On the future of biomedical publishing…

Commercial STM publishing came of age at the end of the Second World War when governments around the world realised that investment in scientific and technological innovation was essential for national economic competitiveness and for providing effective healthcare. The subsequent massive growth in scientific activity could not be absorbed by scientific societies alone and for-profit journal publishing rapidly moved to fill the gap.

Today, the digital information age is driving a second revolution in STM publishing. The volume of information generated by science has mushroomed and its structure become more complex, there is a compelling need for data sharing and re-use, and traditional methods of quality control such as peer-review are broken. The question is, “Are the conventional STM players up to this new challenge?”

Journals are just one of the peaks in the information mountain range confronting today’s researchers. Two recent reports sponsored by the JISC and RIN have highlighted the immense amount of time scientists spend searching for and using information. Not surprisingly, the information gathering activity of a single researcher will take them to several different data resources during the course of a working day, requiring the user to learn the intricacies of several different interfaces and systems of nomenclature.

Taking biomedicine as an example, a plethora of “-omic” databases such as Ensembl, Genbank and Uniprot must be regularly consulted. Centres such as NCBI, EBI and KEGG began life as sources of data consensus for the biomedical community, but now house interactive models of entire genomes (see for example the UCSC Genome Browser), while companies such as Genego and Ingenuity publish simulations of networks of protein and gene interactions. The scientific article increasingly becomes, at best, a simple annotation linked to these virtual models of how biological systems function, or how clinical decisions are made.

The public’s perception of scientific progress is of a sequence of “breakthroughs” reported by “vanity” journals such as Nature, Science and The Lancet. But the truth is that, if it could be sketched on a sheet of paper, progress would often look more like a random movement than a perfect straight line. Indeed, it has been argued that the results of most published biomedical research are wrong or biased, that not publishing original datasets, as well as negative or confirmatory findings, makes it impossible to weigh evidence effectively, and that the peer review process contributes little to the assessment of the true value of a piece of research.

So my point is that STM publishing has moved from being part of the solution to making content available, to being part of the problem of making it useful and realising its full value for the tax payer. And herein lies the commercial opportunity. There is a huge amount of work needed to reintegrate these huge volumes of data and information being created by modern research technologies.

We have plenty of technological tools, but what we lack is a blueprint of sufficient scale to inform us how these tools should be used and how development resources should be directed. To achieve this scale of vision, the process must be led by the payers (NIH, Wellcome, UK research councils, HHMI, and so on). Open Access could have provided a beginning to this process by shifting the publishing business model away from a subscription basis and towards a service-based one. But instead it has lurched into the green and gold ruts leading to even more “balkanisation” of information resources. Unless this changes, the next decade is going to be fairly boring for STM publishers, as Springer’s Derk Haank has predicted.

Viewed in this light, Elsevier’s development of navigational tools such as Scopus, and support for UK institutional repositories, even its ill-fated attempt to move into the electronic lab notebook market, could be seen as strategic moves. But to engage with the vision of the payers, the successful publishers will need to focus on a market vertical, as have non-trad players such as Lockheed Martin, SAIC and IMS, and to take a much broader view of “content” than is currently the case. From this perspective, Open Access seems like a serious distraction. Payers please take heed, your guidance is needed…

What would you say to Elsevier about Open Access?

An excellent post from Cameron Neylon, who works at the Science and Technology Facilities Council, the UK’s major provider and supporter of large scale academic research facilities, including synchrotrons, neutron sources, and high powered lasers. He has been invited to speak at a forward planning exercise at Elsevier later this month. He says,

"My plan is to focus less on the arguments for making more research output Open Access and more on what happens as a greater proportion of those outputs become freely available, something that I see as increasingly inevitable. Where that proportion may finally be is anyone’s guess but it is going to be a much bigger proportion than it is now. What will authors and funders want and need from their publication infrastructure and what are the business opportunities that arise from those."

Focusing on the bigger picture is the way to go, I think, for commercial publishers. The easiest way to grow is to find new markets and to add new value - and its all there if you redefine the role of STM publishing in the way he suggests...

Google vs. Bing - at a split screen near you...

