Drugs - the ones you get from doctors - haven't had a great press recently. From some antidepressants making troubled children worse, to Alzheimer's disease remedies that might not do all that is claimed, success stories are thin on the ground.

So presumably the UK Clinical Research Collaboration's initiative to increase the number of patients taking part in clinical trials has to be a good thing? Surely the more people who join trials before drugs are used more widely, or indeed before a decision is made to stop using them, the better? As an Association of the British Pharmaceutical Industry (ABPI) factsheet states: "Clinical trials mean that NHS patients have potential early access to the newest forms of treatment together with the highest standards of medical care."

In other words, join a trial, and you will get the best new treatments, Rolls-Royce standards of care and do your bit for medical knowledge. Quite possibly not, on all counts.

First, you won't necessarily get the newest treatments in a trial. You may get the placebo, a "dummy" with which the test drug is compared or, if you get the real thing, it may be worse than nothing, or worse than what you could have had without being on the trial. Indeed, it's such uncertainty that makes trials essential.

Second, it's a medical mantra, supported by logic and anecdotes, that people who join trials do better than others because they get regular check-ups and care from doctors with time to chat. But hard evidence for such benefit simply isn't there.

So if the hopes of getting newer, better pills, or better care, are not served by joining trials, can you fall back on the comforting knowledge that you're doing something for future patients - that the next person who gets ill will fare better because you've tested the tablets?

The answer is no because there is no requirement to make trial results public. As Sir Iain Chalmers, a proponent of evidence-based patient choice explains: "Patients who agree to participate in controlled trials assume they are contributing to knowledge, but disappointing results may never see the light of day: researchers can report only those results that show test drugs in a good light. The law doesn't oblige companies to disclose the findings of their research, and scientists, doctors, patients and even public organisations currently have no legal right to inspect the evidence that leads to a drug being licensed."

Chalmers edits a website, the James Lind Library, which explains why carefully controlled treatment trials are needed and why incorrect conclusions can be made about effectiveness. And he does not pull punches about the impact of secrecy, saying "Biased under-reporting of clinical trials kills patients and wastes money."

While a firm believer in the importance of trials, he is campaigning to ensure they happen within a transparent registration and reporting framework. "Until 1990, heart attack sufferers were often routinely given drugs to prevent heart-rhythm abnormalities. There was evidence suggesting this might not be a good idea, but it remained unpublished. It is estimated that these drugs killed more Americans annually than died in the Vietnam war. Had evidence suggesting the drugs were lethal been published, many of these deaths would have been prevented."

Commercial or political drivers behind keeping results under wraps are painfully apparent. Chalmers was asked by the department of health, in 1994, to advise on who could review the evidence of effectiveness of evening primrose oil for treating eczema, an approach that was costing the NHS more than £7m annually. He recommended distinguished professors of dermatology and pharmacy, who considered published and unpublished results and found no evidence that the treatment was effective. Chalmers says: "The manufacturers responded by asking the department of health not to disclose or discuss the contents of the review without their agreement. Incredibly, officials accepted these demands and evening primrose oil remained on the NHS drug list."

Although it was eventually removed, five years later, Chalmers adds: "The raw results underlying such decisions have still not been published."

Given such cloak-and-dagger dealings, why might you or I join a trial? The simple answer is because they are the best way to find out whether a treatment or procedure does more good than harm. Without controlled trials, and dissemination of results to doctors and patients, a doctor may prescribe a drug simply on the basis that he or she has used it for 30 years, or because it's brand new or because the company that makes it sends the glossiest Christmas cards and gives away the best pens.

Trials are not always for patients' benefit. Dr Andrew Herxheimer devotes much of his working life to collating experiences of illness and pushing for adverse treatment effects to be taken seriously. He says, before joining a trial, we should ask "In whose interest is this?"

"The trial may be aimed at finding out what works best for a specific problem, or just for a drug company to get marketing approval," he adds.

Dr Marcia Kelson, director of the national institute of clinical excellence (Nice) patient involvement unit, agrees that knowing who a trial serves is important. Speaking in a personal capacity, she says: "It's well and good to say get more patients on to trials, but they need careful thought. They must be designed so that if it will take 20,000 patients to answer a question, you won't get stuck having recruited 2,000, unable to get more.

"And they need to be ethical. Patients need reassurance that declining to join a trial or pulling out, won't affect their care. They must have a balanced explanation of what is involved so they can understand what a trial involves and make an informed choice about participating."

In addition to ensuring that trials reflect patients' as well as doctors' priorities, Kelson says more are needed for things that might make real differences to people's lives, but not make money for drug companies: "There's sometimes a bias towards drug-related research at the expense of lifestyle change, complementary therapies or talking treatments, because these aren't generally what industry backs."

Professor Jeremy Wyatt, Nice's associate director for research and development, is also focused on what matters to patients. Asked where those who want to know about current, past and future trials can find out more, he says: "It's a ridiculous situation for the consumer. There's a real need to bring them together with self-help groups, funders and the NHS so all parties can easily find out about trials and all their results, not just the good ones."

Surely Wyatt, and Nice, as a key player in the government's healthcare armoury, are in the right position to make this happen? "Sadly not," Wyatt says "We have influence to make recommendations, but lack money to push them forward." And others fear that up against the industry, Nice has little clout.

So where does that leave you, if the doctor suggests you might do well on a new drug but that you would be taking it - or possibly an inactive placebo - as part of a trial? According to Herxheimer, this is where patient power must be pulled: "You must make it clear that your joining the trial is contingent on it being done properly. This means knowing how to get full details of it and having any parts you don't understand explained, that the trial is registered and its results openly reported.

"There are other awkward questions you might ask too, like whether the doctor is being paid to recruit you. If the answer is yes, there may be good reasons, such as that they are getting a fee for doing the paperwork because it is beyond their normal job. But does this mean they will also be tempted to recruit people who aren't suitable? It happens."

· Sophie Petit-Zeman's book, on patient and professional perspectives on healthcare, will be published by Routledge next year.