On Wednesday, journalist John Bohannon revealed to the world how he "fooled millions into thinking chocolate helps weight loss." In a boastful piece for i09, he details how he and German television reporter Peter Onneken performed a faulty clinical trial and used flawed statistics to make it seem like chocolate was a weight loss wonder. The team then wrote a bad paper and managed to publish it in a (non-peer-reviewed) journal. They intentionally concocted an enticing press release to tell the world about their not-so-reliable results, and managed to get a few large sites to bite the hook they carefully baited. "For far too long, the people who cover this beat have treated it like gossip, echoing whatever they find in press releases," Bohannon wrote to explain why he agreed to the elaborate sting. He hopes that the shame of being called out for bad journalism will be enough to get reporters and the public to be a bit more skeptical of science news. Of course, some were quick to point out that Bohannon mostly fooled the most well-known churnalistic sites, and that overall, science journalists didn't fall for the ruse. I'm inclined to agree with their criticisms both of ethics of how the sting was conducted and the bold conclusions about the lazy nature of science journalists drawn from it. But it's hard to stand on my soapbox, fist in the air, when it seems like every week there's another example of just how shoddy science journalism often is, even when the studies reported on are actually quite wonderful.
Diane Brinkman is first author on a new paper that tells us more about the venom of the deadly box jelly Chironex fleckeri than ever before — too bad the news media has done such a shoddy job of reporting about it. Image from Wikipedia. You see, I'm in a particularly sour mood because I didn't want to bring up John Bohannon or the failings of science journalists today. Instead, I had planned to write this awesome post about a fascinating new paper published in BMC Genomics. I wanted to talk about how this research (which details the venom transcriptome and proteome of the largest of the deadliest class of invertebrates in the world, the box jellyfish Chironex fleckeri) is an incredible, fresh look at an evolutionarily old venom. I wanted to expound extravagantly on the novel toxin types Diane Brinkman and her colleagues from Queensland found in the terrifying tentacles of a species that has killed more than 60 people and caused serious injury in multitudes more. Most importantly, I would have loved to dive deeply into the study's methods and results, discuss what this new information tells us about some of my favorite venomous animals, and how it builds the foundation for future studies. But instead, I was so nauseated by the coverage of this study that I feel obligated to take the time to correct the lazy reporting of others. Bohannon's chocolate fake-out may not have been right, but it's hard to say he's wrong about science news coverage. For example, most of the articles about Brinkman's paper focused on implications for antivenom. "Box jellyfish antivenom a step closer after breakthrough by Queensland scientists," says the headline for the Australian Broadcasting Corporation (ABC), a government-funded conglomerate that claims to reach 71% of Australians and whose site receives over 6 million visits every month. The International Business Times (IBT) hailed that "Queensland Scientists Discover New Toxins That Could Help Create Box Jellyfish Sting Antivenom," saying that "developing the antivenom for the box jellyfish sting could take a lot time...they could probably accomplish their goal of developing the antivenom in five or 10 years." Smaller audiences were given similar news, with outlets in China, India, the Philippines and Canada all publishing under the title "Box Jellyfish anti-venom closer: Australian scientists." There's one not-so-tiny problem with all of these headlines: Brinkman's research wasn't about antivenom. In fact, the word 'antivenom' does not appear in the paper. Not in the abstract, introduction, methods, results or discussion. Not once. Where did the antivenom idea come from? I have no clue, considering it's not in the research institute's press release, either. Once it was brought up, though, it was perpetuated through incestuous sourcing — news articles riffing off news articles that riffed off of ABC's article about antivenom. Why look for a new angle on a story, I guess, when you can just use someone else's? (Oh yeah — because that's not journalism.)
