Assessing how well an epidural controls pain during labour means looking at its duration, the mode of delivery and various hormones; but in only two out of nine studies that were reviewed in 1992 were women asked how they felt. Similarly, in the review of a drug for schizophrenia, it took a question from the father of a woman affected (what, he asked, was the meaning of a 20 per cent shift on an apparently important scale?) before researchers admitted that one measure was uninterpretable. They added: "The lay input there clearly focused us on what was really important and changed the tone of the whole review."
I forgive purists who say "prove patient input helps", but it must make sense to widen research beyond researchers. If I need an epidural, I would prefer to know that other women had found it useful than just that someone had once measured it working in a lab.
Involving potential beneficiaries in research helps ensure that it happens with, not to, them. Through a national advisory group, Involve (www.invo.org.uk), the National Health Service aims to improve how research is prioritised, commissioned, conducted, communicated and used. According to its website: "Research which reflects the needs and views of the public is more likely to produce results that can be used to improve practice in health and social care."
But defining "public" and "patient" is not easy. While the dividing line is thin (today's public, tomorrow's patients), there is a difference between public interest in science and enabling patients and carers - those at the sharp end of research - to have a say.
Funders want to support both public and patient involvement. The Biotechnology and Biological Sciences Research Council strategy panel includes non-scientists and consults widely during priority setting. And while the Medical Research Council's advisory group on public involvement steers clear of patients, the MRC (with the Department of Health) backs the James Lind Alliance (www.lindalliance.org) to support patients and clinicians shaping research together.
Patients are integral to many charities. Some of the 112 members of the Association of Medical Research Charities were set up by patients, others have patient trustees, and almost all rely heavily on public money.
Uniting such groups, the UK Clinical Research Collaboration has allocated £74 million to experimental medicine. As it is defined as "investigation undertaken in human beings", patient involvement is essential. But definitions of "involvement" are broad. Sometimes it means patients doing research - but this is tricky, and many patients want to influence research but not to do it.
Meanwhile, arguments about terminology simmer. Some hate "patient", others find "user" confusing, and as for "consumer" - do we eat healthcare? Isn't it curiously derogatory that when professionals determine the research agenda it is termed "top down", while public or patients having a say is the altogether less appealing "bottom up"?
Whatever it is called, the challenging task of melding medical science with an individual's complex wishes and experiences makes sense. A website, the Database of Individual Patient Experiences (www.dipex.org), tries to do just that. Its medical director, GP Ann McPherson, is pushing for patient experience to be included in medical textbooks.
As DIPEx unites researchers, their assessment tools, doctors, their prescription pads, policymakers and patients, all should heed consumer advocate Hilda Bastian, who once said: "Researchers cannot assume that their own values and priorities apply to others who do not share their world."
Sophie Petit-Zeman is director of public dialogue at the Association of Medical Research Charities
Sophie Petit-Zeman's Doctor, What's Wrong? Making the NHS Human Again will be published by Routledge on June 21 (£12.99).