Design, deployment and use of digital libraries in the NHS: implications for policy

Design, deployment and use of digital libraries in the NHS: implications for policy
By Ann Blanford, Professor of Human–Computer Interaction & Director, UCL
Published Tuesday, 29 May, 2007 - 14:00
http://www.egovmonitor.com/node/10681

Professor Blnford's framework suggests that ICT systems need to be flexible to adapt to evolutionary strategies in public policy, a stance this publication supports strongly.
There are currently great concerns over the introduction of information systems in the NHS containing patient data. Such systems pose great challenges over privacy, security and data integrity, as well as usability and the ways they cause working practices to change.
Research at UCL has been studying the introduction of non-personal information systems. Observations and interviews with over 200 participants (clinicians, management and support staff, and patients) have been conducted across eight NHS trusts. This work has focused mainly on digital libraries such as the National Library for Health, Medline and the Cochrane Library: systems that pose fewer technical and social challenges than those involving patient (or clinician) data. Although these systems might be considered non-threatening, the studies have identified both examples of good practice and projects where changes have been resisted, both passively and actively.
An example of passive resistance was that computers that were initially placed in wards in neutral space (accessed by all staff) were sometimes moved into doctors’ offices. The reason given for this was typically that of protecting patients’ privacy: that there might be unauthorised access to patient data on computers in ‘public’ spaces. However, nursing staff felt that this limited their access to information that they needed to support their work. This action polarised perceptions of technology, highlighting the differences between the various professional groups working within the space.
From detailed analysis of the data, it emerged that one of the key factors leading to this kind of conflict was the different value systems of people at different levels within the health system. For example:
• at the national level, technology is valued because it helps to improve standardisation of practices (e.g. across NHS Trusts) and information access for all;
• at the organisational level, technology can improve staff accountability and efficiency;
• at the team level or in the clinical encounter, it can facilitate effective communication; and
• at the individual level technology can empower users if it provides effective tools for supporting their working practices or managing health.
These values alone do not typically lead to resistance. However, other values relating to individual and team working do. In particular, individuals and teams have pride in their knowledge and ability to find creative solutions to immediate problems. This pride in expertise can be undermined by strict standardisation of practices. Similarly, information democratisation can change power relationships between clinicians and patients: some individuals can readily adapt to the changing access to information, but others find it difficult. Also, patients may have more time to research their conditions, but they typically do not have access to all the subscription-based resources available to clinicians; neither do they have the same vocabulary or understanding for searching and interpreting digital information. The imposition of new technologies that cannot be readily fitted with working practices results in surface compliance in which organisational structures and practices are subverted through personal interactions, such as the example given above.
Where the conditions are created in which technology and practices can adapt to better fit each other, technology acceptance and exploitation can create a positive environment for adopting new ways of working. Conditions for this include placing value on different kinds of knowledge and expertise, creating realistic expectations of the technology, making the technology clearly usable and useful, and giving people a sense of engagement with the design and use of technology and with the related social norms (e.g. concerning privacy and security behaviours). One example of this was an awareness server introduced in a hospital to protect potentially sensitive information: when the system had been inactive for a short time, a screen-saver displayed general hospital information; interviews with staff showed that this solution worked well because the awareness information only appeared when staff were engaged in low-intensity work (e.g. talking or making a cup of tea, rather than writing up clinical notes or finding information on-line).
What are the implications of these findings? The quality of patient care depends on buy-in from all levels and roles within the organisation. One example of this is the need to value the expertise of clinicians. Another is the requirement for systems that fit working priorities (not interrupting work inappropriately; supporting interpretation of information as well as access; etc.). Systems that do not fit result in active resistance or surface compliance.
In terms of policy, this work suggests various actions.
First, there are now many examples of effective system implementation and also many that have been less successful than anticipated. System effectiveness should be reviewed systematically and openly, and lessons should be learnt. If those lessons aren’t learnt, future advances will be much slower than desired. 
Second, bottom-up studies of actual practice should inform design – not to constrain what is possible, but to preserve the best of existing systems and develop a richer understanding of the context into which new systems have to fit.
Third, system commissioning and purchasing mechanisms should ensure that user requirements are considered from the outset and continually through design and deployment.
There is a creative tension between top-down (policy-led) and bottom-up (user-led) design; neither on its own provides a good solution. New systems have to be situated in existing values and practices in a way that supports evolutionary change