The US Food and Drug Administration (FDA) will meet next week to consider how to calm the firestorm over the use of antidepressant medications in children (see page 122). It is now armed with much more material than it had at its last meeting on selective serotonin reuptake inhibitors (SSRIs) in February. None of these data have altered the conclusions drawn by FDA epidemiologist Andrew Mosholder ahead of the February meeting, however. As a class, SSRIs are associated with an increased risk of suicidal thoughts and behaviours in young people.

The FDA is reluctant to discourage the use of SSRIs, because there are few good treatments for childhood depression, which can by itself lead to suicide. But there is little to suggest that children are better off on the drugs than not. Only one of the drugs, Prozac, has consistently improved childhood depression more than placebo treatment.

The FDA needs to make a statement reflecting these facts, to counteract the long holiday that drug companies have enjoyed. Crucial data about the lack of efficacy and side effects of SSRIs in children have not been clearly described or even published in medical journals. Without access to the full findings, US doctors wrote 10.8 million prescriptions for the drugs for children and adolescents in 2002. The FDA is responsible for stopping this widespread ‘off-label’ use.

Drug companies are now allowed to market their products directly to US consumers. But they do not reveal all the information they have about these drugs to potential patients. A new method of disclosure must be created that is as understandable as the TV and magazine ads that spread rosy messages about today's blockbuster drugs.

The drug companies have begun to respond to such demands by posting information about approved drugs on their websites. But the industry's reputation is so tarnished that this is unlikely to be sufficient. The FDA and other agencies should collaborate to develop a clinical-trials database containing information on every clinical trial sponsored by a drug company. It should include all the clinical results from drugs that have been approved for any indication. Such a move would go some way towards restoring the FDA's dented credibility.