Challenges for the next 25 years

Library 1.0

Every day in the UK, 819 000 patients have a consultation at a general practice, 122 500 patients attend an outpatient clinic, and about 60 000 patients are treated as day patients or inpatients in hospital.¹ It is estimated that a million clinical decisions are taken in a day in these settings.

High quality, safe healthcare requires health-care professionals to access knowledge and the best available evidence efficiently and, increasingly, at the point of decision making.

If we describe the current health library service—Library 1.0—as a passive support to decision making, providing physical and digital knowledge services to users away from the clinic, then in 25 years the business case for the health library will depend much more on providing a more active support to healthcare, not just providing knowledge but participating in its use in decision making. This will increasingly be delivered not only in active clinical decision support systems but also in the clinical and health management context, by which we mean the work of developing care pathways and innovation of service delivery through networks and management institutions.

Library 2.0: knowledge as a web service

There was a period when the only access point for libraries was the front door of the library. Now there are numerous digital resources accessible and, to a limited extent, e-mail supports the virtual reference function. That trend will only continue in the future and libraries will have to enhance and develop their digital offering to reach out to their users through the web.

In the future the search box will be on the toolbar of the student's browser. The commissioner will access the ‘evidence button’ on his/her ‘Service Modelling Tool’. The clinician will see ‘evidence recommendations’ in their pathology results screens. Document delivery will be digital end-to-end—no one will print because printers will be removed to ‘green’ the organization. Users of Google, if they have not switched to Baidu <http://www.baidu.com/>, will personalize it to be their search engine of choice for finding NHS and scholarly content.

Libraries will become a ‘content community’—like Flickr <http://www.flickr.com/>—where users will tag, rate and group knowledge and share knowledge items among peers and clinical teams. They will be supported by ‘Team Knowledge Officers’ within their digital space using collaboration and conferencing tools like Skype and they will seek support from local librarians within that same digital space.

Interoperability with the future

Obtaining the above facilities requires a radical transformation of the way health libraries procure and package knowledge and knowledge services. Increasingly, there will not be any one predominant website or collaboration space, and these will change every 5 years, therefore digital services will have to be capable of being surfaced in more than one place.

The key to keeping digital health library services visible in whatever the future digital space of choice is, will be the application of Service Oriented Architecture (SOA) principles and ruthless pursuit of common standards, in all layers of the application stack. In an SOA world, each discrete piece of functionality in a knowledge service or OPAC say, is separated out and provided with its own machine-to-machine programming interface. When combined with standards for metadata, vocabularies and messaging you create a powerful set of services that can inter-operate and be re-used in multiple implementations. In an SOA world, data, and the systems that process that data, are separated. The user can find and search collections of ‘Evidence-based Guidance’ or ‘Systematic Reviews of evidence’—not portals like OVID PS or SCOPUS containing only what that aggregator has pulled together.

Library services can use this architectural approach to provide their own programming interfaces to the collections and services they create from their own knowledge generation or to content they procure from suppliers. In this way knowledge can flow through to NHS clinical applications or common desktop applications used by health-care staff.

A more equal partnership with publishers

The Web is a highly competitive environment in which library service offerings must compete for users with dynamic and agile commercial providers. The users will vote with their mouse if they are not satisfied. This means that health libraries must co-ordinate and act strategically in developing digital services and must adopt user-centred design methodologies to develop service offers which users view as indispensable to their work.

A recently commissioned report for the JISC Strategic Content Alliance concluded that: ‘While much attention is given to making material available online, very little is given to making sure people become aware of it and can find it. We find few digital resource projects have devoted substantial financial or intellectual resources to understanding user needs, preferences and behaviours’.^2,3

Content publishers have more sustainable models than public sector institutions, are more powerful than public institutions. We must continue to breakdown the walled gardens of publishers’ offerings so that we address public sector objectives: open access to knowledge.

This will only happen if we seek to partner with content providers rather than out-compete them. Nevertheless, such partnerships will only deliver if we can redress the imbalance of power between producers of knowledge and publishers of that knowledge.

Knowledge re-engineering

To complete this picture we also need to address the engineering of knowledge from the outset. For example, we need to transform NICE guidance from monolithic PDF documents into discrete pieces of knowledge such as recommendations and evidence, to code it as XML which can be transformed by other applications, attach persistent and reliable identifiers (Digital Object Identifiers). This gives rise to the concept of knowledge as an object, which can be separated from the journal article that contains it, transformed in a format which can be re-purposed, given labels to code and identify it and pushed through clinical applications to surface on a patient record.

NICE guidance is given only as an example here, the same rules could be applied to any knowledge produced by the NHS or procured for the NHS.

Conclusion

Within this 1000 word-limit, the reader will recognize that much detail has been omitted, however, this merely serves to focus on the critical path:

Conflicts of interest

IM has declared no conflicts.