WASHINGTON — During the Bush administration, the Food and Drug Administration was mostly a place of black-and-white decisions. Drugs were approved for sale or they were not, and the agency’s staff was expected to publicly support those decisions.
But as Thursday’s landmark decision on the controversial diabetes medicineAvandia makes clear, things have changed under the Obama administration. Certainty, staff unanimity and even the approval status of big-selling medicines are no longer so black and white.
Presented with what seemed to be a choice between keeping Avandia on the market or withdrawing it, the Obama administration decided on an unusual middle path — allowing sales, but with tight restrictions. Even more unusually, the agency admitted that many of its top scientists disagreed, some passionately. Competing memorandums were posted immediately on the agency’s Web site.
And the agency’s three top officials co-wrote a highly unusual explanation of their action in The New England Journal of Medicine.
Some of these changes have been in the works for years, but they have accelerated under the Obama administration, driven by increasingly sophisticated measures of drug safety and growing skepticism about whether the F.D.A. is making the right decisions and making them appropriately.
“I think that F.D.A.’s credibility really depends on being able to explain its decisions well,” said Dr. Joshua Sharfstein, F.D.A.’s principal deputy commissioner. “We can’t expect people to think that F.D.A. has decided, therefore it’s the right answer.”
Some of the changes have been driven by people like Dr. Steven Nissen, a cardiologist at the Cleveland Clinic whose 2007 analysis of Avandia’s heart risks stunned doctors, patients and legislators, who asked why the F.D.A. had not done anything similar. When the agency revealed it had done an almost identical analysis a year earlier and found the same result, the controversy intensified.
“You have these third-party analysts setting the agenda for the agency in ways that never happened before,” said Daniel Carpenter, an F.D.A. historian at Harvard.
For the F.D.A., the Nissen analysis presented major challenges. It demonstrated that the agency no longer had a monopoly on the information needed to make drug and device safety decisions. Data from crucial clinical trials are increasingly being posted on public Web sites. And academics are using sophisticated techniques to test whether popular medicines are safe.
In March, for instance, a team of academics found that a children’s diarrhea vaccine contained harmless but apparently extraneous pieces of pig virus. Blindsided, the F.D.A. had no idea what effect the particles would have. While the agency studied the problem, the commissioner, Dr. Margaret Hamburg, asked its maker to stop selling — a request she had little power to enforce.
Two months later, the agency allowed sales to continue.
Like the vaccine finding, the Nissen analysis flummoxed the agency because the science behind it was controversial. Dr. Nissen combined the results of many clinical trials to suggest that Avandia substantially increased heart risks. Other studies suggested that there were higher risks.
None of these studies met the rigorous standards that the F.D.A. demands when approving new medicines, but they were among the only information available to explore whether popular medicines contribute to common problems like heart attacks.
The agency was torn about how to interpret the studies, a problem it rarely faced until recently. “In the past, we would approve the drug after a couple of efficacy trials and that was it,” Dr. Janet Woodcock, chief of the F.D.A.’s drug center, said in an interview. “We didn’t know too much more about the drug. It was simpler.”
Now, sophisticated analyses present the F.D.A. with a complex picture. “It’s good for public health that we’re learning more, but it creates a more complex environment in which to regulate,” Dr. Woodcock said.
It is an environment in which top agency officials are in some ways at sea. The agency has no systems or standards to follow in deciding which studies deserve their attention or should lead to changes in a drug’s status. And since new tests are being created constantly, creating such a standard would be an ever-evolving process.
Dr. Lynn Goldman, dean of the School of Public Health and Health Services at George Washington University, said the F.D.A. was being forced to become more comfortable with studies done in academic rather than regulatory settings. “They have to get used to a less controlled environment,” Dr. Goldman said.
And the agency’s decision to create a unique distribution program for Avandia is not one it can repeat often or doctors and pharmacists — who must learn a new system for each program — will give up.
“We have to get some standardization,” Dr. Woodcock said, “or we’ll burn out the system.”