91É«ÇéƬ

Doctors and patients are being misled by the results of drug trials published in major international medical journals. According to a new analysis, reports on trials designed to find out how well a drug prevents strokes or heart attack, for example, include only the most flattering statistics, which makes a drug or procedure seem more effective than it actually is. Of 359 articles published between 1989 and 1998 in five major journals, all included a relative measure of how the drug reduces health risks, say Jim Nuovo and Joy Melnikow at the University of California, Davis. But only eight reported the more significant figure of how many people a doctor would need to treat before one of those patients sees the benefit.