Digital Transformation of Emergency Medicine: How commercial mathematics is helping hospitals adapt to COVID challenges

As travel restrictions began to ease across Australia at the start of 2022, hospitals experienced an increased strain on resources, brought about by a peak in COVID cases.

Planning to expand their emergency department (ED) to compensate for lost space from accommodating COVID query patients, one of SEQ’s largest public hospitals engaged the services of Biarri to ensure they were achieving optimal utilisation of their facilities.

Though the new building was originally designed for the treatment of minor injuries, the impact of COVID, as well as historically high levels of demand on the ED prompted a reconsideration of the design and intended usage for the expansion. 

The hospital needed to understand how they should change the existing ED whilst utilising the beds and resources of this new physical area, to best improve the flow of patients, reduce their waiting times, and treat them as efficiently as possible while continuing to provide excellent patient outcomes.

Over the 4 months in 2021, the average time a patient spent waiting for a ward bed exceeding the expected time of 75 minutes by 25%. 

Within a restricted time frame, Biarri was able to deliver insights based on a custom-built simulation model, that showed the department how best to optimise throughput, maximise their use of space, and minimise the overall time patients spent in the ED. 

Simulation Modelling

You have a problem. You’ve come up with some possible solutions that work in principle, but you want to be certain of their effectiveness prior to implementation. Often, it’s impractical to test them in reality, as they’re either too expensive, too time consuming, or there would simply be too much interference with the normal operation of your business. You may need a simulation

Simulation is a time and cost effective way of testing ideas and theories across a huge number of disciplines, from wind tunnel testing on scale models in the aerospace industry, to predicting animal behaviour in large groups, or recreating the formation of our Milky Way galaxy. 

Because they are easily configured to take advantage of randomness, running a simulation multiple times can produce a wide spectrum of possible outcomes, which sometimes makes them a better choice to model reality than more traditional techniques. 

Using this method of exploring the problem space, a business can develop their operational plans based on a typical or likely day (or month), while also gaining visibility of, and preparing for the worst-case scenarios. 

By representing the allocation and flow of resources as a series of discrete events, simulations can serve as a digital testing ground for a business, and provide the opportunity for fine tuning and optimisation of business processes. 

Hospital Emergency Department – A Case Study

A simulation algorithm was developed for the ED that allowed the identification of bottlenecks in the department, and could be used to explore the impact of reducing wait times, reallocating beds to different types of patients, and expanding the areas reserved for COVID patients. 

The simulation treated patients as individual agents who were tracked from arrival, through to triage, into a particular area in the ED, where they could then either leave after being examined, or be admitted to the ward. 

The algorithm made use of 4 months of historic data regarding the expected frequency of patient arrivals, expected time in triage for each patient type, and processing treatment times by ED area. Expected flows between the areas of the ED, for example the percentage of acute patients admitted to ward were also estimated. 

The outcomes of the simulations were measured in terms of a NEAT score. 

“The National Emergency Access Target (NEAT) stipulates that a predetermined proportion of patients should be admitted, discharged or transferred from Australian emergency departments (EDs) within 4 hours of presentation”

For example, it was found that reducing the time that patients wait for a ward bed could significantly improve flow in the ED. If patients only waited 45-60 minutes for a ward bed, compared to the current average, the NEAT for Resuscitation patients could increase around 25%.

By tweaking the parameters of the simulation, different scenarios could be investigated. In one such scenario, it was found that if the COVID area was not expanded, even 40 additional patients per day would lead to significant queuing, as the current allocation of 10 beds would be insufficient to handle the demand. 

COVID queue length (10 beds)

Length of queue for beds over a two day period for COVID patients in the situation where the COVID area has not been expanded. The green curve represents the “average” (or 50th percentile), the yellow a worst case scenario (95th percentile) and the blue a best case scenario (5th percentile).  

COVID bed utilisation (10 beds)

The utilisation of beds in the situation where the COVID area has not been expanded. Within a few hours there is no longer any capacity for additional COVID patients. 


However, expanding the COVID area into the acute and clinical decision units (adding 12 beds) meant the ED could handle an increase of roughly 70 COVID query patients per day.

COVID bed utilisation (22 beds)


The utilisation of beds once an additional 12 beds have been added, in the case of 70 additional COVID query patients. If the COVID pod was expanded into the Clinical Decision Unit area, this amount of space should be enough to meet the anticipated capacity.

Simulation in Mining and Construction

Biarri has used simulation modelling in a variety of industries, such as modelling the movement of roof supports in longwall mining applications to significantly reduce the time taken for their recovery, transport, and installation. Another application for simulation is traffic and the movement of goods through a network. In the animation below, trucks with a random arrival time move through an underground parking structure to deliver pallets to a goods lift, which might have a randomised waiting time. The bottleneck in the structure acts like a traffic light system to control the flow of trucks through the structure. 


Would you like to know more? 

Speak to an expert today and discover how your business can begin leveraging the power of commercial mathematics and simulation today. 

Get in touch

  • Hidden
  • Hidden
  • Hidden
  • This field is for validation purposes and should be left unchanged.

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published.