Can you publish too many papers?

Covid-19 has exacerbated the over-production of research, which was already a trend before the pandemic

The dictum that “you can’t be too rich or too thin” is commonly attributed to Wallis Simpson, but was almost certainly coined by Babe Paley, socialite and daughter of Harvey Cushing. The academic equivalent to this would be that you can’t publish too many papers. A 2018 Nature paper by John Ioannidis, Richard Klavans and Kevin Boyack, entitled ‘Thousands of scientists publish a paper every five days’ questions this assumption.

Ionnnidis is a Greek American professor at Stanford University, a founder of the discipline of meta-research – that
is, research about research. He and his colleagues searched the literature for authors who had published more than 72
papers (one paper every five days) in any one calendar year between 2000 and 2016, a figure, they wrote, “that many
would consider implausibly prolific.” They were careful to point out, however, that “we have no evidence that these
authors are doing anything inappropriate”.

They found 9,000 such individuals, 7,888 (86 per cent) of whom were physicists. (It is recognised within physics that these ‘authors’ are simply members of very large teams.) Of the remaining 1,112, 909 authors were Chinese.
They were excluded because Chinese universities openly grant cash rewards for publishing papers. In 2016, the
average bonus for a lead author of a paper in one of the top journals (such as Nature and Cell) was $44,000, which
is five times the average professorial salary. This left 265 hyperprolific authors, whose number grew twenty-fold
between 2001 and 2014. About half of these authors were in medical and life sciences.

“We emailed all 265 authors,” they wrote, “asking for their insight about how they reached this extremely productive class.” Eighty-one replied, citing reasons such as hard work, love of research, mentorship of many young researchers, leadership of a research team, extensive collaboration, working on multiple research areas, and “sleeping only a few hours per day”. The median number of full papers was 677. The authors came from several countries, including the
United States (50), Germany (28) and Japan (27). Malaysia (13) and Saudi Arabia (7) were disproportionately represented; both countries, wrote the authors, “incentivise publication with cash rewards”. The Erasmus University in Rotterdam had more hyperprolific authors (9) than any other institution. Eleven authors worked on one large project
– the European Prospective Investigation on Cancer and Nutrition.

Hyperprolific authors were also concentrated in cardiology. I looked up the publication record of a prominent UK-based professor of cardiovascular medicine and found that he has already published 114 papers in 2021 (up to 15 October): That’s one every 2.5 days. (I would be happy if I read – as opposed to skimmed through – a scientific paper every 2.5 days.) Authorship of a scientific publication means taking credit and responsibility. The Vancouver criteria for authorship specify that authors must do the following four things to qualify: Play a part in designing or conducting experiments or processing results; help to write or revise the manuscript; approve the published version; and take responsibility for the article’s contents. Although supervision, mentoring and sourcing grant money are not sufficient for authorship, many authors only became hyperprolific when they became heads of department, suggesting that they become co-authors regardless of their contribution. Ioannidis emailed a survey on authorship to 81 hyperprolific researchers, of whom 27 replied. Most admitted that they did not fulfil all four Vancouver criteria.

The Covid-19 pandemic has only exacerbated research over-production. In December 2020, Nature reported that four per cent of global research output in 2020 was devoted to coronavirus, and that 2020 also saw a sharp increase in article submission on all subjects to journals. Submissions to the publisher Elsevier’s journals (such as The Lancet) were up by 58 per cent compared to 2019. The journal speculated that this rise might be because researchers had to stay at home, and so concentrated on writing papers. Many papers (an estimated 17-to-30 per cent of all Covid-19 articles) were published as preprints on sites such as medRxiv.org. We have also witnessed the phenomenon of ‘epistemic trespassing’ – scientists from other disciplines, who have no background in virology, epidemiology of infectious disease or vaccine development, have begun to publish papers on Covid-19.

A similar phenomenon was observed after the 11 September attacks in 2001, when unqualified researchers, attracted by the large grants available, began to publish on terrorism. I have known a few hyperprolific researchers – I do not envy them. To be identified as one should be a cause for concern, not a badge of honour. That their emergence is the inevitable consequence of the publish-or-perish culture that has dominated academic medicine for decades. Over-production is incentivised because candidates for senior academic posts and applicants for research grants are judged by these metrics. But these figures are now increasingly meaningless: The published product of medical research is now a devalued currency.