What’s a Digital Twin?

A digital twin virtualizes a hospital (or other) system to create a safe environment in which to test the impact of potential change on system performance. In other words, to play “what if?” with system dynamics. This is important because healthcare delivery is massively complex. Common sense, spreadsheets, and statistics just don’t have the horsepower to inform strategic decisions. 

By creating a digital twin of your hospital(s) (and related ambulatory services), we can test potential changes in operational strategy, capacities, staffing and care delivery models to objectively determine what actions to take. A great example of this is our ability to test potential OR scheduling optimization options for their impact on the efficiency of the system – including down stream bed and post-acute resources. We use this tool to conceive and test options which result in specific chartered improvement projects (with targeted metrics) that combine to achieve objectives. It also frequently results in the discontinuation of other projects that are interesting but not critical path to the core performance of the system.

The platform we use to do this was developed by GE data scientists at our Corporate R&D Center in Schenectady NY and leverages a combination of discrete event and agent based methods. The system was designed to specifically account for the nuances of care delivery and how patients flow through a healthcare system – we have used it successfully at over 50 clients including Johns Hopkins, University of Michigan, Humber River and others. GE provides interview and observation request and works with scheduler to develop onsite schedule.

Are Digital Twins New?

Not exactly. Digital twins use discrete-event-simulation techniques which have been around for 30 years and applied successfully in healthcare to model departments like radiology. But modeling a hospital above about 400 beds has proven too difficult for all but the most experienced modelers using the best tools. 

What is New?

What’s new is using digital twins to design efficient new hospitals and to redesign system dynamics in existing large hospitals. "System dynamics" includes bed mix, staffing, model of care, floorplan, bed algorithm, etc. This is becoming more common with better toolkits and more experienced practitioners at companies like GE and EY. For example: GE analytics consultants using our healthcare-specific simulation platform have modeled >1,000 bed academic medical centers 75% faster than teams of PhDs using traditional methods.  

How are Digital Twins Revolutionizing Strategic Planning?

Digial twins enable massively collaborative, data-driven, and scenario-based decision making. Without a digital twin, leaders rely on tribal knowledge and basic analysis to plan new facilities and next year’s budget for existing facilities. This is normal but it leaves much to be desired. With a digital twin, leaders virtually test changes to bed mix, bed algorithm, task assignment, floorplan, equipment, ALOS, model of care, staffing etc.

The traditional answer is to do our best and see what happens.

  • For example: neuro has recruited two new surgeons, medicine is closing a unit, we’re opening a transitional care unit, the State is buying our rehab unit to convert it to psych beds, and we expect to reduce ALOS for knees by .75 days and for general medicine by 0.2 days. What will that do to ED Boarding? What is our maximum volume with different scenarios of growth by cohort? Can we accommodate the neuro volume? What’s the best day to add these cases to the OR schedule?
  • With the Digital Twin, we learn that we can accommodate the volume but only if the ALOS work succeeds. We add the cases Wednesday and shift two orthopods from Thursday to Tuesday. Alternatively, we could upgrade the transitional care unit to an ICU (but that’s expensive). These answers lead to new questions… which are tested in the digital twin.

Digital Twins Revolutionize Planning in Four Ways:

  • Digital twins close the gap from “requirements” to system dynamics. Today this is a leap of faith. The simulation model closes that gap when we design new facilities, when we redesign existing patient flow, and when we convert service-line volume plans to annual budgets.
  • Digital twins target process improvement efforts by putting each process improvement project into larger context. This enables us to charter projects with specific goals tied to both local and system performance. 
  • Digital twins facilitate massively collaborative strategic planning. Health systems are full of super smart leaders with ideas. Those ideas need to be heard and tested. The digital twin gives us the tool do so. In many cases the result is to demonstrate that some ideas are bad. That’s a great result because it allows that leader to move forward and embrace the eventual strategy the Digital Twin helps to clarify.
  • Digital twins can also power ongoing short-term forecasts. For example, when we build a digital twin in our Hospital of the Future Analytics Platform to redesign a medical center's system dynamics, we use the same simulation model to power predictive decision support apps outside-the- EMR.

In the end, digital twins help leaders design and execute models of care which are good for patients, families and caregivers. Revolutionary.

Want to know more about how we can help with Digital Twin? Contact Us