Value of Information Analysis in Models to Inform Health Policy (Kunst-Heath review)
This method helps decision-makers understand how much it's worth to gather more data when making health policy choices. It estimates whether collecting additional information could improve decisions and reduce uncertainty.
At a glance
Use when
Prioritizing future research in health policy; assessing robustness of current decisions; informing reimbursement or coverage decisions under uncertainty; supporting value-based pricing or managed entry agreements.
Avoid when
When decisions are not data-driven or are dominated by political or ethical considerations; when model uncertainty is poorly characterized; when computational resources are insufficient to implement VoI methods.
Inputs
Probabilistic model inputs, including distributions of parameters reflecting uncertainty; cost of data collection; decision context and outcomes of interest.
Outputs
Expected value of perfect information (EVPI), expected value of sample information (EVSI), expected net benefit of sampling (ENBS), and related metrics that quantify the value of reducing uncertainty.
How it works
Value of Information (VoI) analysis is a decision-theoretic framework that quantifies the expected value of reducing uncertainty in model parameters through further data collection. It evaluates the sensitivity of health policy models to different sources of uncertainty and supports prioritization of future research by estimating the expected net benefit of perfect or partial information. Implemented within probabilistic models, VoI includes expected value of perfect information (EVPI) and expected value of sample information (EVSI), commonly used in health technology assessment and economic evaluation.
- HTA domains
- Costs & Economic Evaluation
- Categories
- Cost-effectiveness Modelling
- Assumptions
- Decisions are made under uncertainty; additional data can reduce uncertainty; the decision problem is well-defined with quantifiable outcomes; model structure and inputs adequately represent the real-world context.
- Strengths
- Provides a formal, quantitative basis for research prioritization; integrates naturally with decision models; identifies key drivers of uncertainty; supports efficient allocation of research resources.
- Limitations
- Computationally intensive, especially for complex models; relies on accurate specification of model structure and parameter distributions; interpretation can be challenging for non-experts; may not capture all real-world decision complexities.
- Also known as
- VoI analysis, Value of Information analysis, Expected Value of Information
Questions this answers
- › How much could better data improve a health policy decision?
- › Which uncertainties in a model matter most for decision-making?
- › Is further research on a health intervention likely to be worthwhile?
- › What is the expected benefit of reducing uncertainty in specific parameters?
- › How should research funding be prioritized across health technologies?
- › When is a decision robust enough to proceed without more data?
References & sources
Similar by meaning
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.