Prof. Lex Bouter

Rector magnificus of VU University Amsterdam and professor of Epidemiology.
Lex Bouter chaired of the Society for Epidemiology, the committee of Dutch guidelines for pharmaco-economic evaluations and the Cochrane Collaboration Back Review Group. He participated in a committee of the Royal Academy of Arts and Sciences (KNAW) which concentrates on the evaluation of social relevance of applied health research, is member in the Health Council of The Netherlands, and holds a position as methodologist in the Central Committee on Research involving Human Subjects. (CCMO).

Are people in the medical field familiar with the phenomenon of Open Access?
When you consider the health and life sciences, including medicine, it’s clear that there have been major developments in the past few years. The BioMed Central journals and PloS Medicine are two examples. After an initial healthy scepticism, those in the field now realise that an effective system of reviewers has been set up. People also like being able to access earlier versions of an article and the open referee procedure for BioMed Central, which means that you can follow the whole history. Moreover, page limitation no longer applies, meaning that you can now publish your research protocol beforehand and devote full attention to your measurement tools. There’s also room for data files. It’s also ideal that the article is published as soon as it has been approved. You no longer need to wait until there is room for it in a journal. But ultimately it was the rapid increase in the impact factors that was decisive for Open Access. That was really impressive for some journals.

As far as I can see, people in the medical field are enthusiastic about Open Access. The trouble is, a lot of those who work at university medical centres seem to think that Open Access is already a matter of course. After all, they can access everything and they don’t seem to realise that that only applies to 'the happy few'. That’s a bit naive, but it’s how they think. I gave a course in clinical epidemiology in Bucharest and referred to a whole list of useful websites that make it easy to search electronic journals. After about half an hour, a member of the audience hesitantly asked me “What do you actually mean? We can’t access any of those journals.” It was rather embarrassing. But it’s not much different in the Netherlands. I know people who are engaged in really useful research, but are not affiliated with a university medical centre. They can only access the literature via all kinds of roundabout routes, for example by using a colleague’s account. The same applies to patient organisations or science journalists.

The other side of the coin are the costs. The bodies that finance research are not yet prepared to include the cost of publication in the cost of a project. My own library was prepared to sponsor a number of articles, but only a few because the library’s budget is tied up in subscriptions. That makes the transition from old to new difficult.

So those in the medical field know about Open Access, but is it actually used?
I don’t have any figures, but I am aware of an increase in my own circles. I would estimate it at about 20% to 25% of publications. But people mainly select the top journals. The British Medical Journal is fortunately Open Access, but The Lancet or the New England Journal of Medicine, for example, are not. Only if they can’t publish in one of those journals do people approach another journal or publisher, and then an Open Access Journal is an option. The costs do then play a role also because people seem to think that everything is already Open Access.

The most important thing as regards the further growth of OA is not so much the provision of information but above all that we make OA easy. All those recommendations to the effect that authors should not give away their copyright have no effect. Authors aren’t bothered. They get presented with all kinds of forms by the publisher and they sign them without really reading them. They just want to be published, and negotiating with the publisher about copyright doesn’t interest them.

Don’t they find it a problem later that they are no longer permitted to use their article for teaching or to put it on their website?
They simply go ahead and do that anyway. Scientists have had years of training in opposing bureaucracy. And they fail to notice the consequence that the institution sometimes needs to pay to use their article.

What about including the article in their institution’s repository?
The disadvantage of repositories is that it’s all such a mess; the worst thing that can happen to a researcher is that a whole lot of different versions of his article are knocking about. All the citations need to refer to the same thing, and if a researcher has to go through a great deal of trouble to ensure that they do, then things aren’t working properly.

The route to Open Access needs to simply be the easiest route. Don't make it more difficult for researchers, for example with a website like Sherpa (a website with publishers’ copyright rules, L.W.). No researcher is going to get enthusiastic about that. He simply thinks “My article’s finished; I need to send it out and get it published.” All the rest doesn’t interest him. Providing information for researchers only works if you tell them that publishing by means of Open Access is much easier; it’s a lot faster; you’ll be cited more frequently; and all you need do is something really simple. At the moment, we tell them all about repositories. It takes them a week before they can pronounce the word and when they finally understand what a repository actually is, they don’t want to have anything to do with it. I don’t see any point in it. You can use a repository for 'grey' literature, documents, dissertations, data files...terrific! But for journal articles you have to use Open Access journals, and you have to have serious peer review. Because that’s the problem with repositories: people just dump their stuff in them. But I don’t want more information, I want better information and preferably less information.

