Leah Rosenbaum is currently a graduate student at the University of California, Berkeley, studying journalism and public health. Her work has been featured in STAT News, Vice (Tonic), and other outlets. Follow her on twitter: @leah_rosenbaum.
A study about alcohol use during pregnancy kicked up major media waves last weekend—but was all that frothy coverage warranted? It was another lesson in the pitfalls of reporting on scientific research, as headlines around the world issued some potentially dangerous misinformation.
The Times of London, one of the largest daily newspapers in the UK, ran a front-page story about the study with the headline “Light drinking does no harm in pregnancy.” This claim, so far out of sync with the conclusions of the study, drew ire from the researchers themselves. One of the study authors, Luisa Zuccolo, PhD, took to Twitter after the Times refused to publish a rebuttal from the researchers.
“To say ‘light drinking does no harm in pregnancy’ is a gross misinterpretation of our findings,” Zuccolo wrote. “Such misreporting may boost newspaper sales,” she continued, “but does not benefit public health.”
The Times ultimately issued a correction and re-titled the article online. Now the headline is, “No need for guilt over light drinking in pregnancy,” which is, at best, a marginal improvement as I’ll explain.
Alcohol consumption is an important topic for women everywhere who are trying to conceive or are currently pregnant. While many women avoid alcohol completely once they know they are pregnant, worries can linger about alcohol that was consumed before conception. Fetal alcohol syndrome has been linked to heavy drinking for decades, but little research has been done on the threshold where alcohol becomes dangerous to the fetus. Is one drink at a special event okay? The lack of evidence currently leads many doctors to recommend that women abstain from drinking entirely out of an abundance of caution. Some women wonder whether total abstinence is really necessary.
The study, published in BMJ Open, was a meta-analysis of observational studies looking at harms of “light drinking” during pregnancy. The study was notable in what it did not find: the authors’ main conclusion was that there is little high-quality evidence about whether or not light drinking can harm pregnant women and their children—but that still doesn’t mean it’s safe. The authors concluded simply that this is an important area for future study. In the meantime, the authors emphasized, pregnant women should strongly consider avoiding all alcohol consumption. “Guidance could advise abstention as a precautionary principle,” the authors wrote in their conclusion.
The Times wasn’t the only media outlet with a misleading headline in the days after the study was published. Coverage of the study can be grouped into three tiers: grossly misleading, wishy-washy, and solidly science based.
In the grossly misleading category is the Times, along with Refinery29 (Headline: “Reminder: It’s okay to drink (a little) during pregnancy”), the New York Post (“Light drinking during pregnancy does not harm babies: study”), and Business Insider (“Pregnant mothers can drink 4 units of alcohol a week without putting their unborn baby at considerable risk, according to a new study”).
In the category of wishy-washy coverage, outlets including Reader’s Digest and the International Business Times hedged by saying that drinking during pregnancy “might” not harm babies. Adding meek qualifiers to headlines that overstate study evidence is one of the most common pitfalls in health care reporting. We offer tips on how to headline health stories in ways that are both accurate and compelling.
In the science-based reporting camp sit NPR and Newsweek, which both took a firm stance against drinking at all during pregnancy, citing multiple experts who advised against it out of caution. This is also in line with the recommendations of the study authors.
The best headline however, was from Popular Science, which ran: “We know shockingly little about the effects of drinking during pregnancy.” That, after all, was the study’s only real conclusion.
During the headline debacle, a more nuanced debate came up on Twitter between the study author and several followers. With findings that were so prone to be misinterpreted, some people wondered, should the study have been reported on at all?
The issue was raised by James Pickett, a research manager in the UK, who wrote in a reply to Zuccolo’s tweet, “It’s a shame the times won’t publish [the rebuttal] but, to stir debate, did this finding need a press release to start?”
Its a shame the times wont publish it but, to stir debate, did this finding need a press release to start? https://t.co/8XNp4arNlQ
— JamesPickett (@JamesPickett12) September 14, 2017
Pickett brings up an interesting point. Not all studies are given news releases — we’ve written about BMJ’s philosophy on issuing news releases before. The studies that are promoted with news releases, said a BMJ spokesperson, are the ones that BMJ thinks journalists — and the public — will find most interesting. The whole point of news releases, after all, is publicity.
And although randomized controlled studies (RCTs) are the gold standard of scientific research, a study last year published in PLOS ONE found that journal news releases were more commonly issued for observational studies — a weaker level of evidence — than for RCTs. (Yes, a study about studies, very meta.) It’s worth noting that many lifestyle topics holding great news appeal for consumers — such as the health effects of nutrition and exercise — are frequently studied with weaker methods and rarely subjected to “gold standard” scrutiny via RCTs.
Journal news releases drive science news coverage. In this case, the study was published online on August 3, but stories about the research didn’t start to pop up until after the BMJ news release was issued on September 12. Was it necessary for BMJ to seek out news coverage for a study that offered no actionable information for the general public and which was likely to be misinterpreted? To be clear, the news release was accurate and cautious in its depiction of the findings. Blame for the misleading headlines sits squarely with the editors who came up with them. But does BMJ bear some responsibility for creating the conditions that would likely lead to confusion?
The question of whether or not a news release or news coverage is warranted is difficult, said Liz Fraser, registered midwife and participant in the Twitter debate.
“Science reporting is often inappropriate and often fails to contextualize research in a way that allows the real significance of the research to be understood by people without a science background, and it’s not clear to me that there’s an obvious solution to that problem,” she wrote.
For the authors of the paper, it was not the reporting itself but the headlines that raised real concerns. News outlets issued medical advice that directly contradicted consensus recommendations from health professionals, said study lead Zuccolo. And such misguided advice could potentially have real harmful consequences for unborn children.
“This misunderstanding or intentional misreporting which conflated no evidence with no harm, could potentially damage some children if their mums were to act on the headline and disregard the official public health advice,” Zuccolo wrote to me.
This case highlights the need to pay scrupulous attention to the actual findings of the study—not just what editors think is true or what the public might want to hear. The effects of misreporting on this issue could impact generations to come.