How Long is Long Enough? The FDA is coming under criticism in the wake of the Vioxx withdrawl. Some critics expect drugs with no side effects, even attacking the very old and extremely important antibiotic erythromycin because it has the potential for drug interactions (something that doctors know, by the way, and are usually careful about before prescribing it). Others say drug trials looking for side effects should be longer:

The FDA could take a simple step that would improve clinical trial quality before it approved drugs, observers say. Drug companies with products comparable to Vioxx could be required to conduct longer clinical trials.

"If you're the FDA, you'll say `OK, all bets are off. We're going to make you do studies lasting 18 months,'" said Dr. Sidney Wolfe, director of Public Citizen.

At the time of approval, the FDA had data from clinical trials on Vioxx that lasted for 12 months. The increased risk of heart attack and stroke that prompted Merck to pull Vioxx did not begin to appear until older patients had taken took the drug for 18 months.

It's a valid criticism that drugs that are used long-term (like birth control or arthritis medication) should be tested longer than six or twelve months. But then, what about drugs whose long-term side effects only manifest themselves after two or three or five years, or even longer? No time frame is going to be perfect. And no drug is going to have zero side effects. Every doctor and every patient has to ask themselves if the potential risk is really worth the benefit.