Vaccines, just like drugs, have major adverse effects associated with their use, and death, of course, is the ultimate adverse event. What I would suggest is that if we want to have an effective post-market surveillance system, that means that when we actually provide the vaccine we collect data on the outcomes of interest and we collect it from both people who have reactions and people who don't have reactions.
In our case, I specifically went to pediatric hospitals and said “How many cases of cisplatin ototoxicity do you have?” They said they didn't know, they'd never looked at the hearing loss with cisplatin in a comprehensive way. So we began to uncover these cases. They're there, they're not hidden. And I would suggest that the same thing could be done with this HPV vaccine.
When women and girls are given this vaccine, their outcomes can be studied and the data can be collected systematically. We could have a sample of several thousand very soon. In my case, we collect biomaterial--saliva, or blood, or buccal swab--and we can look at whether the genes play a role in an adverse outcome in people who have the adverse effect versus those who don't. That's one way that we could deal with that issue.
Death occurs regardless of whether people get vaccinated or receive drugs. The question of causality is critical, and how to determine causality is extremely difficult. It's much easier to do when you can look at data from two sets of people who receive the vaccine--not just from people who had adverse reactions, but also people who didn't. That's what I would suggest would be a good approach in Canada.