There were a few different pieces in there. I'll start with the breast cancer item.
All of the cancer prevalence or cancer incidence-type studies require a large population set in order to have reportable results with any confidence. The cohort of veterans who served in the Persian Gulf, for example, was a fairly small number. It's only around 5,000 to start with. If you isolate that down to the women veterans who were part of it, it's a much smaller number and it becomes very difficult to produce reliable or robust statistical estimates about cancer prevalence and comparisons against the population. There's that caveat.
I will say we are now able—again thanks to the census—to conduct a data linkage study, which we're now just embarking on, looking at cancer prevalence among veterans. That's thanks to the existence of the Canadian cancer registry, which has all primary cancer diagnoses since 1991 in Canada. We can link that with all living veterans in Canada. As of just the past few months, we have the data components that make that kind of analysis possible.