Seriously, not even one "antivenom" anywhere. It's not even in the titles of related articles or the 56 references cited, FFS. Such headlines are more than misleading. While there may be a role for proteomics and transcriptomics when looking at the future of antivenom production, at the moment, antivenoms can be (and usually are) produced without knowing all the toxins involved. Bryan Fry, a world-renowned venom scientist whose research bread and butter is examining venoms using combined proteomic and transcriptomic approaches, said that such studies have little to nothing to do with antivenom production, as "antivenom development doesn't really require in-depth knowledge of what is in [the venom]." Not only is the paper not about antivenom research, and thus the news articles fundamentally misunderstood the entire point of the paper, but to make matters even worse, they all implied that scientists are in search of a Chironex fleckeri antivenom (remember, they said an antivenom was five to ten years away!), when in fact, there's already one available. Even a cursory Google search would have told any of the reporters, editors or fact-checkers involved that a Chironex antivenom exists. I'm guessing the antivenom powerhouses at CSL would be insulted if they realized the antivenom that they have produced since 1970 is so easily forgotten. Is it the best antivenom in the world? That's debatable. Other scientists have shown that Chironex antivenom may even hasten death (in an animal model of evenomation), and there are frequent and sometimes severe negative side effects from injecting humans with animal blood proteins (which is what antivenoms are — antibodies and sera from venom-inoculated animals), a fact which tempers the whole idea that antivenoms are some clinical silver bullet. But more importantly, given that box jellyfish stings can kill before the victim has even made it to the beach, there's not a whole lot of reason to think improving antivenom treatments (if that's even what the paper were about) is the best way to save lives. You'd think that there would have been more discussion on that point, since the ABC talked to an 'outside expert' of sorts — Surf Life Saving North Queensland regional manager Colin Sparkes — who noted that antivenom had to be administered in two minutes or less. He's not an outside expert on the paper's actual topic, but at least he provided some insight into why it's so hard to create a good antivenom for this species. The Courier Mail, though, takes the prize for biggest off-target leap: "Queensland researchers hopeful deadly box Jellyfish venom holds cancer answers." Say what? There's not a single mention of the word "cancer" in the paper, nor should there be. As I explained in a recent article for Mosaic, the process of drug discovery from venoms is long and arduous. Venom experts like Glenn King (who actually works on turning venom compounds into pharmaceuticals) already worry that headlines pronouncing potential "cancer cures" and the like long before compounds show clinical promise will lead to greater distrust of scientists and the scientific process. The Courier Mail's headline is a disappointingly perfect example of that kind of attention-seeking drivel; this study didn't even begin to examine whether any of the toxins had anti-cancer properties. (Fry said connecting this study to cancer was "shameless".) The articles continued downhill from there, particularly overselling the research's novelty. "Scientists have analysed deadly box jellyfish venom for the first time" touted The Courier Mail, and "there has never been any research into what causes such severe reactions," explained ABC. That's a funny thing to say when the first author alone has published several papers on box jelly venom proteins. Oh, and earlier this year, another transcriptome and proteome was published (coauthored by none other than Bryan Fry), which comes up as the second Google Scholar result if you search "Chironex transcriptome".
