The art of medicine
6 de octubre de 2023
Por Sean Cleghorn
Comma crunching: a week at The Lancet
Editor's note: This article is part of The Lancet 200th Anniversary Special Issue and is republished here with permission. Find out more about the 200th anniversary(se abre en una nueva pestaña/ventana).
A week at The Lancet tends to start on a Wednesday. Much of what the journal publishes these days is made available online as soon as it is ready, days or weeks before it is bundled up into an issue and dropped through a reader’s letterbox or into an inbox. But the old weekly print cycle, which starts to gear up on a Wednesday, still dictates the rhythm of how The Lancet works. That momentum — the “dum … dum … dum …” repeated beat of issues going to press 51 times per year — has gone largely unbroken for two centuries. The journal has been made up each week since before the existence of steam railways or friction matches. Thinking about it can give me vertigo.
One thing has not changed: editors spend much of their time reading, from new papers and revisions, to peer-review reports and authors’ responses. A huge flow of information passes in front of an editor in the course of producing a weekly medical journal. Where once there was a tottering pile of manuscripts in sun-bleached folders, there is now an online content management system, and Wednesday starts with a meeting to appraise new submissions. The Lancet receives somewhere in the region of 8,000 submissions per year, including around 3,000 original research articles.
So, editors read. They also form judgements about what they are reading. It is an exercise in curiosity. We ask ourselves a lot of questions. What is the hypothesis? Are the methods robust? Do the ethics look right? How does this paper build on what we know? Is this work definitive? And, fundamentally, what is the message of this paper, and is it important for a general medical audience to know about? Answering these questions is not a solitary enterprise. Most papers are read by at least half a dozen editors across various specialty journals. It is a collaborative process and we aim to be democratic. The occasional study is an obvious candidate for serious consideration but much of the time, it is a judgement call. We critique. We tease apart and we reconstruct. We champion strengths and highlight weaknesses. We sometimes disagree. And we discuss. But in the end, one group of new submissions will have become two: a set of papers to be taken forward for external peer review, and a set to be politely declined with expressions of regret.
One old Lancet editor, presumably in a low moment, was once heard to speculate whether readers would notice if we simply took the pile of submissions, tossed them down a staircase, and published those that reached the bottom. Like the chimpanzee that beat the Wall Street fund managers. Given the vast amounts of high quality research being done, are editors’ interventions really all that worthwhile? It is true that The Lancet has to reject many otherwise sound papers for want of space. But quality control is vital to an editor’s work. And it is not just about making sure that checklists are completed and reporting guidelines followed (although there is a lot of that). Ernest Hemingway said that the most essential gift for a good writer is a built-in, shock-proof shit detector; it is pretty handy for an editor too. Outright fraud in the research we assess is rare, but an editor has to look out for a sentence that does not ring true, a conclusion that does not accord with the data, or an argument that does not hang together. Even reports of the strongest research can usually be strengthened further.
There are strands of subjectivity in all of this. Quality is the first priority, but each editor has their own interests that influence their judgement and choices. Editors have their own ideas about what we should publish, what findings are more interesting, and what issues deserve wider attention, with all the attendant biases and privileges of being the ones who choose. A diverse and thoughtful editorial staff will be conscious of their role as gatekeepers and attempt to use it positively. Health equity is central to The Lancet’s ethos, and that entails elevating voices that might not otherwise be heard. But, no matter what, a journal is a reflection of editors and their decisions. As the number of scientific papers published soars, and readers are ever more pressed for time, there is a sense of the editor as curator. The hope is to create a publication that readers trust to be not only accurate and useful, but also challenging, surprising, and provoking.
