A CALL FOR A LARGE ELECTRONIC DATA BASE FOR DRUGS.

A CALL FOR A LARGE ELECTRONIC DATA BASE FOR DRUGS. Gina Kolata had an article last week in the New York Times news service, which described how biphosphonates (for example, Fosomax), which are used to treat osteoporosis, will now be required by the FDA to carry a warning label that “they can lead to rare fractures of the thigh bone.” This highlights the issue that short-term trials with limited number of people are not good at finding out rare side effects from long-term use. I note my usual comment that an estimate of the probability of the side effect (which is characterized as “very rare”) should be given because people understand numbers better than they do words like “rare” or “very rare.” What is important for me is the recommendation by one doctor for “the development of a national electronic drug database that would reveal drug use and complications.” I also note that the preference of “gold-standard” randomized drug tests is so strong in the medical community that two of the doctors “would like large clinical trials after a drug is approved and continuing for years” despite the practical difficulties—such as that such a trial would have difficulty in picking up very rare side effects.