OBJECTIVES : The use of open-source programming languages in health economics has the potential to facilitate model transparency, reproducibility, and shareability. Guidance as to how to structure the required components of model building and analysis is lacking. To address this challenge, we propose a generalizable coding framework for model-based decision and cost-effectiveness analyses.

METHODS : The development of this framework encompasses a collective expertise from research and pedagogical experiences. Our framework is based on the premise that comprehensive model-based decision and cost-effectiveness analyses involve common components, regardless of the type of the model. In developing this framework, we strived for flexibility and generalizability to accommodate a diversity of model types and applications. The framework provides a structure for organizing code to conduct decision analyses, allowing for full documentation of the biological, behavioral, and mathematical assumptions and decisions that went into the model development and analysis processes.

RESULTS : The proposed framework consists of a set of common core decision model components divided into five components: (1) model inputs, (2) model implementation, (3) model calibration, (4) model validation, and (5) analysis. The first four components form the model development phase, whereas the analysis component is the application of the final model to answer the policy or research question of interest, conduct sensitivity analyses, assess decision uncertainty, and/or determine the value of future research through value of information (VOI) analysis. We showcase the framework through a fully functional, testbed decision model coded in R, which is freely available on GitHub for download and easy adaptation to other applications.

CONCLUSIONS : This framework provides formal guidance for coding decision models and facilitates the accessibility of more advanced modeling methods to a broader set of users. Adoption of the framework will improve code readability and model-sharing, paving the way to an ideal, open-source world.