Great post from In the Pipeline on anti-psychotics:

The authors compare reported trials of first- and second-generation antipsychotics, looking to see if potentially biasing factors have skewed the results. One (perhaps surprising) result is that the authors couldn't confirm that the newer drugs necessarily work better through showing fewer extrapyramidal side effects (those are the muscle and coordination problems seen with many drugs in this class). While they may well show fewer EPS problems, that doesn't seem to be related to their efficacy.

Something of a relief is that the efficacy of the various drugs didn't seem to be related to whether or not the drug industry sponsored the trials involved. Given the publication bias of submitting favorable results (and given the obvious commercial interests involved), that's perhaps surprising. But it's welcome data to bring up the next time someone e-mails me about the eeevil Pharma companies and their bought-and-paid-for studies. I don't get a steady stream of that stuff, fortunately, but it still shows up often enough.

Frankly, I'm surprised too. On the other hand, I guess I should have expected that there would be publication bias in studies showing publication bias.

Meanwhile, this is just amusing:

I still keep an occasional eye on the antipsychotic drugs, since that was the first therapeutic area I ever worked in when I joined the industry. The project came to a bad end, which was probably a good thing for my professional development. We took the drug into Phase I, gave substantial doses to normal volunteers, and rejoiced when it did nothing to them whatsoever. Then the compound went into Phase II and into real schizophrenics, and it did nothing whatsoever to them either, sad to say. And so it goes in CNS drug development.

I mean, amusing if, like me, you aren't in CNS drug development.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.