Within CDER itself, we're driving much greater collaboration, cooperation, integration, and communication between the parts and pieces. There's much greater interaction between the Center for New Drugs and the Center for Surveillance and Epidemiology so that those who want to approve drugs and those who are responsible for the safety of the drugs are integrated and working effectively. By effectively, I mean by bringing the various perspectives and points of view together to debate, discuss, and analyze vigorously, and then to make an informed decision of what's best. So it's integration within CDER, it's integration of CDER within the larger FDA portfolio, and then it's the integration of FDA with the other parts and pieces.

When you talk about integration, it sounds as if the safety part and the efficacy part of drug oversight are going to continue to be engaged together. This is one of the ideas that some of your critics in Congress don't like.

In the old model, we are dealing with a pendulum that swings from rapid approval of effective drugs to slowing down approval to be even more certain of safety. In that model, you see things as separate. Efficacy's here, and safety is there.

I came from an environment where if you looked at cancer care and what was best for the patient, it wasn't a matter of saying it's surgery or radiation or chemotherapy, but to realize that a woman with breast cancer needed a solution that required the integration of these pieces because there was a dynamic that had to occur. And yet, as chief academic officer and someone responsible for program development, the last thing in the world you wanted to do was to homogenize that into one thing, because you always wanted the surgeons to be the surgeons, the medical oncologists to be the medical oncologists, and the radiotherapists to be the radiotherapists. You wanted each group to bring to the table its own perspective on what it believed would be the impact of its intervention—even if they argued about what could serve that patient best. But what you come up with out of that rich discussion is a decision as to what's best for that patient.

So you don't want to separate them—but you don't want to homogenize them either into a place where they all think and act the same, because that doesn't benefit anything. You want to create this environment that I call, not the melting pot, but the stew pot. And you need to have a place where that kind of discussion and difference of opinion could go on in an appropriate way.

If you look at the figures on patients with severe arthritis who'd been on Celebrex and Vioxx, something like a third of them now take no medication at all. That probably doesn't mean that they got better, but more likely that a large number of patients have basically been deprived of a viable drug. Now Vioxx is off the market, and Arcoxia doesn't look like it's got a fantastic chance. What about those patients and their right to make a decision for themselves?

First of all, there's the important perspective of public health as well as personal health—and all the decisions have to be made in the context of what is best for the public at large.