This paper takes forward some initial research on the relationship between attitudes and technology in the context of leading edge administrative applications within the UK National Health Service (NHS).

'The NHS was set up 50 years ago and is now the largest organisation in Europe'[7] [Perry7] . For many years, it has had to contend with IT plans which have been frequently changed, sometimes for political and sometimes for organisational reasons.

The NHS is now committed to developing both Electronic Health Records, or EHRs (patient centred records which show summary data over a lifetime), and Electronic Patient Records (EPR) - treatment centred records which may both read and write data to / from the EHR. The EPR systems are expected to integrate data from existing Patient Administration systems, laboratory systems, and so on. A major challenge for EPR systems has always been the disparate nature of the systems providing them with data.

EPR systems were selected for study as they are still unfolding and there is significant use of emerging technologies, particularly XML. Further, although EPR systems are primarily concerned with records, the records themselves constitute the content of a number of organisational processes, particularly in the area of patient care. These two aspects of technology deployment - emerging technologies, which address processes, rather than transactions - were of special interest. The EPR research follows work that has been done on the organisational impact of other emergent technologies like workflow systems (Doherty and Perry - [1] [Perry1] ). Here, a number of banking and insurance organisations had been surveyed to assess the organisational and cultural impact of deploying workflow management systems (WFMS). The research showed that business processes had not only been automated, but had been changed substantially, because the technology had given managers the scope to re-think the business flows in their departments. As a result there had been a number of knock-on effects, including a degree of Business Process Re-engineering (BPR).

A later paper by Doherty and Perry [6] [Perry6] examined the effects of WFMSs on organisational culture. It found that there were significant, though frequently unplanned, changes in organisational culture following the deployment of WFMSs which had changed the business processes and with that, had apparently changed the way people related to their work and to each other.

This study set out to assess the motivations behind using XML in EPR systems, and the effects on the organisation afterwards. It considers the possible effect on organisational culture, and the implications for service delivery.

A small group of NHS Hospital Trusts ^[1] was invited to respond to a telephone interview, which used a structured questionnaire as its base. Intentionally, the study was too small to be statistically significant across the NHS; however its purpose was to provide a qualitative measure of attitudes, and to decide whether further study was warranted. The areas covered in the interview were; ^[1]

1. Project management and project personnel,

2. The procurement process for the system,

3. Methodologies used by the Trust in deploying a solution,

4. Technologies used in building the system,

5. The motivation for implementing the system,

6. The relationship with organisational culture,

7. The overall effectiveness of the solution.

All but one of the respondents were senior Information Management and Technology (IMT) managers, and that one had IT experience. Equally, there was more than one ex-clinician in the sample.

Respondents were asked to describe the stakeholders who were involved in the system design and deployment. They were asked about the involvement of managers and staff from clinical, administrative, nursing, laboratory, and IT areas. Additionally, they were asked to comment on external contractors - a total of 11 categories. All had involved at least 9 of those categories in the system.

All respondents had a senior management sponsor at both the beginning, and at the end of the system implementation.

The significance of this breadth of stakeholder involvement, and the presence of sponsors, is well documented in the project management literature [3] [Perry3] , but is also relevant to the discussion, later in this paper, of the effects of the systems on organisational culture.

Of course, understanding the relationships - administrative to clinical, clinical to patient, nursing staff to laboratory staff - involves understanding the culture as well. For the purposes of this study, we used the following definition of culture: 'the set of important assumptions and beliefs (often unstated) which members of an organisation share in common'. [5] [Perry5] Respondents were then asked to comment on how closely their own organisational culture fitted with one of four types, based on a Handy and Harrison's ^[2] model, cited in Yeates and Cadle [3] [Perry3] . Within this survey, the model helped as a discussion point, rather than as a framework. Every respondent commented that there was more than one type of culture in his or her Trust, depending on whether one took a Trust-wide, divisional or departmental view. Centrally and historically, culture was seen to be power based, in the hands of a few individuals. ^[2]

To some extent, this may explain the presence of senior management sponsorship and the wide inclusion of stakeholders, in an attempt to get the holders of power 'onside'.

