Healthcare Technology
North Carolina Creates Agent-Based Model to Simulate Disease Spread in Healthcare Settings
Researchers have developed a flexible simulation framework to track patient movement through North Carolina's health facilities, enabling forecasts of hospital strain and intervention outcomes.
Researchers at the North Carolina Modeling Infectious Diseases Program have built an agent-based simulation tool designed to help public health officials understand how patients flow through the state's healthcare network. The framework, documented in a formal protocol, allows epidemiologists and planners to model different disease scenarios and test the real-world impact of control measures.
How the System Works#
The simulation treats individual patients as autonomous agents moving to and from hospitals, clinics, and other facilities across North Carolina. Each agent's movement decisions and the specific facilities they interact with are modeled to reflect realistic healthcare-seeking behavior. This agent-based approach makes the framework adaptable: researchers can plug in disease-specific rules and transmission dynamics on top of the underlying patient movement model.
Core Applications#
The tool supports a range of practical uses for public health planning:
- Intervention testing: Evaluating whether proposed measures—such as isolation protocols or vaccination campaigns—would meaningfully reduce disease transmission
- Resource forecasting: Predicting how an outbreak might strain hospital capacity, staffing, and supplies
- Capacity planning: Anticipating which facilities might face bottlenecks during surges
Design and Documentation#
The project follows the established Overview, Design Concepts, and Details (ODD) Protocol, a standardized framework used across computational biology for transparently describing agent-based models. This ensures other researchers can review, replicate, and build upon the work. The model was configured to reflect accurate data about North Carolina's healthcare landscape and facility locations.
The framework is designed to work alongside disease-specific add-ons that would capture transmission mechanisms unique to particular pathogens. The base model itself does not include such details.
