The FDA will now require the manufacturers of antidepressant drugs to include a new black box warning about the risks of suicide. This is important information for physicians and patients but bound to cause a lot of confusion and worry.

A pediatric panel strongly recommended the FDA take this action after reviewing studies on the medications' use in children which demonstrated their lack of effectiveness in childhood depression and a doubling of the rate of suicide on the drug compared to placebo treatment.

I was at the FDA meetings in February and September. I'm sure the panel was also affected as I was by the dramatic tragic testimony of family after family whose children died while taking the drugs. But many psychiatrists are concerned that this group of medications, they still consider quite effective, will be refused by a multitude of children who could potentially benefit from the drug. What should a parent believe and should they allow their child to be medicated?

The first thing parents should know is that psychiatric drug prescription for children's depression is much more "art" than "science." Therefore, knowing the biases of the doctor whose advice you're following is quite important. Most child psychiatrists these days are primarily prescribing medication so there will be a tilt towards their recommending those kinds of interventions. Virtually all the psychiatric researchers who are quoted in the newspaper or television have studies funded by the drug companies or are paid consultants.

I prescribe psychiatric medication to children on virtually a daily basis but because I publicly challenge the use and overuse of these drugs to children in our country I have never accepted drug money. While it means I fly to professional meetings in coach instead of business class, I can't afford the risk to my credibility by accepting money from the drug companies, if I also suggest that medication can be effectively and safely used in children.

The anti-depressants and here I'm referring primarily to Prozac and her sisters, the SSRIs, which include Zoloft, Paxil, Celexa, Lexapro, and Luvox, have never impressed me as being particularly effective in treating childhood depression, say compared to the effects Ritalin has on ADHD. This is borne out by the drug company studies of 4600 kids reviewed at the FDA meeting of which only three of fifteen showed the active drug (it was Prozac) more effective than the placebo.

But not so fast. It isn't easy demonstrating a positive effect on childhood depression because the placebo cure is sixty to seventy percent in most studies. Any drug would have to be superior to that already high rate in order to prove its effectiveness. Just last week, this point came back to me in my own practice. A twelve year old girl was remarkably improved two weeks after I prescribed her Prozac for what appeared to be a depression with hallucinations (she was hearing voices). She attributed her better mood and elimination of the voices to her taking the medication. I was very impressed. But then later her parents called to remind me that I'd forgotten they had decided to give her a vitamin pill instead of the medication because they were nervous about the Prozac. One more cure for placebo.

So if a psychiatrist is convinced anti-depressants work, the science for the moment doesn't support that position. However, the drug could still be used in children. Evidence for Prozac's effectiveness in treating childhood anxiety is much stronger. And even using the drug for depression could still be justified but expectations should be realistic.

I would still consider using these medications for teenagers who are actively suicidal or have tried to kill themselves recently. Teens who self mutilate would also seem sufficiently impaired so as to risk the possibility that medication could be helpful. These children should be followed closely with visits and phone check ups in the first weeks of treatment and any time there is a increase in the dosage.

The most important take home message though is about the side effects. Prior to the FDA warning, if a child started an anti-depressant and complained of feeling worse, anxious, agitated, extremely restless, thinking thoughts never thought before (like killing himself or his parents), the doctor was likely to reassure the patient and his parents. "Hang in there. It's your depression. The medication takes two to four weeks to work. You'll feel better soon." Sometimes the doctor would increase the dose or even add a second medication.

STOP! That's the message from the FDA. If in the first days or weeks of treatment the patient is doing worse stop the medication -- there is a small group (about three percent) of kids who seem worse off on the medication. It may be necessary to hospitalize the child (on no medication) until the drug effects wear off. One could try another medication then but only under very tight supervision and perhaps treating the child without any medication may be the best course for that child.

Oh, if your kid is already taking the medication and is "doing better" you don't have to stop (though it may only be the placebo effect). However, if you plan to stop the medication don't do it abruptly. Turns out despite years of denial from the drug companies and their professional representatives there is a mild to moderate withdrawal from the SSRIs (mostly headache and feeling lousy). It isn't the depression returning as claimed for many years. Just ask the babies who were jittery and fed poorly for two weeks in the study of newborns whose mothers had taken Paxil during pregnancy. I don't think these babies were depressed.