Will technology kill the healthcare library?

Abstract

Healthcare libraries are dying. Over the next 5 years we can expect to see emerging technologies destroy the need for those institutions, especially libraries, that today are principally responsible for the delivery of health information. New technologies are empowering the individual, whether it be in the laboratory, clinic or the community, to gather, access and assess information without the need to refer to experts for assistance. However, it is not just the accessibility of the information that spells doom for the library system—it is the whole philosophy of what constitutes a document and how it is published. Once again, technological progress is reducing society’s need for a group of experts who, for the last few centuries, have been the guardians and guides to information. This stark vision should not be allowed to go unchallenged. In order to create arguments to show why it is incorrect or to develop strategies in which librarians may have a future role it is necessary to understand where the technology might take us and how one can incorporate the expertise that librarians currently have with this technological revolution. There are marked differences between subjects and in the type of resource that will be affected most by these changes. The physical sciences have led the way in online publishing and many researchers only use electronic sources of information. 1,2 In these disciplines peer review is either left to the individual reader or is conducted dynamically online by appending comments to the published article. A similar system could be used in the biological sciences, but in general those working in these areas have a very focused expertise, with a specific jargon, that does not make it easy to evaluate the wide breadth of the subject. Individual researches tend to rely on expert peer review as part of the editorial process, but this may shift with time as younger scientists who are more used to assessing online information for themselves enter biological sciences. It is also important to distinguish between books and journals. In this article I focus on the impact of technology on journals, where it is currently most noticeable. To read a book on a computer is, in general, a ‘soul-destroying’ activity that detracts from all the pleasures of reading. 3 There are early signs as to how technology will change this, but new inventions such as electronic ink and paper will first have to become widespread. 4,5 There is little doubt that the World Wide Web and the Internet have been responsible for the profound changes over the last 5 years. This review gives a brief introduction to the technology and then examines how this may lead to the death of the document, the index and catalogue and ultimately the library itself. Web technology provides an easy, user–friendly interface to information distributed from Web servers scattered throughout the world. The ‘techno-babble’ that was once associated with the introduction of computer technology has disappeared and information distributed over the Internet is accessible through telephone lines, cable and satellite and through digital television channels. Information can be linked together irrespective of where it is held in the world and many of us already take for granted the ability to read a journal article online from a publisher based in the United Kingdom where the references are held at the National Library of Medicine in the United States. 6 Internet technologies make it easy for individuals or organizations to provide information from a single server and to allow users anywhere in the world to access it with standard, free browser technology (see Table 1). The resource itself can be enhanced with a full range of multimedia, and delivery systems are extremely robust having been tried and tested by many hundreds of millions of users around the world. Publishing information is now so easy that the classical role of the journal publisher is severely threatened and many of the traditional publishers fear for their revenue stream. Some journals such as the British Medical Journal have decided to make their entire contents freely available online at no cost and remarkably have not seen a fall in subscription rates so far. 6 An ambitious project by the National Library of Medicine, the E-biomed project, aims to make all primary published research output freely available using internet technologies. 7 Not surprisingly there has been considerable opposition to this proposal from the established core of publishing houses. It is worth speculating on how the financial resources released in healthcare libraries could be spent if they are no longer needed to pay for the bulk of biomedical journals. Until recently the publishing process has produced printed information in neatly packaged formats. Articles, books, newspapers and journals undergo a production process that results in the release of a final or near final version. The standard process is to classify the work in a number of ways, for example date, subject, author and keywords and issue a catalogue number, so that the resource may be retrieved by any of these routes. Librarians are the undisputed masters of these techniques. So far electronic publishing on the Internet has not changed this basic publishing process significantly, although components of it are beginning to blur. Documents no longer need to be made available from a single source (see below)—so the references or perhaps the illustrations may be part of another previously published document. This ‘borrowing’ is not new. Paper-based articles often contain cited or copied pieces of information from other works and, providing appropriate acknowledgements are made in the new article, it is a valid process and is the basis of scholarship. However, in the current paper-based documents a reproduced figure from elsewhere will never alter; anyone reading the article in the future will always be able to refer to the citation that the author intended. In the new electronic world of publishing this ‘borrowed’ resource may be updated, changed or even removed without reference to those who have cited it. How does this updating process affect all the key components of the secondary document (for example, the authorship, the publication date and the indexing)? A good analogy is the ‘darned sock’—at what point does a sock in which the original fabric is slowly replaced by darning become a new sock? However, advances in technology may well obscure this issue still further. In the next 5 years we may see much greater use made of technology that builds a document ‘on-the-fly’ from individual components that exist in underlying databases. Each part of the displayed document might be written by different authors at different times, but concatenated and stitched-together