Computing and Information Systems
Volume 3 Number 1
Towards a Decision Support Aid to Assist with the Discharge Decision Making Process in the domain of Mental Health Care
The aim of the research is to develop a decision support aid to assist with the complex discharge decision making process in the domain of Mental Health Care. The method is seen as useful as it focuses upon learning about the whole domain out of which may come the elicitation of information about decision making within the domain.|
In this paper the use of AIM to produce models of the discharge decision process in the domain of Mental Health Care is described and illustrated and the use of these models as a valuable learning exercise for those involved and as a basis for a technology-based decision support aid is discussed.
Keywords: Mental Health Care, Decision Support Aid, Appreciative Inquiry Method (AIM), modelling.
Over the past decade there has been media and public interest in the discharge of patients hospitalised under the Mental Health Acts. Media interest has been stimulated by:
- reports of discharged patients, such as the cases of Christopher Clunis (HMSO, 1994, "The Report of the Inquiry into the Care and Treatment of Christopher Clunis") and, more recently Andrew Robinson (discussed in the Blom-Cooper report), who have carried out violent attacks on members of the public; and
- the Government's policy of 'Care in the Community.'
One result of the increased interest in mental health was contained in the Queen's speech, at the opening of the 1994/95 parliament, which announced the Government's intention to introduce new legislation to tighten controls on patients being discharged into the community. This new act will be introduced in April 1996 in Scotland and in England and Wales.
The project is being carried out by modelling the discharge decision of individuals hospitalised under the Mental Health (Scotland) Act (1984) using an appropriate knowledge elicitation method. When all those involved are satisfied that the model (or models) produced represent what the discharge decision making process is all about then the resulting model (or models) will be used as a basis for developing a technology-based decision support aid. The difficulty of eliciting an individual's and/or a teams' 'knowledge' in complex domains has resulted in the development of the Appreciative Inquiry Method (AIM), a method of inquiry which appears to be appropriate to the elicitation of what we refer to as judgmental and experiential 'knowledge' as well as the more factual, rule-based aspect of human expertise. This method emphasises exploration and learning on the part of all those involved as a way of building up a rich appreciation of the domain, as it represents an attempt to operationalise Vickers' notion of appreciation.
Having modelled the discharge decision making process the intention is to develop a technology-based tool to help those involved in decision-making manage the discharge decision-making process itself (i.e. a tool that focuses upon supporting the process that has been modelled in the elicitation exercise rather than in 'simulating' or 'automating' the process as with traditional expert systems and decision support systems).
THE DISCHARGE DECISION
The Mental Health Act (1983) in England and Wales and the Mental Health (Scotland) Act (1984) in Scotland makes provision for the hospitalisation and treatment of individuals judged to be suffering from a mental disorder. When deciding if a patient should be discharged the Responsible Medical Officer (RMO) consults the various professionals who comprise the multidisciplinary team (MDT) which is responsible for the decision making process. Expert advice is sought from all of the medical team involved with the patient's treatment such as nursing staff, including the Community Psychiatric Nurse (CPN), Occupational Therapists (OT) (if any) and Social Workers. It is recognised that the practice of the act has to be balanced against the individual's civil liberty rights (Carson, 1990). The decision making is complex as it is a result of MDT decision making and it results from an assessment of a mixture of rules (derived from the Mental Health Acts) and 'personal' knowledge gained from professional judgement acquired from previous experience. If the RMO decides that the patient should be discharged then, in consultation with the MDT, a discharge plan or Care Programme is formulated to assess the services and care the patient will need in the community.
Figure 1 shows the groups involved in the discharge decision making process. In Scotland the Sheriff, the Mental Welfare Commission and the nearest relative all have the power to discharge a patient but the most common method is by an MDT decision within the hospital.
Figure 1. A System Picture showing the groups involved in the discharge decision process (Stowell, West, Smith and Begg, 1995)
NATURE OF THE RESEARCH
The aim of the research is to model the discharge decision and when all those involved are satisfied that the model (or models) produced represents their perceptions of the discharge decision-making process then the model (or models) will be used as a basis for developing a technology-based decision support aid.
