Science Goes Tabloid: In scientific journals, if it bleeds, it leads
In the United Kingdom, most of the respected broadsheet newspapers have cut costs and increased circulation by adding a tabloid edition. Some argue that this downsizing has led to a dumbing down of the papers' content. But, in both Britain and America, it is not just the news industry that is shifting to a more sensationalistic attitude. Some scientific journals are abandoning scientific neutrality in favor of policy stances and headline-grabbing scare stories, favoring style over substance.
A prime example is the British Medical Journal (BMJ), which recently published a news story that suggested that Eli Lilly, the makers of Prozac, had failed to disclose links between the drug and violent behavior and suicide. The story alleges that certain documents detailing the alleged links—and provided to the BMJ by an anonymous source—had not been shared with the Food and Drug Administration (FDA) and had gone "missing" during the trial of man charged with murder in 1994. In a strongly worded response, Lilly pointed out that the documents had been in general circulation for years. In fact, the BMJ's one example of missing scientific data had been published by Lilly in the Journal of Clinical Psychopharmacology in 1992! The BMJ refused to allow Lilly to review the documents prior to publishing the story; Lilly was only able to do so after obtaining copies from Democratic congressman Maurice Hinchey of New York, who had received copies from the BMJ. Moreover, the BMJ told Lilly that it would have an opportunity to discuss the issue further after the Christmas holidays—but then went ahead and issued its story on January 1; it also sent the documents to the FDA.
This is "gotcha" journalism at its worst. The story would not have survived a cursory glance by a newsroom editor because the authors neither gave Lilly a chance to respond in an informed fashion nor checked the relevant background to see whether the documents were credible. Yet it was published by a scientific journal supposedly bound to uphold the highest standards of accuracy.
Why, then, would the BMJ publish the story? The journal is not commenting publicly except to say that it "takes the issue very seriously." The incident arouses the suspicion that the BMJ editors thought that a major whistle-blowing story would earn them headlines around the world, as it indeed did. But if this is the case, one is entitled to ask why a medical journal felt the need to take on this role.
Yet the problem goes wider than journals. Even some scientists are keen to trumpet their claims by means of sensational press releases, thus doing a disservice to sound science policy. For example, last November, a Johns Hopkins University research team found a correlation between consumption of high doses of Vitamin E and early death. They released the results to the press the same day they were published online by Annals of Internal Medicine. This resulted in a USA Today story headlined, "High Dose of Vitamin E May Increase Death Risk." The research concentrated on the elderly, especially those with heart disease, and could not be generalized to healthy people under 60—but this important fact was buried deep in the press release and was not given the emphasis it should have been by any news coverage. In addition to this significant problem, the actual difference in risk of early death—5 percent—was small enough to raise question marks among anyone familiar with epidemiological principles.
These significant limitations did not stop the researchers from recommending changes to the U.S. dietary guidelines. In doing this, the researchers jumped the gun. The purpose of scientific publication in journals is to allow other researchers to examine the results and either critique them or try to reproduce them. Treating one's own results as so final that they should lead to policy recommendations short circuits this process. Yet it is an increasingly common tactic.
Meanwhile, Nature, Britain's premier journal of natural science, has overtly abandoned neutrality in favor of specific policy stances in certain areas, most notably global warming. For instance, in the past year it has published an editorial favoring restrictions on the aviation industry; a news article that concludes that greenhouse-gas emissions trading shows that industries must reduce emissions; an essay on Napoleon's scientist Joseph Fourier that asks whether a future Nature essayist will look back and ask why we ignored evidence that the Earth's climate can change dramatically; and a speculative study on species extinction that endorses reducing greenhouse-gas emissions.
Perhaps the most blatant example of this "tabloidization" of science is the recent change in the cover design of the prestigious British medical journal The Lancet. For years, its cover featured a simple table of contents. Now, however, it is emblazoned by one extracted quote in a style worthy of Britain's Daily Mirror. For example, its October 1, 2004, issue's cover quotation read, "The prospect that vitamin pills may not only do no good but also kill their consumers is a scary speculation given the vast quantities that are used in certain communities." Yet the journal also ran an editorial comment on the study that pointed out that the elevated risk of death originated from "one trial in an anomalous population of smokers, ex-smokers and occupationally exposed asbestos workers. The other high-quality trials...do not suggest increased mortality." The news outlets that covered the study mostly picked up the "killing consumers" quote and ignored the qualifications. It is hard to see what medical purpose the pull-quote served. It certainly led to extensive media coverage, however.
If this trend continues, the scientific and medical communities are playing a very dangerous game. MIT scientist Richard Lindzen once commented, "Science is a tool of some value. It provides our only way of separating what is true from what is asserted. If we abuse that tool, it will not be available when it is needed." Yet a troubling number of journals and scientists are doing just that. If the institutions of science do not face up to this problem, we face the prospect of a "post-scientific," relativistic reality. The public that trusts scientists to benefit them deserves much better.