Sure. Good questions.
There are two points. One is the incomplete nature of reporting, and the other is the quality of the reports, which is what we're talking about.
So the first is incomplete reporting. Why don't people report? There are lots of reasons that people have cited. I believe, after having studied this for 20 years, that the reason people don't report is that there's no reason to report. What happens if I report an adverse reaction? Health Canada already knows about that one. This is well accepted. Those are some of the things that I hear from clinicians. They're not reporting reactions because they recognize that somebody already knows about them. They don't understand the purpose of surveillance. The purpose of surveillance isn't to uncover whether a drug is responsible. It really should be to uncover whether there's a solution to the drug safety problem that can be produced as a result.
In 2004, a group of researchers in Toronto in Canada found that for toxic epidermal necrolysis, the most serious form of skin reactions that occur from drugs like Bactrim or cotrimoxazole, actually where the skin turns black and falls off in very large sheets--it's largely fatal in most cases, an awful condition--most of the patients end up in Canadian burn centres. There is almost no non-drug reason for that particular reaction to occur. So they went to Canadian burn centres and said, “How many of these cases have you treated in the last five years?” I think that was the length of time. They found only 4% of those cases were actually reported to Health Canada.
So I think part of the problem is that people don't see a reason to report in terms of helping them find a solution to these problems. It's just another report of a reaction that's idiosyncratic, from the Greek idios meaning “compound” or “mixture” of reasons. We don't know why. So I want to find a reason why.
The second question you asked was about the quality of reports. Why are the reports that come in not complete? And it's not just my reports. It's the ones in Health Canada, the ones I've seen through the Canadian pediatric surveillance program, the ones I've seen for the network I've built in the United States. The reason they're not complete is the physicians tell you the information they think you want to know. So if they think the purpose of reporting is to link drug and reaction, they'll report “Paxil: suicidal thinking”, and then sign their name. They don't provide all of the other information that's useful to determine the biological reason for the response, because they don't understand how you're going to use that information or how it's being used.
So I think that's part of the problem. We have to do more than just sell clinicians on reporting for the public good. We have to show them why it's useful--because we can get solutions.
I found when I built the network across the country that oncologists didn't want to participate in it initially. But once you start showing high odds ratios, high risks associated with a particular gene trait, they start to listen. The arms become uncrossed and they begin to listen and move to the edge of their chair and provide funding. Once they see a solution possibility, they begin to report. So I think those are the reasons why.