The only potential problem I see with Steve's analysis is inability to be sure certain batches weren't given to populations who were more/less likely to die anyway. As he mentions, by only looking at the under 60 age groups, you can be sure nursing home patients aren't included, but what about all the other classes of people who are more…
The only potential problem I see with Steve's analysis is inability to be sure certain batches weren't given to populations who were more/less likely to die anyway. As he mentions, by only looking at the under 60 age groups, you can be sure nursing home patients aren't included, but what about all the other classes of people who are more/less likely to die? I mean, if certain batches were given mostly to the poor, who have a higher mortality anyway, you'd expect that batch to look bad in this analysis. I'm not saying there isn't something here. I actually think the chances are good his conclusion is correct (i.e., there's no way the poor are 9x more likely to die in any given year). All I'm trying to do is be honest about the limitations. Maybe somebody has a better way to look at this.
The only potential problem I see with Steve's analysis is inability to be sure certain batches weren't given to populations who were more/less likely to die anyway. As he mentions, by only looking at the under 60 age groups, you can be sure nursing home patients aren't included, but what about all the other classes of people who are more/less likely to die? I mean, if certain batches were given mostly to the poor, who have a higher mortality anyway, you'd expect that batch to look bad in this analysis. I'm not saying there isn't something here. I actually think the chances are good his conclusion is correct (i.e., there's no way the poor are 9x more likely to die in any given year). All I'm trying to do is be honest about the limitations. Maybe somebody has a better way to look at this.