Respondents were asked to prioritise their top 5 motivations for introducing EPR systems. They were offered a list of 13 motivations, together with an option to selectg their own.The motivations, or business drivers, showed some marked differences between these NHS respondents and the earlier study on workflow management systems in financial services as described by Perry and Doherty [1] [Perry1] . The dominant drivers in the financial sector, beginning with the most significant, had been to improve customer service, to increase productivity, and to reduce costs. While a service-oriented motivation was the principal driver (at 77% of respondents) in that study, the finance- or administrative-based drivers of productivity and costs at 57% and 45% respectively, were significant. The NHS respondents however, had a clearer division between the two types of driver. (See Table 1). Overwhelmingly, the motivation for introducing these systems was to improve patient service. This was endorsed by everyone, and rated first among the 'top five' by all except one respondent. The next most popular factor, by both numbers and degree, was 'to empower clinical staff'. Of the finance / administrative-based choices, only 'to increase productivity' had as many respondents, but at a lower level of importance.

It may not seem surprising that organisations in a 'caring profession' should put service, or relationship-based motivations at the top of their lists. However, these are also organisations, which are under pressure to meet government targets and to deliver within strict budget guidelines. More surprising then, is the comparatively lower rating given to the finance / administration-based motivations.

Several of the respondents commented that they had found ways of circumventing the accepted NHS rules of cost/benefit models, authorisations and so on, in order to deliver either a pilot or a fully developed system. They made the point that the existing procurement structures and rules do not appear to satisfy either the expectations of the stakeholders generally nor the needs of the end users. Prima facie, there is a need for a more goal-centred way of looking at procurement when the motivations are based around organisational goals and aspirations (i.e. patient care, empowering staff).

McBride [4] [Perry4] describes the traditional approach to investment analysis of IT systems - a focus on cost savings, reduced headcount, redeployment, increased productivity and so on. Once the savings have been made, he argues, the search then moves on to the next area. He proposes a relationship-based approach, which will assess the accumulating benefits to those using the system. Such an approach needs to identify the stakeholders in the system, and to understand the relationships which are in some way mediated by the IS. It should then possible to assign a value which represents the additional benefits brought by the IS. Aggregating these values brings an overall organisational value, and may point the way towards new benefits and added value to be gained.

Whether as a result of the influence on technology suppliers, a recognition of the lack of in-house resources, or as part of the general move towards using software packages, all respondents had implemented systems using customised packages. None had built, or commissioned the build, of a system 'from scratch'. There was widespread use of a 'toolkit' approach underlying the packages - Java, ActiveX, ASP, VB and SQL were all represented. The only common thread was that all the systems were using, or shortly to implement, XML.

There were a number of reasons given for the use of XML. Certainly a substantial influence was its endorsement by government; "There is a strategic decision to adopt XML and XSL as the core standards for data integration and presentation" [10] [Perry10] . This appears to have created a 'politically' powerful reason to select XML. However, no respondent gave government approval as the major reason for using XML. Instead, the most definitive comments about the selection of XML as a technology were made about two particular features; its ease of deployment and its cost effectiveness. Firstly, XML is seen as being relatively easy to learn and therefore XML based systems could be modified in-house, at a lower cost and with the benefit of internal expertise. Secondly, it is seen as being cost-effective even when systems are initially developed by outside contractors (which each of the respondents were using initially). This financial argument appears to be strengthened because the XML based EPR systems are being developed within an already cost effective (i.e. intranet / internet) model.

Unpublished research by Graphnet [8] [Perry8] a UK based XML solutions provider, found that of 300 clinicians questioned about their requirements for patient information, almost half wanted summary information, almost half wanted the full record, and one wanted information 'beyond cradle-to-grave'. By using XML, they have been able to offer all the clinicians exactly what they wanted. For most routine consulting, clinicians simply need to check summary histories and results; when complications arise, they may need to search all the patient's records very thoroughly. For each type of requirement, XML provides a fast, and more importantly, an intelligible tool for finding information. Graphnet report that many Trusts, which use their systems, and many users within those Trusts, prefer their information to be displayed in a way, which is meaningful to them. Again, by using XML stylesheets, they have been able to help clinicians use information in ways they can understand quickly and easily. Further, by offering such ease of use, Graphnet report that Trusts are beginning to re-develop or streamline some of their working practices.

When we come to assess the effects on organisational culture, it is worth considering some key characteristics which make XML-based EPR projects unusual: there is this inclusive stakeholder approach, the cross-departmental nature of EPR systems, and this new technology (XML) which can radically alter process rules and the presentation of content. To what extent had the EPR systems affected organisational culture, either at the Trust or departmental level? There was no real evidence for any change either way at the Trust level, but anecdotally, evidence began to emerge of change at departmental level - i.e. among the people who actually used the system. Several respondents offered observations which suggested that the systems' implementation had had a more radical effect than simply a more efficient system. "This is the first time I had seen strong clinical support for [an Information] system".