Blackdog has created a split screen view so that you can compare and contrast search results using Google and the new Microsoft search engine, Bing.

The great thing about Google is that it does have some semantic capabilities, so when searching using terms such as "Ganesha" and "Associates", Google "knows" that the second term means I am probably not after references to the Hindu deity named in the first term. Bing doesn't have this sensitivity - yet.

Likewise, searches using the terms "drug" and "information" came up better on Google - a better quality of results, with a UK bias. Still , early days...

Some thoughts on Open Access publishing

The JISC ‘Economic implications of alternative scholarly publishing models’ and the RIN ‘Activities, costs and funding flows in the scholarly communications system in the UK’ both set out to identify opportunities arising from the shifts towards online publishing and various open access business models.

Both reports could have made a better job at introducing the reader to their respective models . Both reports present a lot of data, but fail to deliver adequate executive summaries. As such it is impossible to reconcile the two descriptions completely without detailed study of the spreadsheet models.

Recommendation: JISC/RIN should together review the different modelling approaches taken and identify and agree basic common assumptions made.

As part of this accord it would be extremely helpful if the ‘dimensions’ for further investigations could be identified, e.g.

• Overheads. RIN study appears in include ‘research costs’. The JISC study includes library ICT overheads. Which overhead costs should be in and which out ?
• Overall scope. Are the JISC/RIN reports addressing identical library constituencies. Also RIN focuses mainly on journals, JISC on journals and books.
• There seems to be a net shift of costs from the developing to the developed nations under both models. It would be worth making these benefits more explicit.
• Disciplinary differences. The impact of these changes will be very different across the disciplines, due, for example, to the large differences in reading habits and the need for access to other linked sources of digital information. Compare for example economics, mathematics, biochemistry and oncology. It may be that the reports’ recommendations would differ significantly for different disciplines once this level of detail has been introduced.
• Both studies identify the high costs associated with searching for information, reading and writing. This extends the definition of a scholarly publisher from one who processes article content to one who is tasked with the development of digital tools that enable a variety of different data resources to be used effectively. The JISC report includes an assessment of ‘research performance savings’. For example, in biomedicine, some of these will result from investments such as PMCUK and its links with the many molecular databases at NCBI and EBI. How will this be modeled ? What, in the future, is the JISC/RIN vision for the respective roles of commercial and governmental publishing resources in this area ?

Recommendation: As a next step compare and contrast two disciplinary areas using a broader definition of information needs which includes other database resources.

Impact of open access on publishing costs. The JISC report concentrates mostly on distribution and production cost savings - hence it shows a big reduction in costs for shifts to online, open access and self-archiving. However, the sales and marketing focus will need to shift away from institutions to authors (see for example, PLoS One) and this will probably cause a net increase in these costs. Some of the production savings on the publisher side may shift to the authors as they are expected to take more responsibility for the formatting of articles prior to publication.

Recommendation: The opportunity offered by OA may not be cost savings per se, but the opportunity for the payers to redefine the role of the commercial STM publisher, at least in certain disciplines such as biomedicine.

Impact of open access on libraries. JISC report seems to predict a much smaller impact on library costs and expenses. This is probably because the RIN model has a more sensitive approach of fixed and variable costs, and JISC covers a much broader academic constituency and includes books. While the self-archiving option appears to create the largest savings, where are the costs associated with the indexing this material and making it findable?

Recommendation: One of the objectives of the disciplinary studies suggested above would be to define the future roles and skill profiles of libraries and librarians.

PubGet - a step forward ?

"Each year, scientists spend at least a quarter billion minutes searching for biomedical literature online. This is time they could better spend curing disease and building the future. Pubget's mission is to give them that time back." So the Mission Statement says...

But doesn't PubMed already have these features ? Less clearly signposted, perhaps, but researchers need fewer, better integrated interfaces if they are to become more efficient.

How the Semantic Web might improve cancer treatment

This PriceWaterhouseCoopers interview with Lynn Vogel describes M D Anderson's semantic technology research and development from the perspective of healthcare. It offers, I think, a fairly complete overview of the opportunities and obstacles for semantic technologies in this area.