Feature

posted 1 Jan 2001 in Volume 4 Issue 4

The healing power of KM

Effective clinical decision-making relies on providing healthcare staff with integrated access to the most reliable information sources. Paula Kingston, Lynne Cooper, Dr Elizabeth Hughes, Dr Andrea Mayne and Jooli Atkins (left to right) look at how Sandwell Healthcare NHS Trust has developed and evaluated knowledge management processes to support best practice in patient care.

Quality assurance in clinical decision-making relies on access to up-to-date, authoritative clinical and management information, both locally produced (within the Trust, Health Authority and so on), for example the Path.Finder system, WardLinks, and so on, and that which is available externally. A wealth of high quality external information is now available on the Internet via sites such as Doctors.net.uk and via various databases, such as CINAHL, Medline, Cochrane, Best Evidence and HMIC, but is only accessible to those who have the skills to effectively search for and retrieve this information.

A need to provide integrated access to these information sources was identified within Sandwell Healthcare NHS Trust. There was also a need to support healthcare staff in developing the skills, understanding and approaches that underpin the development of a knowledge management culture, thus helping to embed knowledge management processes into working practices.

A proposal to develop a knowledge management framework was formulated last year and following a successful bid to the Black Country Training and Education Consortium by the Trust, funding was secured for a five-month pilot project.

The project aimed to develop and evaluate knowledge management processes to support best practice in patient care. It looked widely at knowledge flows across the Trust and within a local GP practice. The project involved working in close liaison with Marjorie Hammond, the WISH trainer for Sandwell and Dudley.

The areas covered by the project included:

The knowledge management infrastructure - the information systems that support the management and transfer of information

The policies, processes and procedures that support and enable the sharing of information, and its conversion into knowledge

The culture, attitudes and working practices of the staff involved.

Five main objectives for the project were identified. These were:

To develop a knowledge management framework that aligns information systems, policies, practices and procedures, and staff development to support knowledge management approaches in clinical decision-making

To extend the local knowledge base currently accessible through the Path.Finder system and incorporate access to external information sources

To identify and work with multidisciplinary staff groups to pilot the framework and develop processes and procedures for knowledge management

To identify training and support needs and develop a programme to meet them, in collaboration with the WISH project, as appropriate

To evaluate the contribution of knowledge management processes to improved clinical decision-making through in-depth interviews, case studies, focus groups and questionnaires.

This report aims to show how the project set out to achieve its objectives and will focus on:

The methodology employed

The project analysis - a quantitative and qualitative analysis of training

The feedback - a critical look at the issues raised (Pugh Matrix)

Conclusions and recommendations - taking knowledge management forward within Sandwell. Methodology

Following discussion and review of a number of methods of identifying key issues across the trust, such as trust-wide surveys and in-depth questionnaires, focus groups and targeted questionnaires were employed as the principal means of acquiring the information and, subsequently, of evaluating the success of the project. The focus group methodology was chosen as it is a tried and tested technique for gaining an accurate overview of issues and needs across an organisation, using a well-selected representative group of staff.

Initial meetings were held with 13 key personnel from across the Trust to discover their views on knowledge management and what they hoped the project would achieve. Staff from both the Acute and Community areas of the Trust (18 in all) attended one of two focus groups. Human resources, training and development, or the director of nursing nominated most participants. A third focus group, comprising six members of staff, ranging from a doctor through nursing support to clerical staff, was held within a local GP practice.

It was agreed that the project would concentrate on three pilot areas, rather than attempting to run across the whole trust in the limited timespan of five months. Initially, this comprised all focus group participants, three hospital wards (one medical, one surgical and one general elderly care) and a GP practice. At the same time, general basic IT skills training, incorporating Path.Finder, the Trust’s knowledge database, would be offered as widely as possible across the hospital but would be restricted only to the pilot GP practice. The project pilot areas had the full co-operation of the management of the ward or practice.

Training needs

Training need was established by use of two training needs analysis questionnaires - one to identify basic IT training needs and the second to identify more detailed database requirements. Training began with the IT Basic Skills and Path.Finder sessions. This training had the following main features:

No more than two hours long

Delivered on site (GP practice/ward) or in training room

Delivered to suit flexible times (7 am - 10 .30 pm).

This training was followed by further sessions on WISH (see introduction) and/or other Internet databases, depending on the identified need.

