Bayesian Meta-Analysis of Health State Utility Values: A Tutorial with a Practical Application in Heart Failure

Area of Science:

• Health economics
• Biostatistics
• Pharmacoeconomics

Background:

• Health state utility values are crucial inputs for economic models.
• Multiple, often conflicting, utility values from different studies complicate policy decisions.
• A single study is frequently insufficient to represent the best available utility data.

Purpose of the Study:

• To provide introductory guidance on conducting Bayesian meta-analysis of health state utility values.
• To generate a single parameter input for economic evaluation using R.
• To illustrate the process with a real-world example of heart failure patient utilities.

Main Methods:

• A step-by-step tutorial using R for Bayesian meta-analysis.
• Includes data setup, imputation of missing standard deviations, prior definition, model fitting, convergence diagnostics, result interpretation, and sensitivity analyses.
• Utilizes data from a systematic review of 21 studies on heart failure utilities measured with the EuroQol-5D (EQ-5D).

Main Results:

• Demonstrates the application of Bayesian meta-analysis to pool health state utility values.
• Discusses interpretation of posterior distributions for pooled effect size (mean utility) and between-study heterogeneity.
• Provides a clear workflow for generating reliable utility parameters.

Conclusions:

• Bayesian meta-analysis offers a robust method for synthesizing health state utility values.
• This tutorial aims to encourage the use of Bayesian methods in utility meta-analysis.
• Facilitates informed policy decisions through reliable economic modeling inputs.