The FDA is likely to approve an implant that stimulates a portion of the brain as a treatment for depression.

The device which is already in use to treat severe epilepsy, is supposed to help moderate depresison through stimulation of the vagus nerve.

However, what I found most interesting about the article, were the quotes from the FDA panel, specifically those by A. John Rush who was testifying for Cyberonics Inc, the group hoping the expand the market for their product.

“We lost four of these individuals in the last 2 1/2 hours,”…basing that figure on the high suicide rate among patients with resistant depression.
…
“In the time to do another trial, we will lose another 1,000 patients a month, 36,000 if the trial takes three years.”

I’m not sure what else was said to the panel, but it seems to me that those statments are deliberately misleading, especially for a treatment whose randomized trial didn’t show statistically significant results. 15% improvement in the treatment group versus 9% improvement in the control group is not impressive, and certainly would not correlate to the saving of “1000 patients a month” as Rush seems to be implying.

Which reminds me: several medical groups, including the AMA, are pressing for a database of clinical drug trials. Much of this call comes after GlaxoSmithKlein was accused of supressing the results of clinical trials for Paxil in adolescents

This comes back to a long standing problem in research, which is that negative results are rarely published, for a variety of reasons. Firstly, it isn’t very interesting or exciting to say “we didn’t find anything”; such results are rarely going to be published in the leading journals. Secondly, if the trial or research was sponsored by a specific company for a specific product, that company is not going to want information published saying their product was ineffective.

The other option, which has been discussed, is the creation of a journal for publishing negative results. However, a database, preferably one available on-line, would be a good idea, because it would make knowledge of those trials widely available.