FLAME’s Future
277
accepted that process management is one of the key reasons. Using the
agent-based modeling techniques, a decision support system was developed to obtain information ahead of time in order to optimize the
management of human and material resources. By modeling each patient, members of staff and other related resources as agents, this model
has the nature of flexibility and capability of providing any detail. With
the benefit of modularized design, this platform can also be easily customized for any ED.
Decision support system: The system model is based upon the detailed
set of agent types which have been developed based on not only observations at the hospital but also from the use of historical data going back
three years. The decision support system is composed of a web-based interface and the ABM model running at the back-end. The architecture
of the proposed system is displayed in Figure 9.1. The system could
connect with the real-time patient record database for presenting the
current status of the whole department. Upon user’s request, when a
prediction is needed, a snapshot could be generated, and from there a
faster than real-time predictive simulation would be run based on historical data. The predictive results would be provided to users to enable
them to look ahead of time. If an intervention is necessary, users would
be able to apply the intervention to the model, and re-run the simulation
to test its effectiveness.
FIGURE 9.1: Diagram of ABM-based decision support system.
For example, on arrival in the green zone reception (self-admitting patients) the receptionist agent will collect information from the patient
and when this happens the patient agent will be given an id and a basic
set of details which are then ‘attached’ to it in the model as attributes.
The patient then waits in the waiting area unless they are identified as
an emergency. The next stage will be triage and for this a nurse, when
available, will examine the patient and record the basic aspects of triage.
277
accepted that process management is one of the key reasons. Using the
agent-based modeling techniques, a decision support system was developed to obtain information ahead of time in order to optimize the
management of human and material resources. By modeling each patient, members of staff and other related resources as agents, this model
has the nature of flexibility and capability of providing any detail. With
the benefit of modularized design, this platform can also be easily customized for any ED.
Decision support system: The system model is based upon the detailed
set of agent types which have been developed based on not only observations at the hospital but also from the use of historical data going back
three years. The decision support system is composed of a web-based interface and the ABM model running at the back-end. The architecture
of the proposed system is displayed in Figure 9.1. The system could
connect with the real-time patient record database for presenting the
current status of the whole department. Upon user’s request, when a
prediction is needed, a snapshot could be generated, and from there a
faster than real-time predictive simulation would be run based on historical data. The predictive results would be provided to users to enable
them to look ahead of time. If an intervention is necessary, users would
be able to apply the intervention to the model, and re-run the simulation
to test its effectiveness.
FIGURE 9.1: Diagram of ABM-based decision support system.
For example, on arrival in the green zone reception (self-admitting patients) the receptionist agent will collect information from the patient
and when this happens the patient agent will be given an id and a basic
set of details which are then ‘attached’ to it in the model as attributes.
The patient then waits in the waiting area unless they are identified as
an emergency. The next stage will be triage and for this a nurse, when
available, will examine the patient and record the basic aspects of triage.