February 16, 2004

Interview Keeping open house at the BMJ

The British Medical Journal (BMJ) has pioneered the use of the Internet to provide free online access to medical research. For the last five years the BMJ's website has given free access to full-text versions of all original research articles, as well as editorial and news content. Open Access Now talked to Richard Smith about his experiences as the editor of the BMJ.

The BMJ is published by the BMJ Publishing Group, a fully owned subsidiary of the British Medical Association (BMA). "The unusual thing here is that as editor of the BMJ, I am also chief executive of the publishing group and I am on the same level as the secretary of the BMA," explains Smith. "I think that this unusual structure was actually quite important in terms of what we have been able to do. We are a very editorially driven outfit." All members of the BMA get a free subscription to the print version of the BMJ, and surveys suggest that most consider it an important benefit of membership.

Since he became the BMJ's editor, Smith has been an outspoken advocate of the benefits of online publishing. "Particularly in the context of the BMJ, I could see that it had a lot of advantages for us," recalls Smith. "First, we were less dependent on [direct] subscription income than some other medical journals." The journal has an income from classified advertising, from pharmaceutical advertising and from subscriptions. All BMA members get their print subscription included as part of their membership. Furthermore, the Internet offered the journal a chance to branch out from its traditional British readers and attract a larger audience. "It seemed like a very considerable opportunity to reach out to American readers."

"We built our first website in 1994 and began to make more and more material available, because it became clear that this was not really a threat to us," says Smith, who quickly realized that the BMJ could reach out to more people without losing advertising income. "In those early days there was not much anxiety about loss of subscriptions income. We began with a fairly limited amount of material, but we very rapidly began to put up everything. There was lots to be gained from making this stuff available for free and not a lot to be lost."

"We have a million readers that we never really had before"

Richard Smith

In 1997 the BMJ began a relationship with HighWire Press, a subsidiary of Stanford University Library that has played an important role in making journals available electronically. "From the word go we decided we would make the whole thing available and free," says Smith.

The BMJ's 'open access' experiment has been an undisputed success. "It has worked enormously well for us," says Smith proudly. "We have about a million visitors a month to our website. Many of them come from outside the UK and the vast majority of them never see a paper copy of the journal. So they are almost entirely 'new readers'. We hugely increased our reach - we have a million readers that we never really had before. We have a lot people from Britain coming to visit our website too, but most of them are also keen to get the paper copy of the journal."

In addition to selling the BMJ alone, the BMJ Publishing Group provides a product bundle that includes their twenty five other journals. "The other journals have been able to replace lost subscription income with electronic income, whereas we haven't because the BMJ has been available free," says Smith.

But Smith has been pushing hard for continued Open Access to original research articles. His editorial team has been debating with the Board whether they should continue to make research available free of charge and continue depositing it in PubMed Central. "I have argued very strongly that we should continue to do that and that we should experiment with an author-pays model," says Smith. "I think that in business terms there is a logic that says that people will pay for our value added material, which we have written or commissioned ourselves, but that the original research should be changed to an author-pays model." He wants to see a hybrid publishing model that incorporates both Open Access to research and subscription access to editorial and commentary. "But I haven't yet managed to convince our Board."

Smith is interested in the possibility of introducing market forces into scientific and medical publishing models. He envisages a system in which authors might be willing to pay for different types of editorial services that are offered by the publisher. "We might have different prices for quick revision decisions, more detailed referee feedback, preparation of different versions for print and online, or thorough technical editing. This potentially allows introducing a market into all these things that we do and enable us to get some sense of what people value and what they don't value."

"Three things seem necessary for resistance to be overcome and change to happen: a 'burning platform,' a vision of something better, and 'next steps'."

Richard Smith

The BMJ is currently conducting detailed research on what people think of author-pays models. "I am forever coming across people who have no idea what it's all about," says Smith. "But there is no real reason why the average scientist/researcher should understand the economics or the business model of publishing." He feels that the BMJ has a responsibility to explain to readers what the different publishing options are.

Smith is also keen to continue the BMJ's tradition of exploiting the strengths of both printed and online media. "Paper is portable, attractive, easy to read, and smells nice," says Smith. "People like the paper BMJ, whereas I think that readers aren't so bothered about having paper copies of specialist journals that contain only original research from beginning to end." About 30-50% of the BMJ is made up of original research articles, which are accompanied by editorial, news and essay sections. "It's much easier to read than many other science or medical journals," claims Smith.

"There is a tension for a journal like the BMJ," says Smith. "Although we are a scientific journal, the vast majority of our readers are not really scientists, they are ordinary doctors and clinicians and they don't really want to read that much original research. So we have a tension between readers who are less interested in original research and scientist readers who are very interested. In the old world, that was a tension we had to live with all the time. Now, in the electronic and paper world, we can please both audiences at once. We can give more data, links and information to electronic readers and shorter sweeter stuff for paper readers." The BMJ has developed a system it calls Electronic Long, Paper Short (ELPS) in which research articles are published in full online but the paper-printed version is substantially shortened.

Another way that the BMJ exploits the Internet is its 'Rapid Responses' section, in which letters to the editor are published online within 24 hours. Contributions to this online forum are screened only to exclude libel or breaches of patient confidentiality. Although this section has been criticised recently by AIDS experts who feel it has become a forum for 'AIDS deniers', Smith strongly defends the value of free speech. "Responses come from all over the world; from patients, doctors, lawyers and others. It is probably the place to debate a hot medical issue."

Smith says there are a larger number of driving factors that have influenced the development of electronic publishing and will contribute to the transition to Open Access publishing models. "First, what we do currently is rather poor. There is lots of mess and confusion, and people are overwhelmed with lots of dubious information." Then there is the vision of how much better things could be. "We have this thing at the moment referred to as the 'information paradox' - people are overwhelmed with information, much of it with not much relevance, not necessarily of high quality and not evidence-based. And yet during the average consultation between a doctor and a patient several questions arise that cannot be answered. Doctors have crucial information needs but they are crushed by the amount of information that they receive." Open Access could provide them with better access to the type of research information they do need, as well as contributing to the development of better tools for searching and indexing information.

"Another driver is what I call 'balkanization', meaning that everything is broken up all over the place. We are encouraging authors to do more systematic reviews, but it is such a hassle to find all this stuff and then it's very expensive to get access to it." Smith lists economic considerations as an important driver towards Open Access. "There's a lot of money to be saved in the system."

Smith concludes that all these factors will continue to accelerate the rate of change. He is confident that he can find the arguments to make a case for preserving Open Access at the BMJ. "We will convince the board to move to an author-pays model, by keeping it open and encouraging authors to pay."

bmj.com