Earlier this week I posted a roundup of scary stories, or bad news yarns, concluding it with reference to journalist and risk analysis consultant David Ropeik's complaint about one of the stories, on the AP. It was about pollution due to drilling, such as for fracking operations. Ropeik declared that, by not making clear the distinction between reports by the public of polluted water due to drilling and actual cases persuasively verified by expert investigation, it spawned a lot of headlines suggesting a problem much larger than the data support.
The tracker post told you readers that I had sent an email to him about another story in the roundup asking, if he had time, what he thought of its handling of risk. He has since replied at some length. Reporters who cover technical fields and the risks they may pose will surely find his analysis of great interest, as will most other readers of the tracker. So, after a re-posting of the short section I wrote about the story in question, his full response follows. It goes well beyond my few lines on this story that I appreciated very much – even though it is not good news.
- New Yorker – Dana Goodyear: Death Dust / The valley-fever menace ; A long and detailed story that yo yo's between vivid vignettes on the fairly rare person who breathes fungal spores in the windblown dust of the American southwest and gets very sick, and persuasive health statistics showing that this sometimes near-incurable and fatal disease, cocci, doesn't have to have a high batting average to create tremendous misery and challenges to the nation's health system. Not only that but as the nation warms, Goodyear reports, valley fever's epidemiological footprint is growing larger.
And Ropeik has this to say:
The piece on Valley Fever passes many of the basic requirements of Risk Reporting 101 [See a Ropeik post at Columbia J. School's The Observatory in 2011 that explains the term/ cp] It establishes that the agent in question is a hazard, and what it does, and how it does it, although there is no clear idea of the dose necessary for infection. Will one spore do it? (To be fair, that may be unknown, or tough to calculate. This isn’t like a poison.) And she establishes that there is exposure, the other basic component required for anything to be a risk, and she fills in some of the important details here too: which populations are at risk, both demographically and geographically, and that inhalation seems to be the exposure pathway that matters rather than dermal exposure or ingestion.
But there is one big flaw in the Risk Reporting 101…the failure to include both absolute and relative risk numbers of victims. Both are important to give the reader perspective on the scale of the risk. You can see this when the author offers some general stats.
"Every year, there are some hundred and fifty thousand cases.” We learn the absolute number of victims, but not the relative risk; 150,000 out of how many? If that 150,000 figure is for all of America, the overall rate is tiny. Even if it’s just for the entire population in the region where the fungus is endemic, the rate is still tiny.
And then she writes "Only forty per cent of people infected are symptomatic, and the signs—fever, cough, exhaustion—can be hard to distinguish from the flu.” So now we’re down to 60,000, which is still a large number in absolute terms and worthy of concern, but probably an even tinier relative risk. 60,000 get sick out of how many potentially exposed?
Unfortunately, the author then slips into a more alarmist mode…in the rest of the graph she writes,
" A small subset of patients will suffer long-term health problems;” Okay, so a small subset of the 60,000 get really sick, but she doesn’t tell us how small. Then she divides the at risk population further
" in fewer still, cocci will disseminate from the lungs into other tissue—skin, bones, and, often fatally, the meninges of the brain.” But now we’re talking about a really small, though unspecified, subpopulation.” So now we’re down to a small subset of a small subset…yet that’s what she spends the rest of graph describing in dramatic detail,
" For those with cocci meningitis, the treatment can be brutal. Three times a week, in the hospital, patients are administered an anti-fungal called amphotericin B—“amphoterrible” is how doctors refer to it—with a needle to the base of the skull. To prevent headaches, patients sometimes rest for several hours with their feet elevated above their heads. One patient, a twenty-six-year-old white woman who caught valley fever four years ago, told me that the medicine made her vomit non-stop on a negative incline. She was temporarily paralyzed, underwent three brain surgeries, and has had twenty-two spinal taps. Not long after her diagnosis, the doctors told her mother to make funeral arrangements. Now they tell her she will be on anti-fungals, funnelled through a shunt in her brain, for the rest of her life. That certainly seems to be emphasizing the severity of the risk, while downplaying the likelihood of such severe outcomes.
It’s not that those symptoms aren’t awful and worthy of note, but there is a pretty clear imbalance here in favor of playing up the scariest aspect, which apparently only threatens a tiny percentage of those who get sick. And we never learn just how many get REALLY sick in either absolute or relative terms….just how scary it is when they do. There is some basic and IMPORTANT information missing here, and leaving it out tends to play up the alarm.
Maybe just a caveat at the end of the graph to reminds us of that fact might have been more fair to the overall nature of the threat.
Another big problem I had is her reliance on Dr. Galgiani as a source. He’s the guy who says early on “The impact of valley fever on its endemic populations is equal to the impact of polio or chicken pox before the vaccines,” REALLY? POLIO before vaccines! POLIO!!! That’s pretty dramatic, and she lets that quote land without any challenge. This is also the guy who goes on to say "But terrorists don’t care about that stuff—all they care about is perception. A single cell can cause disease, and you can genetically modify it to make it more powerful.”
and
“We sometimes talk about wishing a President or former President would get cocci,”
So this guy is sounding all sorts of really dramatic fears. But it’s not until late in the piece that we learn that that Galgiani is "a partner in a company that is trying to ready a molecular byproduct of the bacteria streptomyces, called nikkomycin Z, for the marketplace, to control the risk.
It’s not so much that this poses a conflict of interest. I think it is trite and too easy to imply that making money automatically makes anyone dishonest, a charge that is too readily and naively made. But I do think that in a longer piece, in a periodical like this, the reporter needs to do a fairer job of noting the deep emotional as well as financial and career interests of a source who makes such dramatic statements. Dr. Galgiani’s level of alarm really does need to be put in context.
One last item that illustrates some unfair alarmism; in the interest of narrative, and drama, the story ends with the suggestion that this regionally isolated disease could spread, "in the air as in the body, spores can go anywhere they want: in shipping containers to Hong Kong,…” It’s pretty clearly the author saying to the reader who lives outside the endemic area “Hey, pay attention. This could be a risk to YOU!” That is scientifically incomplete and unfairly alarmist. This fungus appears to be constrained by environmental conditions. Yes, it can travel, as she notes at great length with the scary tale of the fellow in Michigan who got it via the vector of a used car, which she describes at great length to close out her narrative. But for any exotic to become invasive…to catch on and thrive in the new places it shows up (West Nile virus)…it needs to find favorable conditions. Otherwise the cases it causes will be REALLY rare one-off sorts of things, like the one she found, but nothing worth worrying about. Which needed to be noted…
But of course such a caveat would diminish the alarming news that the author seems to be reporting, that YOU COULD BE AT RISK!!!!!!!
But of course such a caveat would diminish the alarming news that the author seems to be reporting, that YOU COULD BE AT RISK!!!!!!!
——————-
Thank you David. It shall be interesting to hear from any tracker readers who feel the same, or different. It did seem to me that the absolute number here, 150,000 cases, is indeed enough to make health authorities sit up straight and for me to feel nervous when the wind blows Central Valley dust my way. But I took that to mean cases with symptoms serious enough to get the at times intense and painful treatment that writer Goodyear describes, not – as Ropeik interprets it – the total of those infected but most of which have few if any serious symptoms. The story is ambiguous. It surely should have explained what 'case' means. – Charlie Petit
Leave a Reply