Michael Joyce is a writer-producer with HealthNewsReview.org and tweets as @mlmjoyce
We often write about the problems arising from hyped news releases coming out of medical conferences.
But I don’t think we’ve ever written about an over-the-top news release getting toned down — and re-released — with more cautious language.
Kudos to the media team at the annual meeting of the American Society for Clinical Oncology (ASCO) for doing so with a news release featuring a clinical trial looking at whether pairing a commonly used stomach acid-reducing medicine (esomeprazole; Nexium®), with aspirin, might decrease esophageal cancer in a subset of high risk patients.
But was it too little too late? Had enterprising journalists already framed their reporting based on the first release (distributed to journalists last week) before the toned down version was released over the weekend? Would journalists miss the corrections in the flurry of conference-related communications? We may never know for sure, but some of the coverage suggests so.
If that’s the case, it serves as a compelling example of how more cautious framing of news releases from a meeting can lead to more balanced and informative news coverage.
Some highlights from the first release (that was subsequently updated and corrected):
How these highlights were toned down (in bold) in the second release:
The changes above acknowledge three shortcomings of the first release that had the potential to mislead readers, and even cause harm.
First, is one of scale. If a finding is minor or moderate, why not just say so? Isn’t that scale as relevant and important as ‘major or severe’? By simply adding the qualifier ‘moderate’ to the second news release, readers are now appropriately prompted to ask themselves — or their doctor — “How much is moderate?” They’re also less likely to simply accept the result as a foregone conclusion. Even better would be to document, in absolute numbers, how many fewer people developed these outcomes when they were treated or not treated with Nexium and aspirin.
Second, it’s not infrequent for us to see lead authors — not just news release writers — using unjustified or hyped language. Having the lead author tone down his language in this case was quite justified. Both Nexium and aspirin have important side effects and drug interactions to consider. Given how common heartburn is, the lead author’s comment could have easily prompted a run on both drugs.
Finally, the change in the third bullet point acknowledges an important limitation of the study results; that is, the primary outcome of this study is what’s called a “composite endpoint.” (Read more about composite endpoints HERE.)
In this case, researchers measured three outcomes of the Nexium/aspirin combo: reducing the risk of developing high-grade dysplasia (so-called ‘precancerous’ changes), reducing the risk of developing cancer of the esophagus, or delay death from any cause in high-risk patients with Barrett’s esophagus.
Lumping all 3 outcomes together can make it easier to claim a statistically significant result — because your treatment is more likely to affect at least one of the outcomes. But how clinically relevant is that when your’e left wondering: “Which outcome was affected? One, two, or three?”
Remember, the first news release boldly claimed “25% of esophageal cancers prevented” with the therapy. But they backed off in the revised release to conclude a “moderate” reduction in the aggregate sum of three outcomes. This suggests they felt the claim of a 25% reduction in esophageal cancers was misleading. But that should have been clarified with the revision.
“It’s a very unusual composite endpoint,” said Dr. Emil Lou, a gastrointestinal oncologist at the University of Minnesota who’s attending the ASCO meeting in Chicago.
“You’d think if you were studying a drug combination to prevent cancer, you’d really want to measure whether the treated group had any sort of delay in the time it took to develop cancer, when compared to a control group. With the composite outcome we can’t really know if 25% of cancers were prevented.”
This sensationalistic story from The Sun — Aspirin & Indigestion Drugs ‘can prevent cancer’ — clearly relied upon the first release. The evidence for that comes from the fact that it uses the lead author’s more reckless quote recommending that “any person with heartburn” consider the Nexium/aspirin combo.
The story is even more imbalanced than the release. It takes on a promotional and cheerleading tack in following this wildly erroneous logic: that since heartburn is common, and it can damage the esophagus, and damaged cells in the esophagus are associated with cancer … and because Nexium and aspirin are cheap … then people should “slash their risk of gullet cancer” by taking the combination.
In contrast, this much more cautious story by itv.com — Aspirin Could Prevent Esophageal Cancer; New Research Suggests — is clearly based on the toned down release because it features the more judicious quote by the lead author (ie. ask your doctor; don’t self-medicate).
It also makes it quite clear that these preliminary, unpublished results only apply to people with Barrett’s esophagus, and that aspirin does have serious side effects like internal bleeding.
I wish it had added that Nexium and other proton pump inhibitors have side effects (metabolic deficiencies, intestinal bacterial overgrowth), and also made it clear that only about 1 in 10 people with chronic acid reflux (“GERD”) develop Barrett’s esophagus. Kudos for making it clear that only about 1-2% of those with Barrett’s esophagus ever develop cancer of the esophagus.
What I find intriguing about these two news releases is this: adding a single qualifier (“moderate”) + replacing a single hyped quote + including a single, important limitation can all add up to a potentially big difference in coverage. Although I can’t be sure the two news releases dictated two completely different types of coverage, it certainly seems possible.
And if that’s the case, it shows just how great a responsibility news release writers have to get it right the first time — before their words and framing get picked up by news outlets around the globe and shape what readers think about research that has the potential to help, or to do harm.
See more of our ASCO 2018 meeting coverage:
Editor’s note: We remind readers that HealthNewsReview.org is committed to doing more than criticize imbalanced media messages. We also offer to review any health care-related news release prior to publication and before it reaches consumers. We encourage public relations teams to take us up on this free, no-obligation service.