I'm not necessarily sure what the dangers would be. Assuming that privacy issues aren't at play and nothing is personally identifiable, then no. We're talking in slightly broad terms here, but I believe it absolutely should be the case that researchers—even someone who has training but not necessarily institutional affiliation, and who is interested in looking at rates of illness in Canada—are able to look into that.
On December 7th, 2010. See this statement in context.