When a medication is being evaluated to modify the behavior of a person with autism, one must assess the risks versus the benefits. The benefits of the medication must outweigh the risks. Some medications can damage the nervous system and other internal organs, such as the liver. These risks are greatest in young children because an immature nervous system may be more sensitive to harmful side effects. A good general principle is that the use of powerful drugs should be avoided in young children when the risk is great. The younger the child the greater the risk. For example, it would be justified to give a young child Prozac to stop severe self-injury, but it would probably not be justified if the only effect was that it made him slightly calmer. If a medication improved language, its use would probably be recommended.
The brain of a teenager or an adult is fully formed and the risk is less. Many teenagers and adults with autism may benefit from Prozac or Zoloft. See my book, Thinking in Pictures, or other papers I have written for this Internet web site (http://www.autism.org/). There is a possibility in some cases that if too many drugs are given to young children they may not work when the child needs them when he becomes a teenager. This may be a problem especially with drugs such as Haldol or other neuroleptics. Practical experience has shown that the nutritional supplement DMG is safe for young children.
A medication that works to change behavior should have an obvious and dramatic effect. One of the best ways to evaluate a medication is a blind evaluation. A simple way to do this is to start the medication and do NOT tell the teacher at school. If the teacher says "WOW, your son's behavior has improved remarkably," then you know that the medication works. To evaluate a medication, it is important that the other therapies are not changed at the same time. Change only one thing at a time so you can see the effects. A new medication cannot be properly evaluated if the child goes to a new school around the same time that the medication is tried. If a medication does not show enough benefit to outweigh the risk then you should get rid of it. Medications should work. If the change is not obvious and is not dramatic, it probably is not worth giving the medication. It is also important to start only one medication at a time so that its effects can be evaluated.