by computer algorithms depending upon the interests of the reader. Thus, an online newspaper may record what you have read (or at least browsed) in the past, register the length of time you looked at the articles, matched it against your general reading profile, etc., and then construct the text of an article based upon this ‘knowledge’ of your interests. Similarly, scientific papers may be written, stored and displayed in different ways depending upon either the knowledge of the reader or the scientific argument being presented (see the ‘dialectic tree’). Traditionally libraries have been the repositories of printed information; the librarian has provided the expert guidance and knowledge to find and catalogue the information and to help the users of this information. It was essential that each organization had their own locally stored collection of resources tailored to their own needs so that information could be retrieved efficiently. This stock of selected printed information was usually managed by an appropriate guardian or librarian. For journals this has now changed. Information is available in electronic format and can be distributed over networks. It can be stored in one central repository and accessed by millions around the world simultaneously. There is no longer a need for the information to be replicated and stored locally, indeed any organization that has only this collection will whither and die because they cannot have access to the most up-to-date information. As a historical archive of healthcare, these legacy institutions may have some value. However most of the information of value to the biomedical scientist is contemporary or only a few years old and these ‘historical’ documents might soon all be available in online format from central data stores. The healthcare library (but not those in the Arts and Humanities) with miles of shelves of bound journals dating back to the last century will become a mausoleum for the printed journal. Many librarians would happily say that it is not the distribution and storage of journals that really defines their skills; it is the ability to search and find information. This area is also under threat from emerging technologies. Traditional cataloguing and searching required each document to be carefully analysed and appropriately coded with key words, then filed in the correct place so that it could be easily retrieved. The librarian was the expert in systems used to code and catalogue printed resources and any non-librarian user who wished to find information would seek the expert advice of these trained professionals. However, the last few years have seen considerable advances in the process of document classification and retrieval using powerful search engines. These are far from perfect, but they are easy to use and they give the impression to the end-user that they have successfully retrieved a large numbers of documents. The riposte of the expert librarian to this observation is often that the information retrieved is of poor quality and that many resources are missed. This may not be evident to the individual user who may also be assisted through intelligent agents or ‘Web bots’. These intelligent search agents (for example Autonomy 8) are capable of learning from the previous searches conducted by an individual user and refining their strategies to greater levels of accuracy Individual search engines no longer require specific key words, but are capable of matching natural language requests to individual searches and creating a hierarchy of documents of relative importance for each individual. It is significant that the pressure for this evolution in document searching and retrieval has not come from librarians but from the users. The hundreds of millions of documents scattered over the Internet has been the driving force for the development of sophisticated search engines, which threaten to supersede traditional cataloguing systems. In the United Kingdom, the tardy evolution of the healthcare system has protected medical librarians in the NHS. Libraries are often under-resourced and the NHS framework in which they exist is so wary of emerging technologies that it is only the more adventurous librarians who have been able to develop innovative systems for access to online information. This is in contrast to the United States where the most successful medical libraries not only provide all the online access facilities for staff, but in some cases, for example the HighWire Press at Stanford University, have become world-wide publishers of electronic journals. I think that with only a few exceptions the healthcare libraries we see today with shelves filled by journals will disappear entirely and users will browse the Internet from their own computers for access to the information they require. The space currently occupied by journal stacks will be used for other purposes; the money spent on journal subscriptions will be redistributed or paid to a central body—perhaps representing universities or the NHS—who will negotiate licence fees for those resources that are not free. I believe that there may continue to be a role for the library and librarian even when users can access all the information that they want from their desktops and the search engines can provide apparently adequate retrieval of resources. Throughout this article I have looked at how technology will enhance the way information is distributed, searched and accessed and the effect that this will have on the information that libraries have traditionally distributed. Although most libraries will not be publishing international journals, all organizations, however small, have a need to distribute their own information—it is often easier to get electronic access to medline than the local telephone directory. It might not seem glamorous to be responsible for publishing the telephone directory, staff list or citation to in-house publications, but there is little doubt that this knowledge is of value to the organization and will probably be accessed more frequently. Cybrarian and Digitarian 2 are just two of the names that have been suggested for the librarian of the future. If the Cybrarian is going to evolve to lead this revolution they may not need to be competent at searching and finding information, but should be fully computer-literate and skilled in the use and application of these emerging technologies. While the views in this article are my own, I am particularly grateful to Peter Morgan and Wendy Roberts in the Medical Library at Cambridge University for their input and allowing me to share ideas with them. I am also indebted to many of the other Librarians throughout the NHS Eastern Region, who through their comments on the Local Implementation Strategy document for the NHS, have allowed me appreciate their visionary grasp of the future.