Discrete-Event Simulation Modeling Framework for Cancer Interventions and Population Health in R (DESCIPHR): An...

Area of Science:

• Health policy
• Decision science
• Computational epidemiology

Background:

• Simulation models are crucial for health policy, especially for cancer, a leading cause of death globally.
• Discrete-event simulation (DES) and Bayesian calibration are powerful tools for modeling complex health conditions and quantifying uncertainty.
• Existing guidance for end-to-end DES cancer modeling and Bayesian calibration for policy evaluation is limited.

Purpose of the Study:

• To introduce DESCIPHR, an open-source R framework for cancer intervention and population health modeling.
• To provide end-to-end guidance on structuring DES models, parameter estimation using Bayesian calibration, and policy evaluation.
• To facilitate the development of future decision models for cancer interventions.

Main Methods:

• Developed the DES Modeling Framework for Cancer Interventions and Population Health in R (DESCIPHR).
• Integrated a flexible DES model for cancer natural history with Bayesian calibration for parameter estimation.
• Applied DESCIPHR to calibrate bladder and colorectal cancer models using real-world cancer registry data.

Main Results:

• Demonstrated the framework's utility by calibrating bladder and colorectal cancer models.
• Introduced an automated method for generating data-informed parameter prior distributions.
• Enhanced a neural network emulator-based Bayesian calibration algorithm for improved functionality.

Conclusions:

• DESCIPHR provides an adaptable template for constructing decision models in cancer research.
• The framework supports robust evaluation of cancer interventions and health policies.
• Facilitates better understanding of cancer epidemiology and data uncertainty in policy decisions.