OBJECTIVES: Probabilistic sensitivity analysis (PSA), deterministic sensitivity analysis (DSA), and scenario analysis are typically conducted to evaluate parameter and methodological uncertainties in health economic models. While PSA and DSA commonly use data visualization techniques like scatter plots and tornado charts to display outcomes, a standard/recommended method for effectively visualizing scenario analysis data remains undeveloped. This abstract explores various potential visualization techniques to effectively analyze data from multiple scenario analyses.

METHODS: A diverse range of data visualization techniques were examined, spanning categories such as comparison, correlation, part-to-whole and hierarchical, temporal, distribution, etc. For scenario analysis results, net monetary benefit (NMB) was chosen as preferred outcome measure over incremental cost-effectiveness ratio (ICER), as ICER can include a combination of numerical and textual values ("Dominant", "Dominated"). Scenarios with NMB>0 were classified as favorable, while those with NMB<0 were deemed unfavorable. For evaluating these visualization methods, a hypothetical model comprising 100 scenarios was used.

RESULTS: Our analysis identified two-part (diverging) bar charts and radial column charts as the most suitable data visualization techniques, given their ease of implementation in MS-Excel - a widely used tool in HTA modeling. In two-part bar charts, columns are arranged from highest to lowest based on the NMB values for each scenario, with favorable scenarios color-coded in green and unfavorable ones in red. Radial column charts are also effective featuring two distinct charts for favorable and unfavorable scenarios, sorted by NMB values. Additionally, for specific scenario categories, such as varying drug pricing or multiple time horizons, heat maps were identified as another viable option.

CONCLUSIONS: Effective data visualization for scenario analysis offers stakeholders a clear and comprehensive understanding of various scenarios, aiding in informed decision-making. Additionally, use of more advanced software for data visualization can further enhance this process, adding depth and clarity to the analysis.