Computers supporting experts: A system to assist in logistical decision making

Mr Trevor Matthews1

1SA Ambulance Service, Adelaide, Australia

Abstract:

We have developed a computer-based decision support system (CBDSS) which uses machine learning tools to provide time estimates for SA Ambulance Service (SAAS) retrieval of high-level care patients.  SAAS is the primary provider of pre- and inter-hospital transport of patients within the state of South Australia.

Where the patient’s condition warrants high level care, MedSTAR, SAAS’s retrieval arm, is tasked to undertake that transfer.

MedSTAR has multiple modes of transport available with a fleet of road vehicles, fixed- and rotary- wing aircraft to choose from.

The characteristics of each mode of transport and the location of the patient sometimes require judgment decisions to be made with respect to how to send the team.  Another consideration is when to send the team – sometimes delaying a team who is near to shift completion will not significantly affect patient outcome.  In this case,  an estimate of total on-task time will allow the logistician to consider whether there will be shift overrun and the result impact to the roster for subsequent shifts.

The CBDSS we have developed provides decision making support to the experts within MedSTAR. Coordination staff can enter a small number of details known about the patient into a web-based form. The system will then provide a predicted length of time at the retrieval location and total case time.  If required, the model can also provide a total case time estimate across a range of transport options for that case.

These estimates can then be used to provide further guidance to the clinicians making these logistical decisions. This poster will outline the process of building the support tool and the benefit it provides to coordination staff.

 

Biography:

Trevor Matthews is a Rescue and Retrieval Paramedic who has been working within MedSTAR for the last 12 years.  After over 20 years working as a Paramedic, he is now also a master’s candidate with the School of Mathematical Sciences at the University of Adelaide, undertaking some operational research into part of SA Ambulance Service’s dispatch process.