Too much pressure to publish? ‘Publishing is important for humanity’

09-11-2018

Some researchers publish an article every five days, according to an analysis in ‘Nature’. Radboud university medical center professor Mihai Netea is one of them. ‘I never just write an article.’

It was a remarkable analysis that was published in Nature recently: there are a few hundred scientists worldwide who published at least 72 articles a year in at least one year between 2000 and 2016. That is one every five days, an exceptionally high output. In comparison: the Rathenau Institute calculated that a Dutch scientist writes 1.8 articles a year on average.

The author of the analysis, Stanford professor of epidemiology John Ioannidis wants to start a discussion about what a publication means in the current scientific system. Are authorships not handed out too easily? Some of the articles by the ‘super publishers’ list large numbers of authors, sometimes hundreds. Have they really all contributed significantly?

Stanford professor John Ioannidis.

Those questions are relevant as the length of the list of publications plays an important role in the allocation of money and jobs in the academic world. They are the currency of academia. The phrase ‘publish or perish’ is familiar to all researchers, and unions and critical scientists have been complaining about too much pressure to publish for years. Radboud university medical center expects their head researchers to publish at least six articles with an impact factor (a measure of journal status) of eight or more every five years.

One name jumps out on Ioannidis’ list: Mihai Netea, the only one from Nijmegen. Netea is a professor of Experimental Internal Medicine at Radboud university medical center, and scored 76 publications in 2016. He thinks that the attention to super publishers is ‘a storm in a teacup’.

What do you mean?

‘What does the article in Nature want to show? Like a lot of things in nature, publication numbers follow a normal distribution, a Gaussian curve. There are always people who publish a lot, just like there are researchers that publish very little.’

76 in one year, that is a lot.

‘Coincidentally, 2016 was a special year for me regarding publications. Usually I average about fifty – not much different from Ioannidis, by the way. And for about nine or ten, I am the last author. Nothing special. I don’t feel the need to be defensive about it.’

‘If you have good focus and structure in your work, you can accomplish a lot. As an internist, I have the tendency to collaborate. Our research group at Internal Medicine works together closely with all researchers in the department; this is how our structure works. This means that there are a lot of joint publications. I collaborate a lot internationally too, for example, in two large international consortia (large collaborative projects with dozens of research groups sometimes – Editor) where a lot of articles are generated.’

The recently revised Code of Conduct for Research Integrity by the Association of Universities in the Netherlands (VSNU) sets clear criteria for authorships. Researchers need to make a tangible intellectual contribution to the study, need to have read the final version, and are responsible for the full academic content.

Can you always fulfil these criteria, with such a large output?

‘I always fulfil the author criteria. I never just write an article! I make relevant and continuous contributions to every study. I’m not talking about suggesting a control experiment during a large group session – in that case, you might as well designate everyone who adds something relevant during a conference as co-author. No, I’m directly involved with the coordination and decisions within the project groups. Those are meetings with three, four people.’

‘I read and correct all of the fifty articles a year – so one a week. This takes me one to three hours, depending on how well someone writes. What if that didn’t fit into my working week? It is part of my job, and that is the way it should be.’

‘Science revolves around trust.’

Dozens of your articles list twenty or more authors. In a few instances, for large collaborative projects, there were hundreds. Do they really need to all be named?

‘If we contributed samples, or did a small control experiment, that is not necessary. But if it gets more complicated, a researcher has to get involved and therefore needs to receive an authorship. You can’t simply have them do lots of experiments and then just say ‘thank you’.’

How can you take responsibility for the full content, as the VSNU wants?

‘That is very difficult indeed, I also don’t know how to best deal with that. You can’t keep track of all parts. I cannot guarantee that no one has ever pipetted something incorrectly (using a lab instrument to measure and move a liquid – Editor). No one can guarantee that 100%.’

‘For me, science revolves around trust. Cases like the one with Diederik Stapel do not mean that everything has taken a turn for the worse. We should be on the lookout for malpractice, but we should not overdo the quality control. That is not feasible either. If someone really wants to do something wrong, they will manage to do it. You can only be certain of the content if you look over someone’s shoulder in the lab. But how do you know that I can be trusted? Do we need a third person for that? And how do you know that I didn’t pay them off?’

‘The current system is not perfect, but it does a good job of filtering itself. When a result is incorrect, it will quickly show up if subsequent research cannot confirm it.’

Is the focus in science too much on quantity instead of quality?

‘It is easy to say that less should be published. But say that you publish one article in two years. Can you justify your work towards your funding sources and society? And does that make the article better than those of someone who publishes more?’

‘Of course there are disadvantages to this system, but it also encourages quality. Ever since Radboud university medical center set publication criteria for researchers, the number of publications and the quality have soared, just like the number of collaborations. It would be a pity if we had to reverse that because of a few drawbacks in the system, and in turn, lose the incentive to try our best.’

So the ‘publish or perish’ culture has not gone off the rails?

‘Publishing is important for science and knowledge in general; I don’t just publish for my career. If I treat a patient with rare symptoms, I consult the literature. If someone treated a case successfully, but didn’t publish it, I would not know. My patient’s chances for receiving good treatment and recovery would be diminished. If Pasteur had not published his findings, we would not have vaccines. It is not ‘publish or perish’ for researchers, but ‘publish or perish’ for civilization!’

Leave a comment

Vox Update

A daily or weekly newsletter with our articles in your mailbox!

Weekly
Dutch