On Thursday, a lot of the content for the next issue is solidified. There is animated discussion of what is happening in the world of health and medicine. There is also a fair amount of sipping tea and staring into the long distance. Subjects for Editorials—the voice of the journal—are chosen and arguments sketched out. Thursday is also a day for a manuscript meeting, when papers are reconsidered in light of peer-review reports. At each step, the candidates for publication in the limited space in The Lancet are thinned out. Our website might have unlimited space but a print issue does not, and besides, the capacity of The Lancet’s workforce is finite. The production of The Lancet is a colossal team effort. Copy editing, drawing illustrations, processing paperwork, arranging layout, writing press releases: they all require skill, effort, and time. It is one of the quirks of editing that we probably spend much of our time working on papers that will not end up being published in the journal.
Rejecting all these papers (the acceptance rate for research Articles submitted to The Lancet is about 5%, although many papers are passed on to Lancet specialty titles), it can be easy to view editing as an exercise in negativity: a process of winnowing, removal, and exclusion. But there is a positive, constructive side too. An editor’s job is to help authors strengthen papers—to make them as good as they can be, and help authors refine and deliver their message. There is also what editors call convening power. Journals are part of a community, and editors have an opportunity to bring together this community to confront the issues of the day. It speaks to the Enlightenment origins of scientific publishing— the idea of using knowledge to drive progress.
The Lancet has a long history in this kind of work, although in retrospect some looks more pressing than others. In 1902 the journal published the findings of a Special Analytical Commission on Brandy, urgently dispatching a Commissioner to France for extensive sampling. There were, apparently, some genuine concerns about adulteration, but it is hard not to imagine some old fellow with extensive facial hair rattling from one cognac vineyard to the next under a thin veneer of scientific inquiry and a thick cloud of fumes.
These days The Lancet directs its energies to more sober projects. We have published an increasing number of Commissions and Series over the past 20 years on subjects central to human health and wellbeing, ranging from newborn health to the value of death. These are often monumental pieces of work; a typical Commission can take around 2 years from inception to publication. Whenever the thrum of the weekly publication cycle has a lull, editors have a chance to work on projects running on a different timescale. Friday can be a good day for such work for those editors not finalising the sections of the journal that have to come together before the weekend.
By Monday, the print issue is falling into shape and the editorial team becomes ever more frantic. Last points are checked with authors, language polished, images dropped in, and copy shuffled around. Layout is an art as much as a science, and editors can be found in small bands, hunched together over computer screens or page proofs, shuffling text back and forth, cutting a line or adding a clarification.
Tuesday is press day, and the intensity has built to a white heat by early afternoon. Just enough time for any small tweaks. There are final reads and final final reads. Editors have a reputation as sticklers at best, officious pedants at worst, going round unsplitting infinitives, shoving in em dashes, and allowing no exceptions to their arcane and impenetrable rules. At a village fete I once told a jovial but steely eyed woman working the tea urn that I was an editor. “Ah”, she replied with relish, “a comma cruncher.” It is hard not to see the truth in her characterization. Everyone has their pet hates. One former editor managed to prevent the word tool appearing in The Lancet for 15 years at least. Editors have had all sorts of grammar and style rules drummed into them pretty thoroughly—it is difficult to let them go. After just a couple of years working at The Lancet, the absence of an Oxford comma on a billboard or a restaurant menu will be forever noted.
The Lancet’s style guide runs to 82 pages and covers everything from how properly to refer to Christmas Island to the correct orientation of the optic disc in ophthalmological images. There are legacy rules that do not help the journal sound modern (telephone not phone, although editors are permitted to use fax rather than facsimile), next to guidance on how best to cite a tweet. There are plenty of entreaties to use common sense—the worst rule of all to overlook.
By 3 pm, proofreaders have made their last esoteric marks. Editors-in-chief give their blessings. Production editors receive final confirmation that the issue is ready to pass for press and send the journal to the printers. Not so long ago, it was dispatched on the back of a motorbike; these days an email will do. After that comes relief. Press releases are sent out, administrative business tidied up. Some colleagues still recall the days when putting an issue to bed would be marked by passing around the sherry bottle. It is a tradition I wish we could have brought back before remote working put paid to it for good (and who drinks sherry anyway?). So instead, there is just a moment of calm. Take a deep breath. Silence. And then a noise. Dum … dum … dum.