For communication professionals, the pace of COVID-19 news and developments has gone beyond difficult to handle. It is unmanageable on many days. This is why communicators need to know when they are in over their heads, says Gary Schwitzer, veteran health care journalist and publisher of HealthNewsReview.org, a longstanding health media watchdog project.
Whether you work for a news outlet, a nonprofit, or a corporation and you’re covering COVID-19, you need to have a robust understanding of how to write about and report on health and science news. And, most importantly, you need to know when to ask an expert for help, says Schwitzer.
In this in-depth interview, Schwitzer spoke to Content Science about the most significant challenges he is seeing in health and science communication today and how communication professionals can avoid these problems. [Stay tuned for part two from our interview with Schwitzer, coming next week, for his take on the state of COVID-19 vaccine science so far and how communication professionals can cover the news responsibly.]
SCHWITZER: I think that we are, at times, seeing some of the best health and medical science journalism that we have ever seen. Now, I am always quick to say that every dog has his or her day for good or for bad, so I am a little slow to make sweeping comments about any one institution always being good or always being bad, but there are folks that stand out. Helen Branswell, at STAT news, is a veteran public health and infectious disease journalist who has just done some excellent work. Caroline Chen at ProPublica had a terrific piece this week called How to Understand COVID-19 Numbers, and I actually have saved it as a resource, so that is one that I think that I would recommend to others. Maryn McKenna, a longtime public health and infectious disease journalist, a freelancer now, but very active on social media and writes in a variety of formats, is always right up there.
Then in a big newspaper, where it is easy to get lost and difficult to stand out, I have started to notice more and more, I have known about her work for quite some time, but more days that I see she really includes things that I look for that usually get left out, and that is a woman named Carolyn Johnson at The Washington Post. I will stop there, but let me emphasize that is a very incomplete list, but they are people that came to mind.
In general, but not always The Washington Post, the New York Times, ProPublica, and Kaiser Health News. In television, the PBS NewsHour is on my daily must-watch list. Again, that is an incomplete list.
What stands out from that is a good news story that in this deepening, widening valley of the drumbeat of dreck that we see, thankfully some of the big boys and girls are standing out and rising to the need, as they often do. The takeaway there is if your organization can’t do what it is that we will be talking about here, can’t apply the rigor and the time and deliver the context and the balance and the background that is just inarguably required with these topics and these issues at this time then you should back away and not try to do it all yourself because in any of these categories—traditional news outlets, large corporations, nonprofits, and others—in their communications, they obviously can’t excel at everything, and this is one of the most difficult areas to excel. Don’t try to reinvent the wheel, don’t try to stand out if you’re not up to it and if you can’t devote the time and resources that are required. You can still come away looking good by pointing readers towards some of these sources of excellence.
Those are the sources that do not read too much into single studies, avoid single sources, and avoid single patient anecdotes. Those three areas of singularity can be misleading and dangerous. This is what you should look for to avoid and look for in what you want to include in your daily reading. No single study is like Moses coming down from the mountaintop with the word from god to deliver to the people. Any new study must be put into the context of what else has been shown, what contradictory evidence has been shown, what else is being studied that is comparable or is a competing theory. That is hard to do so this goes back to my point that if you can’t do it, don’t, because you might be in over your head and there are others who do it very well.
SCHWITZER: Here I would quote from my JAMA article:
“Government reports, journalism, talk shows, and public relations news releases from industry and academic institutions have often failed to communicate the results of studies well, and these failures have important consequences. Failures of communication include (1) a focus on single study results without the context of other studies or acknowledgment that single studies are rarely definitive; (2) overemphasis on results, particularly relative effects, without recognition of important limitations; and (3) communications based on incomplete reports of studies and reports of studies that have not been adequately reviewed.”
Let me read between the lines here.