The decision support aid will have 4 main uses. These are:
- to assist with the responsibility for risk management associated with the discharge decision making (Smith, 1993);
- as a representation of the overall decision making process it can be used to help formalise and standardise the decision-making process;
- as a 'training aid' to help those new to the decision making process to see how the process takes place. The decision support aid will allow them to explore the entire discharge decision process; and
- as a foundation for discussion between the experts involved in the discharge process.
THE DESIGN OF THE DECISION SUPPORT SYSTEM (DSS)
A recent suggestion by Isenmann (Isenmann, 1993) is that the DSS could be based on a view that problem solving takes place by structuring a discussion about the domain. Isenmann discusses the solving of "wicked problems" where "all participants in the problem solving process may have differing interests, opinions, and judgements on every aspect of this process" and where "processes of problem solving involve more than just finding solutions to well described problems." A DSS could be produced which is based upon: issues emerging when dealing with a problem; different assessments; alternative procedures; controversial opinions; and various lines of reasoning. The most important elements are issues and questions emerging in the context of the decision making process.
The discharge decision-making process is a "wicked problem" as the discharge decision-making process is a process that is complex and difficult to describe where the different members of the MDT have their own interests, opinions and judgements on the discharge of individuals. This idea of a DSS may be very important as it may give a view which could be used in the development of the computer-based decision support tool to assist with the discharge decision by concentrating upon the process of inquiry that needs to be undertaken for good decision-making rather than on solutions.
MODELLING OF THE DISCHARGE DECISION MAKING PROCESS
The modelling of the discharge decision process is being carried out by using an appropriate method of inquiry. The domain of Mental Health Care can be described as being ill-structured, subjective and complex in that it is a mixture of clear 'rules' and the interpretation of these 'rules.' The former derives from the Mental Health Acts and the latter is a result of personal judgements based upon experience and individual interpretation of clinical evidence. The chosen method of inquiry needs to cope with this level of complexity and some research into an appropriate method has been undertaken and one selected which it is believed will enable the task to be addressed.
There are many different knowledge elicitation (KE) methods such as Protocol Analysis, Repertory Grid and Structured Interviews (Neale, 1988; Cordingley, 1989). In reviewing the wide range of KE approaches, Stowell and West (1989) have offered criticisms and based upon these criticisms which are relevant to the domain under investigation in this research the KE method that has been selected for use is the Appreciative Inquiry Method (AIM). This method has been developed out of work from a previous research project (West, 1991), as a means of eliciting not only the 'facts' and 'rules' of the domain but also the more intuitive and other forms of 'tacit' knowledge within the domain of expertise (Stowell and West, 1990; West, 1992, 1995). AIM represents an attempt to produce a KE approach that operationalises Vickers' notion of 'appreciation' and the process of appreciation (Vickers, 1965).
KE, PHASE I
The inquiry method used in this research as indicated above is AIM. Details of the approach can be found in West (1995). Briefly, the approach consists of 3 phases which relate to each planned meeting with the domain expert rather than a particular division of the task.
Phase I: Elements relevant to a central subject are recorded in a pictorial fashion in the form of a "systems map". Environmental influences can be shown as tertiary layer "bubbles." The approach allows the expert to work on this representation alone without interference from the knowledge elicitor since there are no "rules" to the process other than the constraint of the form of the map itself. At the end of the first session it is expected that the map will offer a full, but relatively low level, view of the expert's thoughts about the problem defined in the central "bubble."
Phase II: Discussion of the map drawn by the expert at Phase I is likely to have revealed additional information which might be used to direct the next phase of elicitation. The aim of the second phase of the elicitation is to explore this information further.
