Why drug trials go wrong

I've written in this blog before about medications causing adverse effects -- sometimes to the point of needing to be withdrawn from the market.

Well, a team of Toronto researchers might have found some of the reason why: The people who end up having problems aren't being used in drug trials.

Robert Fowler, a doctor at Sunnybrook Health Sciences Centre, teamed up with a few colleagues to look at drug testing methods in the wake of pharmaceutical companies being forced to remove new medications from the market after discovering adverse side effects. He presented the findings at Toronto’s Joint Centre for Bioethics recently and spoke with me afterwards about his research.

Fowler and the others found that drug companies were deliberately excluding from their studies the very people who are often the most likely to have adverse side effects from medication: the very young, the very old and young adult women.

The people who end up getting into trouble are those who weren’t included in the study."

He said drug companies have good reasons for being selective. Older people are more likely to be taking other medications that might interfere with what they are testing, and using children in clinical trials for experimental drugs is frowned upon, as is the use of women in child-bearing years.

It’s a reasonable way to set up a trial."

Fowler, an assistant professor of medicine at the University of Toronto whose study was published last year in the Journal of the American Medical Association, said the problem is that once the trials are completed, the results get picked up by a medical journal and doctors start writing prescriptions for all types of patients — including those excluded from the testing.

That’s when the trouble starts, Fowler said, and it’s something he’s seen first hand.

After digging through almost 5,000 studies, they found that, not surprisingly, the type of people who tended to have problems were never part of the original testing.

Fowler and the others said studies should specify how test subjects were selected. Fowler has already begun to see such information in medical journals, which helps doctors decide whether a patient will benefit from a new drug.

A further step, Fowler said, would be for drugs to get only limited approval after a clinical trial that excludes large groups of people, with full approval coming only after wider tests are done. Right now, he said, “there is no appetite and no funding” for such further studies.