Why I blame the FDA for their decision to approve Old Monty
The decision by the FDA to approve Old Monty was bad, in my opinion. Let me explain.
The issue has to do with what is called 'Sub-Group Analysis'. Here's an extreme example to illustrate a point: suppose a drug application was sent to the FDA which consisted of two clinical trials. One of the clinical trials was conducted on people West of the Mississippi, and the other clinical trial had patients who lived East of the Mississippi. Suppose there were a very high number of deaths in one of the groups and not the other group, such that there was no chance that the drug would ever be approved in that one region. Would it be a good decision to completely ignore the results in the one trial and then go ahead and approve the drug for people living in the other region? Of course not. A company may go ahead and attempt to get approval for one group - but the FDA would be expected to reject based on this 'after the fact' subgroup analysis.
Sometimes the sub-groups are different and so it does make sense to consider the differrences. For example, people aged over 65 are obviously different in health than people under 65, so there may be some real reason that the results from their experiences may be viewed differently.
In the AFFY situation, the two sub-groups were a) people with kidney problems who were getting Dialysis treatment and b) people with kidney problems who were not getting dialysis treatment. Logically, we know that there is no difference in these groups, except that group a) is sicker and therefore required dialysis treatment.
The O deaths occured in the b) group (the healthier patients not yet on dialysis) and so it was determined that drug O should not be allowed for people not on dialysis. Strangely, the FDA considered the trials for the two groups independently and concluded that once a patient is sick enough to go on dialysis, then drug O is no longer considered a killer and instead is considered safe enough for approval. Very strange.