Each element of the expert's map can be viewed as a "system" and can be described explicitly as in the process of developing Root Definitions (RD) in Soft Systems Methodology (SSM) (Checkland, 1981). A six-point checklist for RD construction, which is represented by the mnemonic CATWOE, is asked about each element in the expert's map. A RD is then constructed from the elements of CATWOE. The second part of this phase of elicitation is that of converting the detailed definition of the system of interest into a format that can be returned to the expert for comment and which can be used as a means of enabling further discussion about the "system" identified. The RD is converted into an activity diagram in the form of a "Conceptual Model" (CM).
Phase III: As a means of allowing the expert's own revelations about the subjective area and the increasing understanding of the knowledge elicitor to contribute to the elicitation of further detail the different activities defined in the CM are used as an agenda for discussion between the expert and the knowledge elicitor.
PROGRESS SO FAR
The first interviews using AIM were conducted with the RMOs (clients), who are the psychiatrists who are ultimately responsible for the discharge decision. The first phase involves the 'client' in developing a systems map of their perception of the area of interest (Stowell, West and Fluck, 1991; West, 1995). A Composite Map was then developed which encapsulated the components from the psychiatrists' individual maps, as shown in Figure 2.
The composite map has been reviewed by each of the psychiatrists for comment, discussion and where necessary further development. Through this discussion the original concepts of the individual systems maps have been grouped together into 5 main areas, which have been 'validated' by the different psychiatrists as offering an acceptable and useful way of looking at the area of concern at this stage of the project. The five main areas are: the patient's mental state; the patient's or relative's wishes; home circumstances which include community support; resources which includes pressure on beds and transfer to other hospitals; and the views of the MDT. Each grouping has been considered in terms of the overall activity represented in the named group. This is achieved by the psychiatrists considering each grouping as a Transformation and supplying the details necessary to enable the analyst to develop a strict description of the activity in the form of a RD (Checkland, 1981; Checkland and Scholes, 1990; West, 1995). The RDs have been developed as CMs which are used as the basis for a more detailed discussion about the domain.
Figure 2. A composite systems map derived from the individual systems maps of psychiatrists working in IPCU at Leverndale Hospital, Glasgow.
Different members of the MDT have their own specialist areas of interest and it is hoped that by gradually building up a picture of the whole domain a greater awareness by all those involved (researcher and clients) of the decision-making process and the complex interaction of the many factors involved can be achieved. The RMOs were interviewed first, as those who are likely to have the most complete picture of the discharge decision making process, as they have the ultimate responsibility for this process. Then the nursing staff including the CPNs, the OT department (if any) and the Social Workers were interviewed as they tend to look at their own specialist area within the overall decision process. Individual conceptual models of the domain have now been developed for each expert and the aim is to work towards a detailed conceptual model of the domain that can be used as a basis for group discussion between the participants in the discharge decision making process and act as the foundation for a decision support aid.
Figure 3 shows the CM developed from the RD of one of the five main areas, namely assessing the patient's mental state, as identified from the composite systems map of the psychiatrists working in IPCU in Leverndale Hospital, Glasgow.
Figure 3. A CM built from the RD of the area looking at the mental state as identified from the composite systems map of the psychiatrists working in IPCU in Leverndale Hospital, Glasgow in Figure 2.
INTENDED FURTHER WORK
The conceptual model of the discharge decision process will then be used as a basis for the development of a prototype decision support aid. Previous research has shown that there are various possible methods that can be used to link the conceptual model to technology such as object orientation (Ward, 1989; Stowell and West, 1994), the use of data flow diagrams (DFDs) (Sawyer, 1991; Stowell and West, 1994; Mingers, 1995) or the use of the composite conceptual model itself in the decision support aid. Research is being undertaken to investigate possible methods of linking the model to technology.
It is important that the development of the model of the discharge decision process and the DSS is not 'technology lead' but it is envisaged that the important contribution of the prototype will be in exploring the issues involved in the complex decision process rather than in providing an end product. The aim is to use IT to support the decision making process not by supplying a diagnostic aid but by providing a framework for inquiry (Isenmann, 1993) that may be accessed by those interested in making the decision.
The staff of the Intensive Psychiatric Care Units at Leverndale Hospital and Gartnavel Royal Hospital have agreed and are taking part in the field study phase of the project.
