I think one has to talk about the different patient populations. When we're talking about very small patient populations—hemophiliacs, or people who have other kinds of conditions—patient registries become really important. Again, you're dealing with very small databases, you're dealing with a lot of uncertainty in terms of outcomes. You get down to, in many cases, very individualized responses to drugs and treatments.
So having a comprehensive patient registry in which ongoing information is collected.... For instance, every month hemophiliacs load their information, talk on electronic mikes, and send it off to the clinic, the database. So all that information is constantly collected.
That's different from, as you say, a large-scale public drug in which people may be taking it who are not necessarily part of a specific patient population.
I think the most important thing in order to encourage people to report is actually to give them good feedback. If in fact I get feedback in terms of what my report is relative to what others have done, if I get examples in terms of what other kinds of reactions or what evidence of efficacy there is, and if we get the good information back that can summarize this, that is probably the best encouragement in terms of reporting.
I'm really willing to do some very onerous reporting. Especially, I will go back to some of our rare disease patient populations, where there are a lot of adverse reactions that we have no clue to.