Digital Twins: What Healthcare Organizations Can Learn from Other Industries

Digital twins are helping organizations make more informed decisions. These analytical tools use real-time data to represent a product or process, creating a realistic simulation that can be used to test, monitor, or predict outcomes for its real-world counterpart.  

Over the last few years, digital twins of a product (DToP) have grown popular in the manufacturing industry, where creating virtual models and simulating business processes allow manufacturers to better predict maintenance or evolve production workflows. Digital twins have also become popular in retail, where digital twins of a customer (DToC) build on the familiar concept of a marketing persona to map the range of possible customer experiences with a brand, providing companies with predictions of future consumer behavior. 

Now, digital twins are making their way into the patient experience. Healthcare digital twins (HDT) is a quickly emerging technology for better understanding patients and their behaviors, providing healthcare organizations with real-time reporting for data-driven decisions. But what specific applications can digital twins help solve in healthcare? And what can healthcare organizations learn from other industries about applying digital twin technologies to patients? 

To answer these questions, we need to take a closer look at how digital twins work in healthcare. 

It Starts with Data

While other industries can provide some guidance, there are some key differences in how healthcare organizations can implement and use digital twins. One is data. 

Digital twins, and analytical applications like them, consume data as fuel. In industries like retail or manufacturing, it makes sense to choose the application for your digital twin before deciding what data you’ll need to collect to make it run. The more data you have access to, the more accurate the results, and the more applications you can use it for – resulting in some companies collecting the data they need now, the data they’d like now, and the data they think they might want for an application they haven’t yet dreamed up. As consumers, many of us have become inured to it. How many times have you clicked “I agree” on a website’s data capture agreement without a second thought? 

Choosing the data before deciding on the application might seem backward.  But any healthcare-related digital twin must start with the data you can ethically access and use. This isn’t to say that other industries don’t have to worry about ethics in data capture, but in healthcare, the stakes are much higher. Someone in the zip code 49931 purchasing a TV with a warranty on May 22 is very different from someone in that same zip code on that same day being treated for a rare form of leukemia. Those two individual pieces of data I used as an example are identifiable enough to violate patient privacy. The zip code I used as an example is in a very small community. Someone with a one-in-a-million medical condition, who lives in a place with a population of 50,000, could potentially be identified just by their medical condition and zip code.

Before you start with digital twins, here are the fundamental questions to answer first:

• Can I capture this data?
• How can I use this data and what are the limitations of using this specific data?
• What specific applications will this data be used for?
• Does the person that the data came from know that their data is being used — generally or for this specific purpose?
• How can I capture this data ethically?

What does it mean to ethically capture data?

Before you begin deploying data to create a digital twin, it’s important to make sure your data usage is in line with all the relevant data privacy requirements. One action that can help with this is developing consent-based data practices, like establishing compliant opt-in and opt-out controls for data, or even creating patient-facing portals where patients can manage how their data is being used.

For consent-based data practices, your patient should be able to answer or find the answers to the following questions: 

• Do you use my data to train your proprietary models?
• Are your proprietary models centralized or segmented?
• Can I opt out of you using my data?
• How can you enable me to customize my experience?  
• What are the controls on who can use my data? What are they able to see? 
• How is this data masked or encrypted?

Plan in advance and narrow your scope

Once you know what data you’ll be able to use and narrow down the specific applications you can use it for, you need to decide the scope of your application. What will you be studying, managing, or measuring, and how do you intend to do this? The narrower and more targeted you can be with your use case, the better the data usage and upfront planning. A narrow scope means you won’t be able to solve broad sweeping applications, like a digital twin of a community or a hospital, or a segment membership, but it will allow you to design appropriate guardrails around data collection and management.

Putting that data to work: Potential use cases for digital twins in healthcare

Digital Twin of an Asset

Whether it’s an MRI or a heart rate monitor, hospitals rely on the reliable performance of physical assets. Modeling physical assets and keeping track of maintenance, transportation, and average performance can help healthcare organizations manage these assets, ensuring they are consistently maintained, in good working order, and ready when needed.
   

Digital Twin of a Community

A digital twin of a community helps healthcare organizations create better data models on a block, neighborhood, or zip-code level. This multifaceted simulation, more digital hivemind than digital twin, uses an ensemble model – or a series of diverse models, run simultaneously, to predict a general outcome. A digital twin of a community can simulate how a community uses a specific health service provided in a specific hospital or demonstrate how a hospital opening might impact different populations. A digital twin of a community can’t be used to generate responses in the same way a digital twin of a patient or customer could be, but by running a thousand parallel simulations of slightly different people, you can measure the effects that a specific action will have on the overall health of a community, enabling healthcare organizations to make more informed and equitable decisions and achieve better population health outcomes.  

Digital Twin of a Patient

Healthcare is a constantly changing field. To ensure that healthcare professionals remain updated on new research and procedures, the industry requires frequent education and recertification. By creating digital twins of patients, with different kinds of personas, healthcare organizations can reinforce behaviors, processes, and procedures. 

But medicine is more than process and procedure. Some of the long-term indicators of success in healthcare include softer skills like empathy, warmth, and fixing issues in a way that makes patients feel genuinely cared for. Right now, in retail, digital twins of a customer are used to train contact center agents on how to handle tricky or difficult conversations with an angry customer. 

Digital twins of a patient can provide a similar opportunity for medical professionals by giving them a chance to practice difficult or tense conversations with different patient personas.
If it sounds counterintuitive to suggest humans practice empathy in a digital simulation, ChatGPT recently scored 41% higher than human doctors in a study evaluating empathy in answers to medical questions. Perhaps one of the best opportunities for digital twins in healthcare can be in preparing nurses, doctors, and even insurance providers to better engage with scared, anxious, or difficult patients, resulting in improvements in bedside manner and patient outcomes.

About Aaron Schroeder

Aaron Schroeder is the director of AI solutions and head of the AI Center of Excellence at TTEC Digital. In this role, he helps new and existing clients take advantage of the values that AI offers for improving speed, consistency, and innovation in the customer experience.

manufacturing
medical
healthcare
ai
digital twins
medicine
hospital