The primary study was done by Lazarou et al., three professors from the University of Toronto, in 1998. It was a macro study of hospitals. I don't remember the number of hospitals, but I have a reference for you in the printed material. They determined that 104,000 people die from adverse drug reactions in the United States, in hospitals alone, from prescription drugs taken as prescribed—not in error, but as they were prescribed. Others have extrapolated that figure and asked about the deaths that occur outside hospitals. They think that may be as high as another 100,000, although it's very difficult to prove.
So using the population ratio, that would be 10,000 in Canadian hospitals and then potentially another 10,000 outside Canadian hospitals.