If a media article says something is "first", "ground-breaking" or that there has "never been any research" on a topic, it's a good idea to do a real quick Scholar search to see if that's really true. That's not to say that the research done by Brinkman and her colleagues isn't important. Fry is the first to admit that the new paper contains novel insights."They did an in-depth job and certainly brought new elements to light," Fry told me. "They went deeper than we did." You'd never know that from the vague statements about the "discovery" of "new toxins" in the media pieces. None of those articles actually say what those toxins are, or even how many new families were discovered (the team found evidence for ten different toxin families, if you wanted to know, including metalloproteases, lipases, lectins and others found in well-studied species like spiders and snakes). They also found that there are far more variants of the known toxins present in the venom than ever realized. Previous research on these toxins — nasty, incredibly fast-acting, pore-forming proteins or 'porins' that tear holes in cells — has shown that they are the main driver of the venom's unfathomably fast, lethal activity, and that more lives may be saved by inhibiting them rather than focusing on antivenom. Why do the animals make so many variations on the same theme? That's a good question — one not answered in this paper (that wasn't a part of the experimental design), and one not asked by any of the reporters. The other thing you can't find in any of those articles on the paper is what the team actually did. The Courier Mail called it a "detailed chemical analysis." ABC explained that "scientists took samples from jellyfish venom and searched its contents against a database of proteins." The only place that the words "transcriptome" or "proteome" appear is when IBT repeats the study's title verbatim. I normally wouldn't harp on the use of such jargony words, but it's important to note the combination of approaches used, and it's not hard to explain the big words to do that. This isn't the first study to look at box jelly venom; there have even been multiple forays into categorizing the proteins in Chironex venom in particular. And as I mentioned before, there's even an article that looked at both the transcriptome (the collection of 'transcripts' or messenger RNA molecules which indicate what genes are expressed in the tentacles) and the proteome (the collection of complete proteins found in the box jelly venom-producing cells) in the exact same species. What makes this study unique are the scale and depth of the sequencing and biochemical analyses, not the research question. And even still, there is more to be done. The authors note that they struggled with the data obtained in the transcriptome work in particular, as almost half of their sequences didn't match to any of the toxins they discovered. To a geneticist, that's a lot of data left unexplained. A genome would certainly help scientists figure out what proteins those sequences belong to or what they may do, but there isn't a genome for Chironex fleckeri — yet. That's another step for future research to take. Why do I care so much about a study about box jelly venom toxins? Sure, I'm a little bit biased because I study venoms. But why this really matters is that it's an example of the everyday failures of science communication. It's not a study about some potentially-blockbuster pharmaceutical or a new way to lose 20 lbs. It doesn't have that flashiness, that oomph and pizzaz that makes for an obvious if cliche headline. There are lots of writers and scientists out there that will stand up and speak out against showy bad science coverage, like the pseudoscience of Food Babe or the anti-science of what seems like every member of Congress. This paper isn't political or even really medical — it's just good science, and I'm tired of watching good science be undermined by bad journalism. Bohannon hoped that the chocolate-weight loss scam would rattle reporters and shame their employers into better practices. I doubt the big whigs that commissioned pieces about it even noticed. For the news outlets that create that kind of crap science content, the cost of better journalism is outweighed by the benefits of cutting corners. Catchy headlines attract eyeballs whether or not those headlines reflect the research. Regurgitating press releases or even other news articles saves time, and thus saves money. So long as their tactics keep audiences, and thus keep the ad buyers happy, there's no push to change the status quo. They don't care about the ideals of journalism or even science — they care about their bottom lines. Such terrible reporting not only is a disservice to the science, it erodes the trust placed in both journalists and scientists. In the end, everyone loses: non-scientists are left in the dark (including about what their tax money pays for!), scientists are forced to fight harder for funding and credibility amid back-breaking pressure to publish or perish, and the journalists get undercut if not replaced by anyone who can pump out a 500 word piece of crap that brings in pageviews in an hour or less. Maybe Bohannon is right that by pointing out poor science coverage, the public as a whole will become more critical of what they read and hear. I still don't agree with the way intentionally misled millions (by his count), but perhaps there is a glimmer of upside to what he did. Maybe those who read his confession — or some of you who read this post — will gain from peeking behind the proverbial curtain. I hope so, anyway. It lowers my blood pressure some to cling to the hope that posts like this one make a difference. Citation (Open Access!): Brinkman, Diane L., Xinying Jia, Jeremy Potriquet, Dhirendra Kumar, Debasis Dash, David Kvaskoff, and Jason Mulvenna. "Transcriptome and venom proteome of the box jellyfish Chironex fleckeri." BMC Genomics 16, no. 1 (2015): 407. DOI: 10.1186/s12864-015-1568-3
FYI, there was one outlet that had decent coverage of this paper: Genome Web. So if you want to know what the scientists actually did, head on over and read their article with a perfectly reasonable headline: Box Jellyfish Transcriptome, Venom Proteome Reveal Sting Mechanics, Could Inform New Treatments (but it is, perhaps, a little technical)