Yes. I don't know the details of this study, but I think it's a perfect example of how one needs to balance the benefits and harms of medications in a rational way.
To address your question about whether there are any other downsides to some of the post-marketing surveillance, the fact of the matter is that analyzing these databases that Steven and I have talked about is relatively cheap, because they exist. It's not like a random trial where you have to randomize hundreds of patients.
We need to be careful. There is a tendency, for example, for the pharmaceutical industry to want to mine these databases until they find the result that makes their drug look terrific, publish that, and not tell you that they did 25 studies before they found the one they really wanted. Similarly, some people who might have an axe to grind with the pharmaceutical industry might search these databases 25 times and only report the one that shows harm.
I think it is absolutely crucial that we have a network of these researchers, and that we not make any major decisions such as pulling a drug off the market on the basis of one of these post-marketing studies, but rather on consistent and reproducible evidence from different jurisdictions.