How Digital Twins Can Improve Health System Operations

Digital twins (DTs) are detailed virtual representations of objects, systems, or processes. DTs have been around for decades—the National Aeronautics and Space Administration (NASA) is generally credited with establishing the core idea of building an exact “twin” of a spacecraft for simulation and testing before embarking on high-risk space launches [1,2]. Back then, these early twins could only be made from solid materials.

Fully digital today, DTs can “faithfully mirror real-world systems...and provide predictive insights using advanced simulation, machine learning, and reasoning to help decision making” [1]. DTs have experienced a resurgence in popularity in recent years, with applications from health care to education, agriculture, and beyond [2,3].

DT technology can simulate a hospital’s complete ecosystem, including staff workflow, fluctuations in supply and demand, procurement, and complex patient pathways. Compared to traditional simulation models, which typically involve stand-alone analyses without a consistent link to the physical system, DTs allow a continuous exchange of data between a system and its twin, facilitating real-time updating [2].

“DTs empower hospital leaders to make data-driven decisions by testing different scenarios in a safe, digital environment,” noted GE HealthCare in a recent blog post [4]. In health care, where experimenting directly with patient care processes carries real risk, DTs offer a safe environment to test, monitor, and optimize operational changes and build more sustainable systems.

US health care systems are increasingly enthusiastic about implementing DTs to optimize their operations, with the global digital twin health care market expected to experience a 68% compound annual growth rate by 2030, representing a total value of about US $60 billion [5].

DTs in Health Care Management

In health care management, current DT work aims to reduce costs by improving patient flow and throughput, resource allocation, facility design, and workflow optimization. Potential use studies include the following.

Workflow/Patient Flow

Hospitals often struggle with bottlenecks in emergency departments, operating rooms, and inpatient units. “By creating a digital twin of the hospital’s operations, administrators can test staffing levels, bed allocation policies, and scheduling strategies before implementing them in the real system,” said David Andersen, regional director for AnyLogic, a simulation modeling software company. “In several of our healthcare projects with DTs, emergency department wait times dropped by 20%‐40% and overall patient throughput increased by 10%‐20%.”

Surgical Block Optimization

Many systems still rely on static block schedules and manual planning. “Hospitals often sit on more data than they realize and use almost none of it to model trade-offs,” observed Steve Blatney, director of business development for SimWell, a company that uses DTs to simulate and optimize business operations. “We have used DTs with orthopedic health systems to see how changes to block scheduling, case sequencing, and staffing allocation affect throughput and cost—making final decisions before anything goes live,” he said.

Capacity Planning

Hospital capacity can be affected by several factors. GE HealthCare teamed up with Children’s Mercy Hospital (CMH) in Kansas City, Missouri, to optimize capacity planning [6]. DTs have identified demand and capacity constraints, including during the heightened strain of the winter flu season. “CMH used this new knowledge when planning for winter surge capacity planning including spikes of flu or RSV,” said Bree Bush, general manager of GE HealthCare’s Command Center. “As a result, CMH predicted the winter surge within a week of when it actually struck.”

Facility Planning

Physical layouts can have dramatic impacts on patient flow and staff efficiency. Hospitals frequently use simulation models when designing new emergency departments, specialty centers, or inpatient towers. DTs allow leaders to make data-driven decisions to ensure facilities will meet long-term demands.

Effective Deployment and Use

While DTs have tremendous potential, “at the end of the day, success still hinges on an organization’s ability to execute change effectively and manage project implementations well,” said Julie Gibson, head of US corporate press for Siemens Healthineers, a health technology firm that specializes in digital health solutions. “Learning is crucial for rapid success, but knowledge without action loses value quickly.”

Before a hospital can embrace DTs in a meaningful way, everyone—from chief executive officer to night shift nurse—must understand how the process functions—not how it is designed or documented, “but how it actually works on a Tuesday afternoon when two nurses call out and the lab is backed up,” said Blatney. “Until you can get to that shared definition of reality, you cannot build anything useful on top of it. Defining that process is the foundation and is often more valuable than any software.”

Shared understanding facilitates proactive planning. “Hospitals that understand their operations at a systems level have a real advantage, because simulation-based predictive digital twins allow healthcare leaders to move from reactive thinking to proactive planning,” said Andersen. With regular data updates, a DT system can be used to test additional scenarios as frequently as needed for years to come.

Mistakes to Avoid

DTs go wrong if used improperly or built on weak assumptions. Poor or incomplete data is a particular risk. A digital twin relies on accurate inputs about patient arrivals, treatment times, staffing patterns, and operational policies. If inputs are outdated or incomplete, the model can produce misleading conclusions. In practice, “a large part of building a useful digital twin is validating it against real hospital performance and continuously refining it as conditions change,” said Andersen.

Another common mistake is oversimplifying complex operations. Hospitals are highly interconnected systems; changes in one department can affect patient flow across the entire facility. If a model only represents part of the system, leaders might optimize one area while unintentionally creating bottlenecks elsewhere. High-quality simulation models need to capture systemwide interactions.

A third mistake involves treating a predictive digital twin as a one-time technology project, rather than integrating it as an ongoing decision support tool. The most successful hospitals use these models operationally to repeatedly test policies, staffing strategies, and capacity planning decisions.

DTs and the Future of Health Care

DT applications will likely continue to grow across health care planning and operations.

GE HealthCare recently announced collaborations with two leading US health systems to develop new AI-driven hospital operations software using DT inputs [7]. At Virginia Polytechnic Institute and State University in Blacksburg, systems engineering faculty are developing a DT model for primary care practice. “We are examining both observed and digital workflows with the hope of developing meaningful changes that can be implemented without disrupting that care,” said John Epling, MD, professor of family and community medicine at the university’s Carilion School of Medicine. Virginia Tech is also planning a DT study to predict the burnout trajectory for medical staff [8].

To date, DTs in health care systems have primarily involved simulation for planning purposes. The next stage is connecting DTs to sources of real-time data. “When you feed that data into a live operational model, the platform stops being a planning tool and starts being a live decision support system,” said Blatney. “Everyone in the operation can see what is actually happening and ask what-if questions in the moment. That is a transformational step. Organizations that figure out how to connect those two streams will have a competitive edge as thought leaders in their fields.”