An Open-Access Interactive Platform for Discrete Choice Experiment Analysis in Health Technology Assessment

Background: Discrete choice experiments (DCEs) are the most widely used stated-preference methodology in health technology assessment (HTA), yet the analytical outputs of DCE studies coefficient estimates, willingness-to-pay values, market simulations, and budget impact projections are typically presented in static tables and figures that limit stakeholder engagement and scenario exploration. No publicly available, disease-agnostic interactive platform currently exists to enable HTA professionals, clinicians, and policymakers to explore DCE results dynamically, simulate alternative scenarios, and evaluate their economic implications without requiring statistical programming expertise. Objective: To develop, implement, and validate an open-access, interactive web-based platform for DCE analysis that is (a) disease-agnostic with fully customisable attributes and levels, (b) capable of accepting both simulated and empirical DCE coefficient estimates, (c) equipped with real-time market simulation and budget impact projection, and (d) designed for adoption by HTA bodies and clinical working groups globally. Methods: The platform was developed in R Shiny (R version 4.4) and deployed on shinyapps.io. The architecture comprises seven interactive modules: Study Design, Attribute Editor, Model Results, Willingness-to-Pay Analysis, Market Simulation, Economic Evaluation, and Upload/Export. The Attribute Editor enables users to define, add, remove, and modify attributes (continuous or categorical), their levels, domain classifications, coefficients, standard errors, and calibration sources with all downstream outputs updating reactively. The Upload module accepts CSV files of estimated coefficients from any mixed logit model, enabling transition from simulated to empirical data without code modification. The platform ships with a default preset for Alzheimer’s disease-modifying therapy (DMT) implementation (9 attributes, 3 domains) but is fully configurable for any disease area or healthcare intervention. Results: The platform was validated using a simulation-based DCE calibrated to published clinical trial evidence for Alzheimer’s DMTs. Under the default preset, all seven modules function as designed: the Attribute Editor correctly propagates changes to all downstream outputs; the Model Results module dynamically computes relative importance from any attribute configuration; the WTP module automatically detects the cost attribute and computes marginal rates of substitution; the Market Simulation module generates logit-based preference shares that update in real time as users adjust treatment profile sliders; and the Economic module projects 5-year budget impact under user-specified parameters. The Upload module successfully replaced simulated coefficients with uploaded values, with all outputs recalculating automatically. The platform is publicly accessible at https://sevinc.shinyapps.io/dce-dmt-platform/ with source code available upon request. Conclusions: This platform provides the first open-access, disease-agnostic, interactive tool for DCE analysis in HTA. By enabling attribute customisation, real-data integration, and real-time scenario simulation, it bridges the gap between econometric DCE analysis and stakeholder decision-making. The platform is freely available for adoption by HTA bodies, clinical working groups, and researchers conducting preference studies in any therapeutic area.