Opinion, Berkeley Blogs

Can you trust health news?

By John Swartzberg

Reporting health news isn't easy, especially when journalists have short deadlines and limited space to parse research that's frequently complex, nuanced, and laced with caveats. On top of that, theres often the temptation for scientists, press offices, and reporters to oversimplify and oversell research findings to get more attention. I notice this more and more in health news these days.

Recently two analyses of news stories about health and medicine caught my eye. The researchers gave many, if not most, of the stories failing grades and identified some all-too-common pitfalls.

wellness180c The first analysis, published in JAMA Internal Medicine in summer 2014, judged nearly 1,900 media reports published in the U.S. since 2006 about new drugs, medical devices, screening tests, and procedures. It graded most reports as unsatisfactory in how they discussed the benefits and harms of the interventions, their costs, and the quality of the evidence. Among the faults it found in most news reports:

  • Potential benefits were exaggerated; harms were minimized or ignored.

  • They relied on a single source, often a press release.

  • Benefits were expressed only in terms of relative risk and thus sounded more dramatic than they would have had absolute risk been reported. The researchers gave as an example a news report that stated that aspirin reduced the risk of stroke in women by 24 percentbut that left out the less-impressive fact that this worked out to two fewer strokes per 1,000 women over six years. (We report relative risk on Berkeley Wellness only when we judge the absolute risk to be substantial enough to justify doing so.)

  • The reports focused on observational studies, but didn't attempt to explain the limitations of such research. As we often point out, such studies find only associations and cannot prove cause and effect.

  • They relied on positive anecdotes from patients and doctors.

  • Coverage of expensive new technologies was fawning, with little or no discussion of their cost or their effectiveness compared with existing alternatives.

    • Not just the reporters fault

    The second analysis, published in the journal BMJ in December 2014, concluded that much of the misinformation and exaggeration found in health-related news stories can be traced back to press releases, which are the dominant link between academia and the media. Of course, thats not a good excuse for reporters, since they shouldn't rely on press releases, at least not without making sure they match up with what the studies actually found.

    The researchers looked at 462 press releases from 20 leading British universities in 2011. They found that more than one-third contained exaggerated statements not found in the published studies, and that these were likely to end up in news stories. There were three main types of distortions: extrapolating the results of animal research to humans, making causal statements based on observational results, and giving advice not found in the studies.

    So its the fault of academic press officers? Not quite, since researchers usually get to review the releases about their studies. The blame if it can be meaningfully apportioned lies mainly with the increasing culture of university competition and self-promotion, interacting with the increasing pressures on journalists to do more with less time, the authors concluded.

    Crossposted from Berkeley Wellness.