That is precisely the problem. With clinical trial data, we are talking about very limited patient groups. In the actual practice setting, we are dealing with subgroups of patients who will not all react the same way. Yet we are unable to get at the subtleties and small details on the sole basis of the clinical research. That is precisely why it is very important to have the same volume of information from actual practice settings or based on observation. At the same time, the data you are referring to should be interpreted cautiously, in my opinion, because we have to be sure to compare apples with apples. The fact is I was not referring to exactly the same concepts.
On March 11th, 2008. See this statement in context.