IN THE PAST, the drug safety function of the Food and Drug Administration was straightforward: The FDA's scientists were simply meant to determine whether drugs are safe and effective, and to warn the public of side effects. But in recent years, the public's propensity to be easily scared has made the agency fear that "overwarning" of the downside of certain drugs might be as pernicious as "underwarning". In other words, too much emphasis on suspected but unproven side effects could scare people away from a drug that might help them.

That's the dynamic that appeared to be in play a few months ago, when one of the FDA's drug safety experts wrote an analysis that considered the impact of one class of anti-depressant drugs on children and teenagers. The analysis, which seemed to show that the drugs could make younger patients suicidal, was strong enough to persuade British regulators to ban use of the drugs, with the exception of Prozac, for patients younger than 18. Nevertheless, some FDA scientists questioned the validity of part of the data, particularly the unclear definition of "suicidal behavior." A decision was made to refer all of the raw data to an independent research group -- and to prevent the expert from presenting his views at a public FDA meeting. FDA directors feared that too much attention to the unproven studies could persuade doctors not to prescribe helpful drugs.

That decision is now the focus of congressional investigations. Sen. Charles E. Grassley (R-Iowa) and House Energy and Commerce Committee experts are looking into that incident as well as other cases in which, as Mr. Grassley put it in the Wall Street Journal, the FDA told experts to "clam up." But instead of scouring FDA e-mails for evidence of wrongdoing, it would be more useful if Congress reexamined why more raw data from clinical trials aren't made publicly available, enabling doctors to make judgments for themselves in other controversial and borderline cases. At the moment, laws rightly prevent the FDA from divulging proprietary information, such as chemical formulas and results of tests of drugs that have not been approved. But drug companies are also allowed to keep secret the results of trials conducted after drugs have been approved, including evidence that might help doctors choose one form of therapy over another.

Indeed, one overlooked conclusion that might also be drawn from the controversial FDA analysis is that -- leaving aside evidence about suicide -- the clinical trials of anti-depressants also cast doubt on whether these drugs are more effective than other kinds of treatments for adolescents: Patients in control groups who received extra attention and family therapy often showed as much or more improvement than those taking the drugs. Traditionally, the FDA has not been held responsible for comparing the effectiveness of different kinds of drugs or treatments. But then, traditionally, the FDA did not worry about scaring the public. Times change; so could the FDA, if Congress allows it.