By accumulating them.
We have other approaches. For example, we're working with groups such as the Canadian Medical Association and the Canadian Institute for Health Information in terms of hopefully, in the nearer future, being able to leverage reports without relying on, or waiting for there to be, a voluntary report.
In terms of your point about manufacturers' reports, although 66% of the reports that we get actually come from manufacturers, they get them from health professionals or consumers. So, for example, if there's a 1-800 number on a label and a consumer calls the manufacturer, then the manufacturer has to report it. In other words, the mandatory component is that anything the manufacturer is aware of they have to tell us, no matter where they get it. If they hear about it themselves, or if a physician, a nurse, a dentist, a naturopath or a consumer tells them, they have to make a report.
The problem is that the level of quality may be extensively different, depending on how close they were to the situation. If one were a salesperson in a doctor's office who got a full case report from that physician and then sent it to us, obviously that may be more useful than if we just know that somebody had a rash, but didn't give you a dose, and there's no temporal relationship to when the drug was taken. That's less useful. We would get the full range of that in the approximately 17,000 domestic reports we get per year and 350,000 foreign reports per year. So it becomes a trending exercise.
Then once we've identified what we call a signal, and they're potential signals, then those are further investigated. That may involve actually doing a post-market study. That may involve talking to foreign regulatory bodies. That may involve including academia or provinces and territories looking at utilization information. That may involve purchasing utilization data from groups like IMS, or Brogan Inc., and others who have that kind of information, to be able to put the full picture together.