- Carson, D. , 1990
- "Risk-taking in Mental Disorder; Analyses, Policies and Practical Strategies," SLE Publications LTD., Chichester.
- Checkland, P.B. , 1981
- "Systems Thinking, Systems Practice," Wiley: Chichester.
- Checkland, P.B. and Scholes, J. , 1990
- "Soft Systems Methodology in Action," Chichester: Wiley.
- Cordingley, E.S. , 1989
- Knowledge elicitation techniques for knowledge-based systems, in: "Knowledge Elicitation - Principles, Techniques and Applications," D. Diaper, ed., Ellis Horwood Ltd., Chichester.
- HMSO Mental Health Act 1983
- Chapter 20, HMSO, London.
- HMSO Mental Health (Scotland) Act 1984
- Chapter 36, HMSO, London.
- HMSO, 1994
- "The Report of the Inquiry into the Care and Treatment of Christopher Clunis", Presented to the North East Thames and South East Thames Regional Health Authorities, February 1994, London, HMSO.
- Isenmann, S. , 1993
- How to deal with wicked problems using a new type of information system, in: "Systems Science: Addressing Global Issues," F.A. Stowell, D. West and J.G. Howell, eds., Plenum Press, New York.
- Mingers, J. , 1995
- Using soft systems methodology in the design of information systems, in: "Information Systems Provision: The contribution of Soft Systems Methodology," Stowell, F.A., ed., McGraw-Hill.
- Neale, I.M. , 1988
- First generation expert systems: a review of knowledge acquisition methodologies, The Knowledge Engineering Review 3: 2: 105.
- Sawyer, K. , 1991
- Linking SSM to DFDs: the two epistemological differences, Systemist 13: 2: 76.
- Smith, S.A. , 1993
- Modelling the discharge decision process in the domain of mental health to produce a computer-based decision support aid in Multimedia/Hypermedia, Systemist 15: 4: 205.
- Stowell, F.A. and West, D. , 1989
- Expert systems: ramifications for the knowledge engineer, Systems Analysis, Modelling and Simulation 6: 9: 673
- Stowell, F.A. and West, D. , 1990
- The contribution of systems ideas during the process of knowledge elicitation, in: "Systems Prospects: the Next Ten Years of Systems Research," R.L. Flood, M.C. Jackson and P. Keys, eds., Plenum Press, New York.
- Stowell, F.A., West, D. and Fluck, M. , 1991
- The appreciative inquiry method: an approach to knowledge elicitation based upon the notion of knowledge elicitation as an inquiring system, Systemist 13: 4: 154.
- Stowell, F.A. and West, D. 1994
- "Client-led Design: A Systemic Approach to Information Systems Definition," McGraw-Hill.
- Stowell, F.A. , West, D. , Smith, S.A. , Begg, C.E. , 1995
- Decision-making in mental health care: towards a technology-based decision support aid, in: "Critical Issues in Systems Theory and Practice,"
Ellis, K., Gregory, A., Mears-Young, B., Ragsdell, G., eds., Plenum Press, New York (in print).
- Vickers, G. , 1965
- "The Art of Judgement", Harper and Row Ltd., London.
- Ward, P.T. , 1989
- How to integrate object orientation with structured analysis and design, IEEE Software 6 : 2
- West, D. , 1991
- "Towards a Subjective Approach to Knowledge Elicitation for the Development of Expert Systems," PhD Dissertation, Portsmouth Polytechnic (unpublished).
- West, D. , 1992
- The process of knowledge elicitation and its relevance to the design and development of hypermedia/ multimedia decision support aids, in: "Proceedings of the 3rd International Conference of the Information Systems Developers Workbench," University of Gdansk.
- West, D. , 1995
- The appreciative inquiry method: a systemic approach to information system, in: "Information Systems Provision: The contribution of Soft Systems Methodology," Stowell, F.A., ed., McGraw-Hill.
S. A. Smith is a Research Student at the University of Paisley
© University of Paisley, 1996.