So, [in regard to (2)], the use of relative statistical terms to the exclusion of absolute statistical terms, usually in the context of risk reduction, by journalists or by any communicators is one of the most recurring and egregious flaws that we have seen. One of the leading flaws we have seen over the past 15 years.
Relative risk reduction is, as the name implies, a statistic about one intervention relative to another. So when you hear an effect size of 20%, 30%, 40%—so this drug or this vaccine or this test had this impact. With an intervention with a drug or a vaccine, it reduced the risk by 20%, 30%, 40%, 50%, readers should always have red flags going up and just ask two words—“of what?” So, 20%, 30%, 40%, 50% of what? And the “of what” is what gets you to the statistical context you want, which is the absolute risk reduction.
Let me give you a classic example. When Merck marketed Fosamax for osteoporosis, the best of their clinical trial data, which they touted in their marketing, showed a 50% risk reduction of hip fractures in women with osteoporosis. Sounds really impressive doesn’t it? When you ask 50% of what and you look closer at the data, you find that it was a reduction of hip fractures from it occurring in 2 in 100 women in the untreated group down to 1 in 100 women in the treated group—which is indeed a 50% relative risk reduction. But in absolute terms, it is a reduction of 1 in 100. And using another statistic that can be very helpful, the number needed to treat. How many would you need to treat in order for one to benefit? It is 100. You would have to treat 100 in order for 1 to benefit. And you 100 women don’t know which one of you is going to benefit and the other 99 are going to have to take this drug and run the not inconsiderable risk of side effects and pay for the drug. We don’t do this in most of our communication.
And, the third point—“communications based on incomplete reports of studies and reports of studies that have not been adequately reviewed”—we really have to talk about the impact and influence of preprints in this era. Preprints have been in use for a long time, they are not new. They are usually online. And it is a way for scientists to share their work, their progress reports sooner without going through all of the rigorous peer review and time that it takes to get approved and published in a peer-reviewed journal. If you go on one of the leading preprint websites, you’ll see a box at the top of every page saying this is preliminary, and it hasn’t been peer-reviewed, it should not be viewed as guidance for clinical care and should not be treated as journalism as ready for prime-time distribution. And yet, that happens. And that all-important caveat is ignored.
This is a really important point about literature and about scientists publishing their work in any format, whether it is a preprint or in well-established peer-reviewed journals. Those communication formats and delivery methods were not intended for us, the general population, us non-scientists. Don’t get me wrong; I support smart savvy consumers who learn how to navigate that landscape and read journals and learn from them and educate others. While journals are meant as a forum for sharing and discussions among scientists, those of us who, in effect, eavesdrop among those discussions had better know the landscape or we are in over our heads. And we will either not understand or misinterpret. Especially if we only read the abstract and not the methodology. Because spin has been shown to appear in the abstracts of journal articles, spin that only gets unspun when you get down into the results and the methodology and the disclosures about the limitations.
Too many journalists are thrown on this beat or these topics, especially now. So it is just really important that we understand the landscape of the published literature, or else we not write about these issues without seeking independent perspectives.
Failure of communication is one that many might not be aware of, how often they are getting news from these preprints and that the preprint websites themselves warn you are not ready for prime time. At yet it is being delivered to you, and here we are at the bottom of the food chain, and at every level of that food chain this non-peer-reviewed, not fully vetted message is being spun, and the message is being polluted as it flows down to us at the bottom of the food chain and we don’t even know we are being spun. That is why projects like mine were created. To help people become smarter healthcare consumers.
SCHWITZER: Don’t rely on and look for sources that don’t rely on that single expert source, who may have either—and I don’t know which is worse—a financial or an intellectual conflict of interest. You need to find independent sources. Probably the best piece of advice I can give people is just name dropping a list of independent sources to turn to. We offer some on HNR; it is a list maintained by Shannon Brownlee at the Lown Institute and Jeanne Lenzer, a bulldog investigative journalist, and Adriane Fugh-Berman at Georgetown. The three of them and I maintain that list. There are more than 100 names on that list, which our group reviews from time to time and adds to. If you write to us and show us you are a journalist, we send you a much more detailed spreadsheet that lists more than what you see there, lists their disclosure of either no conflict of interest in the past five years or if someone does disclose they have a conflict, but it was in the past or perhaps they offer legal testimony, but our group feels they have something to offer to journalists and we note that, you’ll get the list and their area of expertise and often their emails and phone numbers.
