Digital twins in healthcare: practical lessons from a Canadian hospital

A group of healthcare professionals, dressed in blue scrubs and lab coats, engaged in a discussion around tablets displaying charts

Throughout the history of the AnyLogic Conference, we have had multiple valuable presentations about using simulation in healthcare. In 2016, Pfizer suggested using predictive modeling in healthcare to simulate clinical trials. In 2023, Davita shared their experience of using AnyLogic models to improve the workflow in dialysis clinics.

Even the concept of digital twins in healthcare was one of the topics at the AnyLogic Conference 2022. The NHS Foundation Trust, together with Decision Lab, studied the opportunities for improving operations and enhancing patient experience with the use of digital twins. Though the approach is highly promising, it is not as widely spread in healthcare as in other fields, like manufacturing.

This blog post has two focus points: the manufacturing experience that healthcare successfully adapted to meet the needs and the challenges digital twins face in the healthcare sector.

Contents:

  1. Digital twins in healthcare: the concept
  2. Challenges for simulation in healthcare
  3. Use cases of simulation in healthcare
  4. Ideas for digital twin applications in healthcare
  5. Conclusion

Digital twins in healthcare: the concept and the leaders experience

The concept of a "digital twin" might be familiar to many of you, though interpretations of the term can vary. A digital twin is a detailed virtual copy of a physical system. The twin helps you monitor, analyze, and simulate processes in real time.

Today, the manufacturing industry is seen as a leader in using digital twins. The Digital Twin Market Report says that by 2027, 29% of manufacturing companies will use digital twin technology. There is a lot more information about how digital twins are used in manufacturing than in any other field.

Comparison of the cumulative number of papers about digital twins distributed among different application domains

Comparison of the cumulative number of papers about digital twins
distributed among different application domains. Source:
A Comprehensive Review of Digital Twin

The case studies on our website can even prove this tendency. If you look for digital twin cases, the manufacturing industry would lead the search results. Here are only a few of them:

An example of adopting manufacturing experience in healthcare

Healthcare can learn from manufacturing, where digital twins have optimized processes effectively. One approach that came from manufacturing was inspired by the Toyota management system. It was adapted to healthcare needs, proved helpful in the field, and was eventually called Lean Healthcare.

A diagram showing Lean healthcare principles organized to prioritize patients and families first

Lean Healthcare System (click to enlarge)

Lean Healthcare aims to reduce "waste" and increase efficiency in hospitals. It involves everyone, from doctors to support staff, working together to eliminate anything that does not benefit patients.

Types of "waste" in Lean Healthcare:

  • Overproduction
  • Overprocessing
  • Untapped human potential
  • Defects
  • Waiting time

So, in Canada, a Saskatchewan Hospital has applied Lean Healthcare. They improved a lot in non-clinical areas such as human resources and supply chains. However, using these principles in clinical decision-making is more complex and less consistent.

Challenges for simulation in healthcare for clinical decision-making

Data availability is one of the primary hurdles in healthcare for simulation integration. Digital twins depend heavily on having access to high-quality data. Unlike in manufacturing, where twins easily get production metrics, digital twins in healthcare face certain challenges when collecting data:

  1. Information Privacy: Protecting patient privacy is paramount in healthcare. Even though there are ways to protect data, like encryption and access controls, they take a lot of time and resources to set up.
  2. Data Accuracy: Healthcare data often includes manually recorded or judgment-based entries, lacking the continuous, precise measurements seen in manufacturing.
  3. Data Availability: The data comes from different systems and formats, including electronic health records, medical devices, and administrative databases. Capturing and integrating it into a digital twin model is a significant challenge. Moreover, accessing real-time data streams for healthcare digital twins adds another layer of complexity to data acquisition.

Another crucial factor is the influence of the community. Factors like family support, access to healthy food, and living in a safe area greatly influence a person's health. These elements make healthcare more complex for simulation.

As you know the challenges, you can try to overcome them. So did the Canadian hospital; they started by integrating simulation into the healthcare system.

Use cases of simulation in healthcare

The Saskatchewan Hospital took the best of the two worlds: simulation modeling and the Learning Health System framework. They developed a multi-layered environment that connects the patient to regional-level decisions. The hospital uses various models tailored to medical problems, enhancing their decision-making speed and accuracy.

The emergency department model

A part of the emergency department model that proved to the management the need for simulation in healthcare

A part of the emergency department model that proved to the management
the need for simulation in healthcare (click to enlarge)

One of the first models the Saskatchewan Hospital team created ten years ago was a simulation model of the emergency department using AnyLogic. It improved patient flow and reduced wait times. The success of this model led the authority to hire a full-time modeler and consider the advantages of using simulation in healthcare.

The COVID-19 dynamics model

During the COVID-19 pandemic, Saskatchewan Hospital used agent-based modeling in AnyLogic to study community-level dynamics and inform managers and policymakers. The modelers combined the community-based model with the hospital model. So, together with the hospital management, they developed strategies for controlling the spread of the pandemic in real life.

The statistics from the COVID-19 spread model

The statistics from the COVID-19 spread model

During the pandemic, the Saskatchewan modeling team used the model's data to answer critical questions, such as:

  1. How does the virus spread through the community?
  2. What is the projected peak and duration of the outbreak?
  3. How will the pandemic impact hospital capacity and resources?
  4. What strategies can optimize hospital resource allocation?
  5. What are the economic and social impacts of different public health measures?

This information has significantly influenced the performance of the hospital’s analytics team. It helped them come up with strategies for controlling the pandemic’s spread and improving patients' experiences.

If you want to know more about how and why it was decided to integrate simulation in healthcare, check out the video from the AnyLogic Conference 2023.



Ideas for digital twin applications in healthcare

In the case of the Saskatchewan Hospital, continuous stakeholder engagement ensures that technologies remain relevant and valuable. Following their experience, here are some focus areas for integrating digital twins in healthcare:

Hospital operations and logistics

  • Predict and manage the availability of medical staff, beds, and equipment.
  • Model emergency scenarios to improve preparedness and response strategies.
  • Optimize the hospital's energy usage and maintenance schedules.

Medical devices

  • Model wear and tear, predict maintenance needs, and improve design for next-generation devices.
  • Monitor device performance in real time and predict potential failures.
  • Simulate and verify device compliance with regulatory standards.

Population health management

  • Predict the spread of infectious diseases.
  • Analyze trends in public health data to inform policy and intervention strategies.
  • Optimize the distribution of healthcare resources based on predicted demand.

Conclusion

Digital twins have the potential to transform the healthcare industry. Despite the challenges unique to this niche, the benefits seen in other sectors suggest a promising future. By embracing digital twins in healthcare, you can make smarter decisions, streamline processes, and ultimately improve outcomes for patients. With continued focus and innovation, digital twins could become a cornerstone of modern healthcare.


Be the first to discover the latest trends in simulation modeling by registering for the AnyLogic Conference 2024.

Learn more about the conference

Postes connexes