biojuice

Great editorial by Donald Kennedy in Science this week about improving US drug regulation:

“The pharmaceutical business is a multinational enterprise of great significance to human health. The U.S. Food and Drug Administration (FDA), once called the world standard in regulating new drug approvals, may still deserve that status. But, as was described earlier in this space, public confidence in the agency has been shaken. That’s why it’s important that we restore the FDA’s capacity to do the quality of work that all health care systems require.

…

With respect to the last of these, the problem is that the United States lacks a system that is adequately tuned to detect adverse reactions. That measure requires a numerator and a denominator: the number of reported adverse events divided by the number of prescriptions issued. The FDA knows neither. Event reporting is voluntary, yielding a record of dubious reliability, and there’s no national prescription record. That’s why the FDA had to use Kaiser, a large health maintenance organization, to find an adequate database for evaluating the safety of Vioxx. Information about all clinical trials reviewed by the FDA should be made publicly available at the FDA’s Web site and also be linked to the National Institutes of Health’s Web site, ClinicalTrials.gov. Why? Recently, a Cleveland Clinic cardiologist analyzed trials conducted on GlaxoSmithKline’s diabetic drug Avandia. His meta-analysis revealed cardiovascular risks that had been hinted at in earlier published trials but were not statistically significant in any one trial. The study, published in the New England Journal of Medicine, illustrates the difficulty of finding important trial data. The journal has since been attacked by the company, some physicians, and an unfortunate editorial in the Lancet, but it deserves praise instead.”

The Times had another good article last month about this issue with respect to Avandia:

For Drug-makers, A Downside to Full Disclosure