Fixing a broken FDA

Dangerous drugs that got approved and are widely and intensively marketed; dangerous imports from China and elsewhere that are never examined; conflict of interest allegations; contaminated consume products from toothpaste to lens solutions; vaccine supplies that go from shortage to glut. Just business as usual at the broken US Food and Drug Administration (FDA). So it's an important problem and fixing it is a priority. It's always a dilemma when a really excellent piece appears and you just want your readers to go read it instead of what we are writing about it. That's the case with an Op Ed in yesterday's Boston Globe by Susan Wood and David Michaels about upcoming legislation to fix some of the problems at the FDA. So you should read it, but I am also going to pull a quote anyway:

As Congress finalizes the first major FDA legislation in years, there is momentum for real change, and resistance from powerful forces that like the system the way it is.

Vioxx, drug-eluting stents, and other dangerous medical products were approved, in part, because the current system has worked exactly as designed: to get the products approved quickly and, only later and half-heartedly, to attempt to collect data on side effects.

New drugs and medical products are reviewed by the FDA based on relatively small, short-term clinical trials conducted by the firms that make the products. These studies are not designed to identify rare, adverse effects or even common long-term risks that occur after months of use. Once their products are on the market, manufacturers have little incentive to discover adverse effects, and the FDA has limited authority and even more limited resources to require manufacturers to conduct the long-term, postapproval studies that would fill in the blanks. (Boston Globe)

Wood and Michaels show how the 1992 Prescription Drug User-Fee Act (PDUFA, pronounced Pah-Doofah) created a serious problem with what looked like a reasonable proposition: that the drug companies should pay for the safety evaluation of their products through user fees. Sadly Congress will continue PDUFA. But there is still much that can be done to improve matters. You should read their piece. They make important points about conflict of interest and whose side FDA sees themselves on. Ideally the FDA is on side of "the public." But there are many ways to interpret this. We need only think of "what's good for General Motors is good for the USA" to remember the public good can be construed in many ways.

There is another fundamental problem that needs urgent attention. The relatively short-term relatively small scale pre-market testing will never be sufficient to detect the longer term rare but serious side effects that can only be seen when tens of millions of people are taking something, which for blockbuster drug is not unusual. One fatality in a hundred thousand is hardly detectable in pre-market testing but would result in 1000 deaths per 10 million users after the drug goes on the market. The only solution to that problem is effective and timely post-market surveillance. Wood and Michaels suggest making the data publicly available would be one way to improve post-market surveillance, but a workable system of surveillance with mandatory reporting from doctors and hospitals is another. Right now reporting of adverse events is voluntary.

Of course if the US had universal health care with an integrated medical records system this problem could be solved much more easily. Just another reason for universal health care -- as if there weren't enough already.