SAVI (Sheffield Accelerated Value of Information)
SAVI is a web tool that quickly estimates how much value there is in doing more research to reduce uncertainty in health-economic decisions. It focuses on the potential benefits of reducing uncertainty in costs and health outcomes.
At a glance
Use when
You need to quickly assess the value of reducing uncertainty in health-economic models, especially when prioritizing future research or identifying key drivers of uncertainty.
Avoid when
The PSA sample is too small or not representative, or when high precision in VOI estimates is required without validation of the regression fit.
Inputs
Probabilistic sensitivity analysis (PSA) output from a health-economic model, including sampled parameter values and corresponding net benefit or cost-effectiveness outcomes.
Outputs
Expected Value of Perfect Information (EVPI), partial Expected Value of Perfect Parameter Information (EVPPI) for selected parameters or groups.
How it works
SAVI is an R/Shiny web application developed at ScHARR, University of Sheffield, that computes the Expected Value of Perfect Information (EVPI) and partial Expected Value of Perfect Parameter Information (EVPPI) using Strong's regression-based method applied to probabilistic sensitivity analysis (PSA) outputs from health-economic models.
- HTA domains
- Clinical Effectiveness, Costs & Economic Evaluation, Aspects Beyond HTA
- Assumptions
- The method assumes that the PSA sample is sufficiently large and representative of the joint parameter uncertainty, and that the regression model used (e.g., generalized additive model) adequately captures the relationship between input parameters and net benefit.
- Strengths
- Computationally efficient compared to traditional nested Monte Carlo methods,User-friendly web interface lowers technical barriers,Enables rapid identification of high-value research priorities,Based on established regression-based approximation methods
- Limitations
- Accuracy depends on quality and size of PSA sample,Regression model misspecification may bias EVPPI estimates,Limited to models with available PSA outputs
- Also known as
- Sheffield Accelerated Value of Information
Questions this answers
- › What is the value of eliminating uncertainty in all model parameters?
- › Which specific parameter or group of parameters contributes most to decision uncertainty?
- › How much could future research on specific parameters improve decision-making?
- › What is the expected benefit of conducting further research in monetary or health outcome terms?
- › Which parameters should research prioritize to maximize value?
- › How does decision uncertainty affect the cost-effectiveness of health interventions?
References & sources
Similar by meaning
Beta record. Based on the original catalogue summary; primary-source enrichment pending.