The National Programme for IT (NPfIT) for health and social services in England has an anticipated cost of around £30 billion. The world's largest ever IT project aims to provide ‘Better information for health, where and when it's needed’. The core strategy is ‘to take greater central control over the specification, procurement, resource management, performance management and delivery of the information and IT agenda’.1 Its top priorities are listed in Box 1.1 Few would question the programme's high-level intentions. Virtually every general practice in the UK is now computerised. A rapidly increasing proportion of all practice team members, not just GPs, use computers face to face with patients every day. Arguably, UK general practice leads the way in the use of computers to support patient care. Yet, as evidenced by the medical tabloids, this key stakeholder group has become alienated and marginalised.
Box 1. National Programme for IT – top priorities.1
The explanation for this lies in part with ownership and control. The NHS struggles to throw off its image as a ‘command-economy state organisation’2 but NPfIT, which is run under firm central controls to very tight deadlines, perpetuates that image. Until recently, GPs owned their computer systems. Over more than 20 years these systems have become feature rich in response to user driven innovation. At many sites, electronic information systems and the administrative processes of running a practice have become highly interdependent. Suddenly, ownership has been taken away and procurement of all replacement systems placed in the hands of local service providers. These new people have little or no experience of the general practice domain. They are charged with providing NHS-wide integrated systems to deliver NPfIT priorities. The future of existing general practice systems, upon which GPs are increasingly dependent for delivering care and generating their income, remains unclear. There is little confidence in the quality of replacement systems, partly because what does not yet exist cannot be assessed and partly because there is a widespread perception that knowledge built up through many years of experience is not being harnessed. There is a fear that existing systems will be uprooted at short notice to be replaced with ‘new’ systems, resulting in severe disruption of vital practice processes. There are further fears that painstakingly collected clinical information will be lost or corrupted during this process, putting continuity of care and patient safety at risk.
In the 1980s Zuboff studied the sociological effects of introducing IT into Canadian pulp mills.3 She recognised three different phases: automating, ‘informating’ and transforming. Years ago, most UK general practices automated repetitive tasks, such as repeat prescribing and patient recalls, but systems evolved beyond this to become increasingly active and participative. They now deliver valuable, timely information about the workings of the practice. They ‘informate’ the practice; better ‘whole system’ familiarity leads to better decision making at practice level. However, this evolution has gone further still in about half of all UK practices where patient records are electronic rather than paper-based. Work processes have been ‘transformed’ as the computer and human parts of the system have grown into each other to become inseparable.
Zuboff observed the changing balance of skills, knowledge, and authority that occurred between workers and managers and the impact of different management styles. She made the point that the culture in ‘informated’ organisations tended to favour learning and innovation.3 Schein later went on to distinguish between the visions of ‘informating up’, which is about using IT ‘to aggregate and centralise as much information as possible about all parts of the organisation, in order to facilitate planning and control by top management’, and ‘informating down’, where system design ‘forces an analysis of the core production and other processes of the organisation’ so that ‘workers become familiar with the whole process and can thus make decisions that previously were made by various layers of management’.4 NPfIT is perceived to be ‘aggregation and centralisation’ writ large with the threat that ‘informating up’ will be dominant along with central control and micro-management. Poor communication with the professions has contributed to this perception.
In examining cases of IT implementation failure Schein characterised senior management cultures between the two extremes depicted by McGregor's Theory X and Theory Y.5 Theory X depicts a hierarchical authoritarian control orientation and Theory Y depicts a belief in collegial or participative relationships that permit high degrees of self-control. A Theory X management orientation has very negative implications for the success of ‘informating up’ or ‘down’ and ‘transformation’ is simply not feasible. In contrast, a Theory Y orientation has positive implications for the success of ‘informating up’ and very positive implications for ‘informating down’ and for ‘transformation’.4 NPfIT has until recently shown scant enthusiasm for participative relationships. It displays a deterministic approach that sees IT as an external force for change and tends to view those with dissident views as being passive, resistant, or dysfunctional. A more participative sociotechnical approach6 such as that loosely described as actor network theory7 is long overdue.
We live in a networked world where increasingly large amounts of data can be exchanged almost instantly regardless of distance and time. Flows require three elements. There must be at least two ‘nodes’, a ‘commodity’ that they wish to ‘trade’, and a transport medium that is fit for purpose. Exchange is most effective and meaningful where the nodes involved share cultural and organisational context. The most successful will develop many links and the more important the flows the greater the tendency for the nodes involved to become culturally closer.8 A modern general practice would represent a very busy node, typically transacting most of its business directly with other nodes close to home. This is at variance with the NPfIT paradigm of a remote central record that takes no account of the importance of culture, context, and proximity.
The further medical information travels away from its origin, the more difficult it becomes to make reliable sense of it.9 First, the completeness and accuracy of the information at its origin should not be taken for granted.10-12 Second, to communicate effectively we need information with its all-important context13 and not just ‘data’. Some of that context is cultural and not captured in the electronic record at all,14 and some may be just the kind of information that patients will be reluctant to share with us if they are worried that it is destined to become widely accessible.15 Reliable communication of meaning over distance is therefore problematic, with further implications for patient safety and confidence.
Respect for patient autonomy and confidentiality underpin patient–doctor trust. However, in a networked society it is becoming increasingly difficult to maintain information privacy without making information inaccessible. Rather than continue to treat this as an issue for individual patients and professionals, it may be time to consider new models that depend on tight codes of accepted behaviour applied to whole organisations and underpinned by surveillance. Health care is increasingly provided by teams of professionals where individuals may have multiple and changing roles, in some of which they may have a legitimate need to access personal information about a patient. In response to this major challenge, NPfIT is instituting role-based access controls coupled with the principle of ‘legitimate relationship’. The details of every incidence of access of a patient's record will be captured in an audit trail and there will be tough surveillance procedures to check for abuses.
Many of the concerns expressed in this article arise because the people, organisations and technology that deliver health care together make up an unpredictable complex adaptive system.16 Thus far, NPfIT seems to have adopted a rational and deterministic approach to management. It systematically gathered and analysed facts to produce an output-based specification and then set clear objectives with tight deadlines.17 This ‘well-oiled machine’ is now driving IT into the health system. That may be fine to get the technology in place, but much more than just IT is required.18 The impact on patients and professionals has yet to be seriously addressed.12,19 A very different approach is needed to nurture culture change. We will need to feel trusted, to be encouraged to experiment in a system that encourages innovation and learning from mistakes. With powerful ‘informating’ systems, we should be well equipped to adapt quickly to change20 and be able to transform the way we work to provide truly patient-centred care. The £30 billion question is not just whether NPfIT will get the technology right but whether it can also win the hearts and minds of the people on whom the NHS depends every day.