[Updated on 9/12/2012 at 10:35 a.m. with a response from Isabelle Boutron.]
In recent years, newspapers have been full of articles touting the health benefits of coffee: It cuts the risk of heart attack, stroke, and various kinds of cancers. Yet some studies have also raised warnings, saying coffee can encourage overeating and, yes, even increase heart-attack risks.
Similar uncertainties—at least as reflected in newspaper articles and TV news reports—surround red wine, aspirin, estrogen supplements, prostate screening, and many other foods, pharmaceuticals, and medical procedures.
What’s going on? Are science reporters unable to make sense out of medical research? Are they overblowing minor fluctuations in study findings just to attract readers?
The answer, according to an analysis published on Tuesday in the journal PLoS Medicine, is neither. Instead, according to one of its authors, Isabelle Boutron, an assistant professor of epidemiology at Paris Descartes University, the tendency to emphasize medical benefits over risks is most often attributable to the presentations by scientists in their own journal articles.
The study was based on an examination of 70 articles that appeared in scientific journals from December 2009 to March 2010, reporting the results of randomized, controlled trials. The study also included a review of the press releases that accompanied those articles, and the subsequent news coverage.
Of the 41 news stories that could be associated with those studies, a majority of them—21—contained “spin,” which Dr. Boutron and her team defined as a specific reporting strategy, intentional or unintentional, that emphasized the beneficial effect of the treatment being tested.
In most cases, though, that spin could be traced back to the journal article itself, Dr. Boutron concluded. Her analysis found spin in 40 percent of the study abstracts published in the scientific journals, and in 47 percent of the accompanying news releases.
“These findings show that ‘spin’ in press releases and news reports is related to the presence of ‘spin’ in the abstract of peer-reviewed reports,” Dr. Boutron wrote.
The 70 studies covered by the analysis included evaluations of treatments for a variety of diseases, including cancers, heart conditions, skin lesions, herpes, pneumonia, diarrhea, diabetes, anemia, fibromyalgia, and postnatal depression. The studies also explored behaviors such as exercise, television viewing, and smoking, and assessed several types of vaccines and remedies that included fish oil, pine bark, hypnosis, acupuncture, psychotherapy, and “flower power.” The research was conducted at a range of universities in the United States and abroad.
In their article, Dr. Boutron and her six co-authors offered one example of spin, describing a December 2009 study by Eleanor M. Walker, director of breast services in the department of radiation oncology at Henry Ford Hospital, in Detroit.
Dr. Walker compared acupuncture with a standard drug therapy for reducing hot flashes in breast-cancer patients, and said she found both roughly comparable in effectiveness. As an unexpected side benefit, the study also found that acupuncture may increase a woman’s sex drive and improve her sense of well-being. The report garnered extensive media coverage, by major newspapers and television networks.
In the PLoS article, Dr. Boutron said Dr. Walker’s comparison of acupuncture and drug therapy “was not statistically significant,” and faulted her for highlighting the improvement in sex drive when that was not “a pre-specified outcome” of the research.
But Dr. Walker, asked about the matter, rejected Dr. Boutron’s conclusions as mistaken. The study, involving 25 women receiving acupuncture and 25 receiving drug therapy, did show both groups’ experiencing “significant” declines in hot flashes, Dr. Walker said in an interview. And, Dr. Walker said, the finding about sex drive was not suggested in advance of the study because it had not been anticipated.
“It wasn’t a spin—everything that was mentioned is in fact in the paper and supported with data,” Dr. Walker said. She said she had not been contacted by Dr. Boutron and was not aware of the PLoS article before being contacted by The Chronicle for comment.
“Why my study should be picked out for this—maybe because it got 1.7 million media hits,” she said.
Dr. Boutron stood by her criticism on Wednesday, saying that Dr. Walker had claimed equivalence between acupuncture and drug therapy despite her small sample size, with 34 percent of her study patients excluded from the analysis. “We presented this example in our article as it is important for readers to understand what spin means and its potential implications,” Dr. Boutron said in a written response.
Dr. Boutron did not describe any of the other examples of spin she found, beyond that one case involving Dr. Walker.
But Dr. Walker offered one: She said she believes Dr. Boutron’s PLoS article is a case of an article overstating the researchers’ actual findings. “To me,” Dr. Walker said, “this sounds like a spin—they’re trying to get a paper out of this.”