Old saying in science that the plural of anecdote is not data. So when you see news stories and news organizations that tend to want to profile single patient anecdotes…who do you think you get when you go to a drug company or university or any source for a patient story? Do you think you get the dissatisfied people or the trial dropouts or the people for whom the trial didn’t work? No, you get the prized patient, and often those people are not representative of the patient population. When you see single patient anecdotes, I would say run for the hills because that is often a red flag. People need data and evidence far more than they need personal stories right now. Personal stories have their place, but not in isolation minus context and data.
Finally what to avoid or look for, when you talk about data, you should emphasize the limitations of the data—the unknowns and the uncertainties. You should look for sources—whether news sources or expert clinician-researcher sources—in my view, who lead with their uncertainties rather than their certainties, because the uncertainties abound and they vastly outnumber any certainties, and I think there are few right now that exist about this virus and about how to test for it and how to treat it.
The last piece of advice I would give in this arena is to emphasize these uncertainties, the limitations and the strengths or weaknesses of the data early in your message, early in your interview and throughout your message or your story or your communication because to fail to do so is to promote false hope and thereby cause harm.
And there is harm being caused, unintentional though it may be.
Traditional news outlets, large corporations, nonprofits obviously care about their reputation, and they want to show a presence, they want to show that they have something to communicate about this pandemic and they want to look good in doing so. I would just warn you and that is in fact where we ended our JAMA piece, talking about issues of trust are at stake here, trust in science, in medicine, in public relations, in journalism is in jeopardy in the intersection where these professions meet. So, whereas you might go into these communication efforts wanting to look like one of the good guys and women you may come out of it damaging your reputation because you are not able to put the resources into this that are required.
To do this at all requires you that you are in up to your ears and that you make a commitment. If not, there is no shame in saying we are going to point you towards others who do this better.
SCHWITZER: For the first 12 years [at healthnewsreview.org] we looked at news stories that included claims about interventions. So that’s really what we are talking about in this era. How do these new drugs in development look? How do these new vaccines in development look? How do these new tests in development look? And what can we say about the studies being done on these interventions? And [for news stories we reviewed] we applied 10 standardized criteria to the review of those stories. By the time our foundation funding ended at the end of 2018, we ended up with more than 2,600 news story reviews.
But in our last four years [at healthnewsreview.org] of those 12, starting in 2015, we got new funding that allowed us to apply the criteria to the review of PR news releases that included claims about interventions. More than 600 news releases were reviewed and the grades were abysmal.
I never thought those 10 criteria were equally weighted. I always thought that there were some that were more important than others. And 5 of them were more important than the other 5. But we treated them all the same nonetheless. But when we looked at our report cards, those 5 that I thought were the most important were the 5 that consistently and on average got poorer grades for performance on both the journalism side and the public relations side. And here those are:
And across the board of all 10 criteria, PR graded worse than news stories and the news stories grades weren’t very good to begin with.
So what can public relations professionals do today? All of those releases are still on our site. In reading how our team of independent experts—journalists, healthcare professionals, clinicians, researchers, etc.—how they reviewed these claims and how they applied these 10 criteria, there are more than 600 such lessons there free for people to learn from. There is no database like this. Please use these and learn from them while they are still there.
Content that uses emotive language performs nearly twice as well as purely factual content. Learn more in this guide from Acrolinx.
Learn why one page is rarely enough to rank for competitive topics and how to build a content cluster that positions you as an authority in this MarketMuse whitepaper.
Make better content decisions with a system of data + insight.