Dan Gorenstein: Early this year, scientists started working at lightning speed to understand the coronavirus.

Moving swiftly means saving lives.

But there’s one problem.

Richard Sever: The traditional publication process takes on average about eight months.

DG: That timeline does not work in COVID times.

So researchers have rallied behind another option — one that allows them to publish their findings in just days.

But it can come at a pretty big cost: accuracy.

From the Annenberg Studio at the University of Pennsylvania, I’m Dan Gorenstein, and this is Tradeoffs.

DG: Traditional publishing relies on a peer review process — it’s cumbersome but it means the work has been vetted.

RS: In the traditional publication process you send your paper to a journal and then they spend a period of time consulting other experts to get their feedback.

DG: That’s Richard Sever.

He helps run Cold Spring Harbor Laboratory Press, which publishes leading research in genetics, neuroscience and cancer biology.

For more than 100 years, scientists have relied on journals — The Lancet, Nature, The New England Journal of Medicine — to share groundbreaking research.

This is how most discoveries make their grand entrance.

But really scientists have been sharing information in a much more informal way for just as long.

RS: This is why people go to meetings, they’re like, I can’t wait to read all about it in a journal in a month’s time. I’m going to go to the meeting, find out what’s happening, and then I’ll go back to my own lab and I can start working and moving science forward.

DG: In 2013, Richard decided the time had come to formalize the back-alley way scientists shared cutting edge biology information and he launched bioRxiv.

bioRxiv is an online site where researchers can post their papers, called preprints, within days.

Editors conduct basic screenings — a review for dangerous or non-scientific content and plagiarism.

But there are no edits and, critically, there’s no peer review.

RS: This is like a really, really big meeting and everyone’s invited, which is basically to say, let’s use the power of the web to allow scientific results to be disseminated almost immediately to anyone in the world who wants to read them.

DG: The popularity of bioRxiv quickly ballooned — with millions of downloads every year.

So Richard with colleagues from Yale and the British Medical Journal added a second site in 2019 for medical or clinical research called medRxiv.

Before the pandemic, medRxiv received about 15 papers every day.

Then COVID hit — and the dam broke.

Now, it’s north of 100 a day.

Lately, between both medRxiv and bioRxiv, there are 30 million views each month — almost four times the normal traffic.

RS: I think the pandemic has really underscored the value of this rapid dissemination. As the whole clinical world is trying to figure out ways in which we should or should not treat patients.

DG: Richard created bioRxiv and medRxiv with a simple goal – share more data quicker.

In this crisis, scientists are telling Richard the sites are helping reach treatments and new knowledge faster.

Exhibit A: the paper from molecular biologist Nevan Krogan at University of California San Francisco who is studying how the virus enters our cells.

RS: Essentially what they did was they they looked at all the interactions within cells when it’s infected by the virus. So basically, you get a big map of what the virus is doing and all the points in the cell that you could potentially look to block in efforts to combat the virus. So it gives you some form of kind of roadmap to start developing approaches to therapy.

DG: The paper was a hit.

Krogan posted it to bioRxiv in late March.

Nature published the research six weeks later, and the article has been viewed over a quarter of a million times.

This paper is a best-case scenario for these preprints.

But, Richard is quick to concede, working so quickly comes at a cost.

RS: You have to make a choice. You can’t be fast and thorough. And what preprints do by doing this, decoupling it, saying, you know, we’re going to we’re going to release this information very fast couple of days in the case of bio archives, three or four days in the case of medRxiv. But that comes with a big caveat emptor sticker on every paper. These papers could be completely wrong.

DG: Can you give us an example of a bad study that you all published that either needed to be retracted or was problematic in some other way?

RS: The infamous uncanny paper, which seemed to have identified similarities between the virus and HIV. And this sort of led to some conspiracy theories. Does this mean that this was a secretly engineered bioweapon or something like that?

DG: bioRxiv posted the paper in late January and reaction was swift and damning.

RS: Literally within hours there were tons of scientific experts saying this is not correct. There was something like 50 to 100 comments within 24 hours by really eminent virologist explaining why this paper was inaccurate. And then two days later, the paper was withdrawn and the authors said they were wrong.

DG: The uncanny paper illustrates the dangers of preprints, and the safeguards.

There’s an open-source quality to these two sites, and Richard counts on scientists policing the work — an informal peer-review process, really.

bioRxiv and medRxiv do allow more work — in some instances, of dubious quality — to get shared.

Ultimately, preprints put the onus on us — the scientists, researchers and lay audience — to determine the value of what we’re reading.

RS: It still may be the worst paper you’ve ever seen. And you, as an expert, have to read it thoroughly and make your own judgments. But if you’re an academic who’s working in a field who really wants to move forward, you will accept that every time. And that’s why you go to meetings because you need to know soon and quickly. Now, for me, if there’s a paper in my field then I might read a preprint. If the paper is not in my field, I don’t understand it, I’ll probably say, you know what? I can wait a few months for somebody to filter it, for somebody to do some checks for me.

DG: And here’s the biggest danger: there will be non-experts who run with bad information or experts who get it wrong in the first place, and that shoddy science could inform high-stakes choices like how to treat COVID.

But to all the preprint handwringers — Richard asks whether the alternative is much better.

News clip: A respected British medical journal retracted a study that the MMR vaccine may trigger autism.

DG: The most revered journals have let junk science through the gates with a paper in The Lancet that linked the vaccine for measles, mumps and rubella to autism being perhaps the most infamous example.

It took The Lancet 12 years to retract that paper.

DG: If any doubt about the value of this work ever creeps into Richard’s mind, he thinks of a talk he had with his dad, who is a physician.

RS: When I launched medRxiv, I mentioned that to him and he said, um, “oh, I’m not sure about. I’m not sure, that’s clinical information that, that could be dangerous. I’m not sure that’s such a good idea.” And then I spoke to him like a couple of weeks later and said, “oh, you know, I’ve changed my mind. I think actually it’s fine. All the crap gets published somewhere anyway.”

DG: Preprints are here to stay. The pandemic has only solidified this.

But Richard has even higher hopes — he sees these sites becoming part of the traditional academic publishing process — with papers being judged first by 100s of expert eyes before crossing a journal editor’s desk.

Preprints may be a messier means to the end of cures and other breakthroughs.

But if it gets us there faster, that’s a tradeoff Richard says he’ll take every time.

I’m Dan Gorenstein, this is Tradeoffs.