OBJECTIVES : Model interoperability is important for model transparency. Being able to test a model quickly, independent of software implementation, can increase model understanding and trust. The objective of this work is to demonstrate how a prototype Java Script Object Notation (JSON)-based Health Economic Model Exchange Format (JHEF) can facilitate the model interoperability. Our goal is to demonstrate how a standard representation can be used to connect to already existing platforms and run models generated on them.

METHODS : We have developed a prototype JHEF implementation in a framework storing model-related content like conditions, events, states, transitions and analysis results (we call that Rapid Health Economic Modeling or RHEM). The JHEF, as a computer readable representation, serves as a counterpart to a humanly writable and readable model representation: the actual RHEM language. JHEF is powerful enough to cover both Discretely Integrated Conditions and Events (DICE) and Markov models. The RHEM framework supports the interfacing with already existing health economic platforms like heRo3 or EviDICE through its import and export functions. The JHEF prototype uses a standardized JSON-based exchange format that is readily supported by most platforms.

RESULTS : These concepts are demonstrated through two examples: a DICE model developed in EviDICE and a Markov model created in heRo3. Both models are converted into our JHEF format by import functions. From the input, model code in RHEM Language is produced automatically. We demonstrate that the simulations obtained by our core framework agree with those of heRo3 and EviDICE.

CONCLUSIONS : Developing a markup language for health economic models represents a foundational step towards improving model transparency and interoperability. Our prototype implementation of a standard exchange format and core modeling framework demonstrates that cross-linking across different model development platforms is feasible. We recommend that ISPOR develop a Task Force to consider adopting a health economic modeling markup language.