You may be right. I'm looking at an MAIG report, and while I trust their analysis even less, it clearly points to newer incidents that I was previously unaware of. This may be an overlap on the data, as I'm quoting a fairly old study, or it may be that the study was incomplete.

I might be looking at a full on amendment here, or there may have been more statistical jiggery pokery that somehow excludes findings incongruent with the expected results. Either way, I'll do more research and get back. Thanks for the correction -- while the link you posted isn't necessarily conclusive, it did spur me to freshen my data sets at the very least.