Presentation (CTI)

OBJECTIVES: While the options for building cost-effectiveness models (CEMs) in R have grown substantially in recent years, modellers still face tradeoffs between a framework's flexibility, its computational performance, and the ease with which models can be built and used. The openqaly ecosystem aims to improve these tradeoffs, providing a simple declarative framework for developing CEMs.
METHODS: Model definitions are composed of standardized components (e.g. variables, strategies, states) which can contain R expressions to represent calculated quantities such as the value of a variable or the probability of starting in a health state. Partitioned survival and Markov cohort models are supported. Computationally intensive operations are implemented in C++ via Rcpp. Tunnel states are declared by referencing time from state entry in expressions, with health states automatically expanded. Decision trees can be placed in front of models or used in calculations of transition probabilities or payoffs. The companion openqalysurv package provides operations for creating, modifying, and combining survival distributions. The openqalyshiny package adds a Shiny-based interface for creating and viewing models.
RESULTS: The openqaly package supports a range of analyses, including probabilistic sensitivity analysis with multivariate sampling, deterministic one- and two-way sensitivity analysis, scenario analysis, threshold analysis, and value-based pricing. All analyses come with their own array of outputs and visualizations, such as tornado plots, cost-effectiveness planes, cost-effectiveness acceptability curves, and value of information.
CONCLUSIONS: The openqaly ecosystem builds on existing modelling packages by providing a structured process for developing transparent and performant CEMs, while remaining accessible to a wider audience. The source code is released under the GPLv3 license, enabling code reviews and contributions by third-party experts. We hope the availability of this package will facilitate the use of R in health economic modelling and help improve the overall quality and reproducibility of studies in this domain.