Spoonful of Medicine, and Risk?

May 30, 2007 — -- It's been a week since the release of a controversial study suggesting the popular diabetes drug Avandia increases the risk of heart attacks and heart deaths.

The Food and Drug Administration is convening a panel of experts to investigate, but there is at least one point on which most doctors appear to agree: The whole process of ensuring drug safety in this country is seriously flawed.

In an exclusive interview with ABC News, Dr. Janet Woodcock, deputy commissioner of the FDA, conceded: "We are finding more problems with marketed medications, but not just with recently marketed medications, but drugs approved long ago."

A report from the Government Accounting Office found that, after approval, 51 percent of drugs were later discovered to have major safety problems requiring label changes. And in 3 percent to 4 percent of drugs, the safety issues were so serious the drugs were pulled off the market.

Dr. Jerry Avorn, author of "Powerful Medicines: The Benefits, Risks and Costs of Prescription Drugs," called that "unacceptable."

"We don't have a good system for finding out if a drug is dangerous until it's been on the market for many years," he said.

There are many reasons for this, one being that drugs are usually tested on younger and healthier patients than those who will ultimately be prescribed the medications — this often hides side effects.

"A drug that works well in handpicked patients," Avorn said, "who are guaranteed not to be terribly sick, is not likely to have the same safety or efficacy as it does in typical patients."

A second reason is that clinical trials are too small. Many drugs are tested on only 2,000 to 3,000 patients. This means serious, rare complications, such as a heart attack — which may be seen in one out of every 5,000 patients — may be completely missed in these studies.

A third is that clinical trials are too short. Most of the Avandia studies lasted only six months to a year. But many of these drugs are meant to be taken for a lifetime, said Dr. Brian Strom, chairman of biostatistics and epidemiology at the University of Pennsylvania. "So we don't know what the long-term effects are at the time they're marketed," he said.