Feedback

Quantitative feedback was gathered through pre- and post-training questionnaires assessing the skills and needs originally considered at the start of the project. Qualitative feedback was acquired through focus groups, where anyone who either attended an original focus group or who had been involved in any of the training courses was invited to attend one-to-one interviews, where more detailed discussions could be held.

Two open focus groups were scheduled in the Acute trust, with a further focus group being held on a pilot ward. In addition, a focus group was held in the GP practice and six one-to-one interviews were carried out.

Project analysis

Quantitative analysis - general IT skills

Forty-six pre-training questionnaires were completed to ascertain individual knowledge of IT in general and Path.Finder in particular. This was based on marking a number of statements in relation to existing knowledge. The maximum total mark available was 55. The mean average total was 25.5.

Only nine of those people completed the relevant post-training questionnaires. Once again, the maximum total mark available was 55. The mean average total was 36.3. Statistically, using a chi-squared analysis, these results are significant at a critical value of .001, which means that the probability of the increase in ability being purely by chance is less than 0.1 per cent.

In addition to the nine post-training questionnaires received, a further 24 questionnaires were returned where no pre-training questionnaire had been received. The mean average total there was 39.7.

Statistical analysis, once again using chi-squared, shows that for these results to be significant, each person’s pre-training questionnaire would have to have shown a total of less than 39. A result of 39 out of 55 equates to 70 per cent of the maximum pre-training knowledge base and, as that was our initial target, it is reasonable to assume, bearing in mind our experience of the IT skill base in the Trust, that the majority were not at that level prior to training.

Quantitative analysis - database skills

Nine pre-training questionnaires were completed to ascertain individual knowledge of specified databases, such as Cinahl, Medline and Cochrane.

Once again, this was based on marking a number of statements in relation to existing knowledge. The maximum total mark available was 40. The pre-training mean average total was 15.8. The post-training mean average total was 24.0.

Only three people completed relevant post-training questionnaires. This provided less feedback than the general IT skills and, therefore, fewer results to analyse. The results received, however, are significant to a critical value of .05, which equates to a probability of less than 5 per cent chance of the increase not being as a result of training.

Qualitative analysis - pre-training analysis

This was carried out through the use of focus groups, as detailed above. A number of issues were identified as a result of these focus groups:

Access to training and development for clinical staff is very good, but less accessible for staff in support departments - more flexible approaches were needed

Need for improved internal customer service from support departments

Continued emphasis needed on increasing skills for evidence-based practice

Need for improvement in IT support services

Increased awareness of Trust goals and identification of mechanisms to encourage ideas from staff

Need to ensure local information, such as Path.Finder is accurate and regularly updated.

These issues were summarised under the following main areas of interest for the project:

Culture - within which the knowledge base is held

Content - of the knowledge base

Technology - supporting the knowledge base.

Qualitative analysis - post-training analysis

The issues arising from the focus groups and the subsequent qualitative analysis were used as the basis for structured feedback sessions, of which two general and one ward-based session were carried out, and one-to-one interviews, of which four were carried out.

Feedback

Analysis of the feedback qualitative data is provided by the use of a Pugh Matrix. Designed by Derek Pugh1 in 1983, the Pugh Matrix was intended as an organisational development tool for assisting when deciding on intervention strategies in complex change situations.

It has been developed for use at Sandwell Healthcare as a matrix for detailing the issues arising from feedback from the knowledge management project.

Three main areas were identified at Sandwell:

Communication

Evidence-based practice

IT infrastructure2.

These areas have been broken down into three matrices each detailing issues arising from feedback at the following levels:

Organisational - the Trust as a whole

Inter-Group - the way in which groups within the Trust, including Acute and Community, inter-relate

Group - the impact of issues relating to groups within the Trust, both Acute and Community

Individual - the impact of issues at an individual level.

Issues are considered in terms of:

Behaviour - what behaviour is being displayed, observed or related

Structure - what is the system within which the behaviour is being displayed

Context - in what setting is the structure based.

It is important, therefore, to view the matrix at all levels and in terms of all three dimensions.

Project constraints

Although the project as a whole was considered by participants to be successful, a number of constraints were identified. These occurred in two areas:

Training

Due to the time constraints on the project, it was necessary to find a GP practice that was willing to take part in the project and had the IT infrastructure to support the training programme. Although Neptune Health Park was initially proposed for the project,

both time pressure and IT considerations prevented it from being a viable option. A member of the Path.Finder support team recommended another, smaller practice whose infrastructure was suitable for the project’s requirements.

