THE NEW England Journal of Medicine's recent retraction of a 2000 article that helped establish the popularity of Vioxx, the controversial painkiller, should give doctors, hospitals and journal authors reason to think harder about how drug safety information is processed. The retraction, published last month, stated that the article's authors -- including scientists in the employ of Merck & Co., which makes Vioxx -- knew in advance of publication that 20 out of more than 2,000 patients taking part in a study of Vioxx suffered from heart attacks after taking the drug, not 17 as the article reported. They left out the information, they say, on the grounds that the three heart attacks in question occurred after the study had technically ended. But as a result, the level of heart attack risk associated with Vioxx appeared in the article to be slightly lower than it should have been.
It is true that Merck disclosed the existence of the three heart attacks to physicians who were sent a copy of the original article, the main purpose of which was to establish that Vioxx caused fewer stomach problems than other leading painkillers. More important, Merck revealed the three additional heart attacks to the Food and Drug Administration, well before the journal article appeared. Had Merck not done so, the company would have been liable for criminal prosecution: The government does have the tools to prosecute companies that lie or withhold information from the FDA, and it does use them.
In other words, this revelation raises legitimate questions about the relationship between scientists, who are often paid by drug companies, and medical journals, which should do a better job of stating authors' conflicts of interest. It should also lead doctors who prescribe drugs and are ultimately responsible for understanding their side effects to read such journals with greater skepticism.
It does not, however, alter our view that the courts are the wrong place to resolve disputes about drug safety. The New England Journal of Medicine's retraction is expected to have an enormous impact on the thousands of lawsuits against Merck. But it is still not clear that doctors prescribing Vioxx several years ago would have behaved any differently had the journal article showed that the risk of heart problems was very slightly higher. Scientific judgments about the risks and advantages of drugs are not black and white -- which is why they are best made by scientists and by the regulatory agency that employs them, not by jurors through the lens of hindsight.