Financial Times- In his acclaimed book Hippocratic Oaths, the physician and philosopher Raymond Tallis describes Sir Iain Chalmers as a doctor who "has most certainly saved more lives than any other doctor I have met".

Tallis was referring to Sir Iain's role in co-founding the Cochrane Collaboration, an international organisation that seeks out, assesses and synthesises every last bit of evidence about the effects of treatments. But in his current role as the editor of the James Lind Library, the 62-year-old continues to focus on how to reconcile the clinical research priorities of researchers, clinicians and patients.

He believes that a lot of research doesn't have to be carried out at all because it has already been done. The problem is that the results haven't been read, searched for, or drawn together in what is known as a systematic review. "The very first thing that all researchers should do is review systematically what is currently known, otherwise they risk doing useless - even dangerous - research," he says. A failure to systematically assess the results of pre-clinical research in animals, for example, has resulted in thousands of patients being given a useless treatment for stroke - nimodipine.

So who should set the research agenda? Should it be patients, not researchers, in the driving seat? "Inquisitive scientists should not be discouraged," he says. "Important things have come from that approach. But on the basis of the evidence available, we do have a mismatch in what patients and researchers think is important."

This mismatch has led Sir Iain to collaborate with Dr Imogen Evans and patient Hazel Thornton on a book, Testing Treatments, which is due to be published by the British Library early next year. I have had a preview of it and can report that it is an eloquent and impassioned plea for better, rather than more, research. It also calls for full admission of uncertainty about the effect of so much of what doctors do and recommend. It argues, too, for a true partnership between doctors and patients in deciding which unanswered questions need to be addressed.

I put it to Sir Iain that the problem with taking patient involvement seriously is going to be that it will be the most educated and capable patients who will be heard loudest. This point is not lost on him. "Take patients who have learning disabilities as well as a psychotic illness. What's the best treatment for them? If you look for relevant evidence, you find a single controlled trial, done in 1954. How do we do right by these people and their carers? These patients can't give informed consent."

The lack of good evidence about treatment for such patient groups prevents doctors from being certain that they are offering useful treatments. Instead, because of the problem in obtaining "informed consent", vulnerable groups of patients often get ignored when it comes to doing research. Yet the alternative, of not doing trials of treatment in this group, means never knowing whether the treatments we offer are doing more harm than good.

Sir Iain became interested in evidence about treatments soon after he qualified in medicine, when he worked for two years for the United Nations in a Palestinian refugee camp in Gaza from 1969. He felt that some of what he was doing resulted in the unnecessary deaths of his patients because it was based on what he had been taught, rather than on reliable evidence.

He returned to the UK and, after more clinical work in obstetrics, proceeded to train further in social medicine at the London School of Hygiene and Tropical Medicine and at the London School of Economics. He then became a full-time health services researcher, particularly into the effects of healthcare, before founding and directing the National Perinatal Epidemiology Unit, which pioneered the collation, coding and systematic review of most of the controlled trials in the perinatal field. This was not a small task - it took 15 years - but when the result was published it was the first truly comprehensive collection of therapeutic research in a field of health care and became a model of development for the Cochrane Collaboration. He was knighted in 2000 for services to healthcare.

It is plain that he is furious about the persisting biases in medical research and the continued promotion of inadequately tested treatments. Even in randomised controlled trials - often seen as a "talisman" of research - there can be serious problems.

"We know, for example, that the results of industry-sponsored controlled trials are more likely to favour the treatment they are trying to sell," he says. "It comes as a deep disappointment to me that so many of our co-professionals working in clinical research have sold out to the interests of industry, and they have actually been encouraged to do so by universities and by successive governments. Where on earth are you meant to look for independent research about treatments these days?

"The only hope is if the public, after being presented with the facts, is encouraged to get extremely angry about this sad state of affairs."

My own view is that Testing Treatments offers a great deal of hope that things will improve. Yet despite my attempts at persuasion, he insists that he doesn't want a revolution in how research is done.

"I just want us to find questions that are pertinent to patients and clinicians and encourage funding of and participation in studies addressing these questions. That's the hope of the James Lind Alliance. That's all."

But regaining and redirecting the juggernaut of medical research for greater patient benefit would be a sort of revolutionary step. I wave him off at the station and realise that such a modest man would never admit to having such radical thoughts. But radical they are, and patients and doctors need them.

www.lindalliance.org

www.cochrane.org

Margaret McCartney is a GP in Glasgow

margaret.mccartney@ft.com

Read more columns at www.ft.com/mccartney