Although having full agreement of the ward managers within the hospital and being delivered flexibly and in line with the needs of the organisation, as well as individuals, the timing of the training delivery, coinciding with the winter pressure, caused problems for release of staff. The surgical and general elderly care wards had fewer problems than the medical ward, which found it so difficult to release staff for even short periods of time, the ward was taken out of the project. At the same time, however, another medical ward originally only involved in the additional, ad hoc training of basic skills and Path.Finder, proved so enthusiastic about the project as a whole that they agreed to take over the place of the original ward.

Although the training offered proved effective in terms of increasing skill levels, a number of areas were unable to access the databases trained, for a variety of reasons:

No access to Internet from ward-based PCs

Access to Path.Finder not available on some PCs due to:

- Need for hardware upgrade

- Lack of software installation

Lack of IT support.

It was noted that the IT infrastructure within the GP practice was better understood and supported than that of the hospital trust.

Conclusions and recommendations

Communication is still the main issue within the Trust and local Health Community. The implementation of the knowledge management strategy, which includes a number of proposals for enhancing communications across the Trust and local Health Community, would have a direct effect on this area.

The strategy proposes the development of an action plan for 2000/2001 in conjunction with the knowledge management group. It is envisaged that the action plan may include items such as:

Team building away-days for multi-disciplinary and multi-functional staff

‘Shadowing’ of roles to facilitate a better understanding of each other’s role both within the Trust, within Primary Care and between local Trusts. (Both of the above are intended to deal with some of the communications issues raised by this project by enhancing communication across the health community and facilitating integrated working practices.)

A facilitator to work with departments on particular operational issues and standards. (This is intended to alleviate some of the frustrations highlighted in the evidence-based practice matrix.)

Benchmarking. (To facilitate best practice across the health community.)

Development of an ICT Learning Centre to provide structured training and less formal ‘drop in’ facilitated workshops. (This is intended to provide a better understanding of the role of ICT within the health community and to provide a readily accessible ICT training provision on the Trust site. It is generally accepted across the Trust that structured IT training needs to be available to all staff. A drop-in centre with facilitated workshops, available in a central location on the hospital site is seen as a positive step towards open access to IT. This does not, however, solve this issue within Primary Care.)

In addition, the upgrading of Path.Finder to an XML format from XGL and its subsequent networking across the whole health community will further enhance the general communication and knowledge flows.

It is important that information about the work of the Trust, its goals and future plans, is available not only to clinical staff, but also to support staff and patients. 24-hour access to this information is crucial and this is where the IT infrastructure can provide the solution.

The Trust’s IT infrastructure and support is still considered an issue to be tackled by most staff within the Trust, including Primary Care, where Pathology Results reporting is seen as important.

The development of a knowledge management strategy within the Trust and local Health Community will develop the three main areas identified above - culture, content and technology and will support Sandwell Health Community in achieving its main goals. Its aims are:-

Culture. To foster an empowering culture in which the knowledge, skills and abilities of all Health Service staff are valued, developed and deployed in support of patient care.

Content. To increase access to high quality information for clinical staff, GPs, community partners and patients.

Technology. To resource and support at all levels, the development of a robust and reliable technological infrastructure and user environment.

The strategy has now to be ratified within the Trust and Sandwell is looking to extend the project to other areas of the Trust and Primary Care groups.

References

1. D. Pugh, People in Organisations, (Penguin, 1983)

2. It is relevant to note that, at the same time as the knowledge management project was running at Sandwell, a parallel project for the rollout of Pathology Results Reporting via the clinical PCs on wards was also running. This had an effect on the feedback received, as people had some difficulty in separating the two projects in terms of the IT infrastructure.

Paula Kingston is a project manager at Sandwell Healthcare NHS Trust.

Lynne Cooper is a project manager at Sandwell Healthcare NHS Trust.

Dr Elizabeth Hughes is a clinical tutor, Sandwell Healthcare NHS Trust

Dr Andrea Mayne is director of clinical effectiveness at Sandwell Healthcare NHS Trust.

Jooli Atkins is a training consultant at Atkins Business Solutions Ltd.

They can be contacted via: jooli@atkinsbs.co.uk