Off-label drugs

This article from slate.com is the subject of my first response.

This article is itself an explanation of an article that appeared in the journal Child and adolescent psychiatry and mental health. Both articles totally missed a much more important point. It's true even the most experienced child psychiatrists experiment every day when they proscribe medications to children. Each study has a population of 1 and tells us nothing conclusive about what the likely efficacy or side effects of the drug might be on the next child. The solution is not to monitor children more closely. The problem is not that children aren't being closely monitored. The problem is that research is not being done on children. It is extremely hard to get IRB approval to study medications in children. Children are not able to consent to participation in research they can only give "informed assent". Parents must consent for them and ethically this is still viewed as an imperfect solution. Institutional Review Boards (IRBs), which oversee medical research, are concerned about this as well as the obvious potential for serious side effects. The demands that an IRB will tend to make on investigators wishing to do research in children tend to be extremely high. Many researchers are unable to meet those demands or find them too dissuading to attempt research. Unfortunately it is children that pay the price for this. As a result of not sponsoring clinical trials in children every child is instead the subject of research. You have a situation where doctors might as well be guessing, and do not always guess appropriately.

Let me give you an example. When we first started doing transplant surgery patients were given extremely high doses of immunosuppressive drugs to prevent rejection. These drugs carry many serious side effects including decreased bone density, increase vulnerability to infection, skin changes, metabolic changes, aggressivity and other mental status changes. It wasn't until many years later when clinical trials proved that giving as little as one tenth of the old standard dose was just as effective in staving off rejection of new organs, and substantially improve side effects. When doctors guess, even with the best intentions and information, they do not always guess correctly, and the patients pay the price.

Retrospective follow-up studies, which ask people to report how much of a drug or food they ate for a given period of time, instead of clinical trials are one approach that can help. However they come with their own set of experimental problems. They can not truly be said to prove causation (ie: drug X alleviates headaches), but rather can only show association (use of drug X is associated with decreased incidence of headaches). This is a start but many of these studies done in other populations have proved inaccurate. Investigators must anticipate all confounding variables and adjust for them in their analysis and there are often factors that contribute to a finding that are not considered. This approach to studying psychoactive drugs in children could still start to give doctors clearer guidelines in the use of medications.

But ultimately I'd like to stress the need for increased openness to studying medications in children. It's risky, its problematic, but the alternative is so much worse.