How can things be made easier when it comes to filling the repositories?
Scientists increasingly have their own personal website. The easiest way for them to organise their CV is to ensure that that personal page is properly set up. You can arrange that by uploading everything that a search engine discovers about a writer – anywhere – to a clickable list and then offering it to the writer with the instruction "Click what you want to have on your personal website." That selection will then be sent to his website automatically and with a little program like 'Publish or Perish' you can immediately assign the h-index to his oeuvre. You can then harvest all those personal websites for your university's repository. That way, they will be high-quality repositories. That's the kind of service that we need to progress towards. If SURF can arrange that in cooperation with the libraries, then you will have something worth communicating. Then there won't be any problem convincing scientists! It'll also make it very easy to produce annual reports and collect material so that institutes can carry out self-assessments.

There is also a “Don’t ask, don’t tell” movement. Scientists simply upload their articles to their own website and forget about what they may have signed.
They can’t actually remember what they’ve signed! They publish hundreds of articles with dozens of different publishers. They have no idea what they need to do with all those forms and statements, digital or on paper. They just chuck them away. Yes indeed: “Don’t ask, don’t tell”. Of course, I'm talking partly as an administrator here and I can hardly call for civil disobedience. In any case, I think it’s a good thing that a lot of journals permit personal use of articles. Journals are also gradually getting more liberal.

But that’s only the author’s version, the post-print.
That’s true. Well, I don’t like that at all. It means that you may get diversion of the citations to these post-prints. I simply want to get the published PDF version circulating. It differs, by the way, from one discipline to another. In computer science, there are all kinds of forums for scientific communication on the Internet, and people have abandoned all that messing around with journals. In physics, discussion takes place primarily via the preprints, and the actual article is merely for the dusty archives. They’re different subcultures. In my subculture, it’s only the actual article that counts and all the citations have to refer to that version. The disadvantage is of course that it takes longer before the publication comes out and then even longer before it’s in the open domain. But it’s not all so very terrible. Scientific results are generally not so urgent that something awful happens if you don’t get to see them until a few months later. Indeed, it can be dangerous to immediately jump on the bandwagon when the latest discovery comes along. Serious professionals in the healthcare sector actually never read primary publications. They work on the basis of guidelines, protocols, and systematic reviews. The worst thing that can happen to you is that you go to your doctor and he wants to try out what it says in an article that he read only the previous evening. You don’t know whether he is capable of assessing the methodology or whether he has weighed it up properly in the light of other relevant research. Journalists also play a role: they’re often too breathless and optimistic. That arouses false hopes. One of my academic role models always used to say: “There are no emergencies in academia.”

There are of course exceptions, for example swine flu recently or previously AIDS. Then haste really is necessary sometimes. Then you have journals that speed things up and promise to review and publish your paper within a week. In the case of a real epidemic, a network is immediately created with a closed discussion list where everything is made available. Later on, when the panic has died down, the articles are then sometimes written.

How can we promote the 'Golden Road', i.e. official Open Access publishing?
The painful bit is the transition. Up to the point when you can cancel subscriptions, there are double costs and the danger is that people will pass on the bill like a hot potato. In actual fact, you need to decide to make the transition en bloc. That way, you can cancel the subscriptions immediately. Sponsoring bodies are extremely important in that regard. The big financiers need to make the move. They can make Open Access publishing a condition for receiving a grant. Universities can do it too, but not on their own. The big ones, like Harvard, Oxford or Cambridge, can do that perhaps, but the others need to do it in combination. Ideally, they should do so in a European context, but the Netherlands could probably do it too. The Ministry of Education, Culture and Science could stipulate, for example, that everything financed with public money needs to be placed in the public domain on the day when it is published or shortly after.

And you mustn’t put any trace of that complex process on the plates of all those poor researchers. They mustn’t be troubled by it for a single moment. What needs to happen is that they simply submit their article and that, after a serious review process, it gets published. At some point later on there will be a bill, but they can submit that to their project, or to the institution when direct funding is concerned. That’s it...finished!

