The trouble of course with this kind of stat is that it is too general: you can always say 100% of cancer victims drank tap water in large quantities or whatever.