"We have better teamworking and have re-engineered some processes".

A third respondent said that the project team had explicitly used the introduction of the system to change organisational culture, and that it had done so.

The workflow management systems studied by Doherty and Perry [1] [Perry1] , were shown to be qualitatively different from most transaction-based systems in that they are fundamentally concerned with improving the process of taking up information, either by re-engineering the 'business' rules or by presenting content in new and more assimilable ways. XML based EPR systems have less emphasis on traditional workflow functions, like routing and tracking work items, but the processes of making the appropriate information available to the right people at the right time, are entirely comparable. This series of interviews certainly indicates that they share a number of key characteristics:

* The ability - sometimes the necessity, of being cross-functional within the host organisation.

* The focus on process, not transaction.

* The adaptability to different processes.

Going back to McBride's work [4] [Perry4] on a relationship-based approach to IT investment appraisal, we can see that (sometimes by default) the introduction of new technology has the ability radically to change our take-up of information, it is possible to alter existing work / professional relationships or indeed create new ones. This is a view reached, through quite separate lines of enquiry by both McBride [4] [Perry4] , and Doherty and Perry [6] [Perry6] .

McBride further takes the view that, if this is the case, then the possibilities of creating new 'added-value' also increase (my emphasis).

In Doherty and Perry's recent paper [6] [Perry6] , they consider specifically the impact of workflow management systems on organisational culture. While the terminology used in the finance industry is very different, the results nevertheless show some strong similarities with the effects of the deployment of XML based EPR systems. As a result, there are also opportunities for those involved in the implementation of EPR systems.

Doherty and Perry found that a new technology (WFMS), with similar characteristics to EPR systems, "has the ability to make significant modifications to the culture of an organisation. Furthermore, there is no evidence to suggest any user resistance, let alone system rejection, as a consequence of IT-induced cultural change. On the contrary, in most instances the systems have delivered tangible benefits to the organisation, and have been positively embraced by the users".

While it is too early to say, from a pilot study, that XML-based EPR is exactly comparable to the introduction of WFMSs in its organisational effects, the findings above are remarkably similar to the comments from respondents in this study (quoted at the beginning of this section).

'Improving customer service' or 'improving patient care' have been described as the major motivations behind the introduction of workflow systems in the financial service sector and behind EPR systems in the NHS. Yet 'improving service' quality has been described as requiring 'an appropriate organisational culture' [8] [Perry8] . Therefore, to use technologies that are themselves reported as having a positive effect on organisational culture would seem to be giving a direct boost, or a head start, to a service initiative. If workflow and XML based technologies can already help, intuitively, to change the way people work, can they be used in a proactive way as well? In other words, can the possible effects of cultural change and attitude change be planned for? And what results, in terms of service, efficiency, and productivity, may be predicted?

Using the ideas raised in this research, it should now be possible, at project inception phase, to conduct a 'cultural audit' to provide a baseline from which to consciously establish a culture which is able to deliver the highest standards of service. The process by which that culture is developed will use the possibilities inherent in the technology, by deliberately involving stakeholders in the understanding and realisation of the system. With a knowledge of the prevailing culture, this would involve, in the case of the NHS, building on what is already an extensive pattern of stakeholder involvement, in order to achieve specific levels of empowerment and co-operation. It would also target specific areas of frustration.

Not only will such an approach benefit service outcomes; future parameters for procurement could be set, as a way of quantifying the so-called 'soft dollar' problem. McBride's [4] [Perry4] work demonstrates how organisations rely on historical cost-benefit models, which show the easily quantifiable savings or gains. By understanding and unlocking the power of cultural change inherent in XML-based systems, it should also be possible to track and predict the relationship-based, cultural outcomes. To date these have been excluded from traditional cost-benefit analyses, even though we know through experience that often they cause the increased productivity, or efficiency, or customer take up.

Don't:

* Assume that XML is just about extracting better information

* Underestimate the power of XML to raise the sights of your employees

* Simply build a slicker or quicker version of what you have now.

Do:

1. Look for the opportunity to make a 'paradigm shift' in the service you are delivering

2. Find out, in a structured way, what your employees really want to achieve for their customers or patients

3. Take the time to understand what your organisation's culture is really like, and why it holds people back

4. Measure 2 and 3; if you're having difficulty, contact us at the Centre

5. Measure the effects, then build them into your next cost-benefit analysis