You can of course mobilise individual scientists as champions or opinion leaders. Harold Varmus is an example. There was also a case when an editorial board as a body forced their journal to allow Open Access.

So now we need to talk about the role of the publishers?
Publishers can continue to organise the peer review aspect within the new system. They generally do that effectively. They need to realise, however, that their product isn’t worth money any more but that they will be paid for the process. The leading journals will need to go along with this too. If they lag behind – The Lancet, for example – it will be difficult to enforce Open Access across the board. Scientists will want to keep publishing in the top journals, and if you were to forbid them to do so they would get very angry indeed. So those top journals will need to climb on board. But that’s quite possible. Even The Lancet has accepted the conditions set by the NIH (National Institutes of Health in the US, L.W.) and the Wellcome Trust (a medical charity in the UK, L.W.). A paradigm shift like this will put an end to the monopoly position of the publishers. For that reason alone, it will be cheaper; their margins will get smaller. The publishers don’t like that, which is understandable. By the way: the British Medical Journal has been making money just fine since it switched to Open Access.

What can patient associations do?
Patient associations too are increasingly focusing on the virtual information available on the Internet. The big problem there, though, is that there is no peer review. Anyone with a crazy theory can upload it to a website and if you have a nasty condition and you go looking for what’s been written about it, you are extremely vulnerable and credulous. Besides a great deal of good information, you find a lot of bad information. In fact, information for laymen ought to be subject to peer review too. Patient associations have started some initiatives in that regard, but that approach hasn’t really been carried through. They need money, but they increasingly get it from pharmaceutical companies, which is a potential threat to their impartiality. That is a serious danger.

Open Access is also important for them because it enables them to access high-quality articles. The trouble is that patients are often unable to decipher those articles. There needs to be an interface in between. The Cochrane Collaboration is a good example. It’s a worldwide organisation of medical literature researchers. They trace all the randomised studies of a particular clinical problem, for example “does ultrasound therapy help alleviate shoulder complaints?” They then carry out systematic reviews. I helped set up the organisation in 1993. It now has a portfolio of about 4000 systematic reviews and tens of thousands of people worldwide are working on reviews, quality control, etc. This means that when you go to see your doctor, he hasn’t just read one article about your complaint but a review systematically summarising 100 studies of the effect of a particular therapy. The reviews are in turn the foundation for guidelines and protocols. Wiley publishes the collection of systematic reviews. You have to pay, but it isn’t expensive: for doctors, it’s USD 60 a year and throughout Africa and in some Eastern European countries it’s Open Access.

Should Open Access also extend to research data?
For a start, the data files that form the basis for the tables in an article should be Open Access. That is important for reproducibility and also for the systematic reviews that we were talking about.

But then you’ve got the separate data files for large-scale surveys or cohort studies. They are always intellectually underused, no matter how many PhD students you’ve got. So other researchers definitely need to be able to make use of them. But not just by dumping the data somewhere where there is Open Access. Things become risky when you are no longer in control of what other people do with your data. Terrible mistakes can be made in analysing data, and you don’t want to be held responsible just because it was your data. So you need to be able to reach the necessary agreement. The safest way is for the person who uses the data to be required to bring in one of the authors of the original study as a co-author when writing his article.

You can also make predictions much more precisely if you can combine similar data files. Here too, you need to make proper arrangements, for example about setting up a consortium to oversee things. That will create enormous added value.

It’s notable that it’s in the area of medicine that the most decisive steps are being taken towards Open Access, for example by the NIH and the Wellcome Trust. Is there a body in the Netherlands with sufficient authority to take the same kind of step?
The Netherlands Organisation for Health Research and Development (ZonMW) – which is like the Netherlands Organisation for Scientific Research (NWO) but for medical sciences – could make Open Access a condition for providing grants. That would probably cause a fuss, but I believe it’s possible. Support from the NFU (Netherlands Federation of University Medical Centres, LW.) is important in that respect. The third major player is the PGO, the umbrella organisation for charities that raise funds from the public. Finally, there is also the pharmaceutical industry, although you need to be careful that it is not involved in deciding what can and cannot be published.

Interview by Leo Waaijers and Annemiek van der Kuil on 30 September 2009.
Photography by Annemiek van der Kuil