 2009; 339: b3335, was published by the British Medical Journal 1st September 2009 http://www.bmj.com/content/339/bmj.b3335