Cost-effectiveness analysis using data from multinational trials: the use of bivariate hierarchical modeling

Area of Science:

• Health Economics
• Biostatistics
• Clinical Trial Analysis

Background:

• Health care cost-effectiveness analysis (CEA) commonly utilizes individual patient data (IPD) from multinational randomized controlled trials.
• Existing CEA methods adequately address between-patient variability but neglect crucial between-location variability.
• Differences in healthcare resources, costs, clinical practices, and patient populations across countries can significantly impact CEA results.

Purpose of the Study:

• To advocate for and illustrate the application of Bayesian bivariate hierarchical modeling for analyzing multinational cost-effectiveness data.
• To demonstrate how this framework can yield more appropriate estimates of overall cost-effectiveness and associated uncertainty.
• To enable the derivation of country-specific cost-effectiveness estimates, accounting for between-location variability and controlling for patient- and country-specific factors.

Main Methods:

• Bayesian bivariate hierarchical modeling was employed to analyze multinational cost-effectiveness data.
• This approach explicitly models patient-level costs and outcomes nested within countries.
• The methodology was applied to real-life data from a multinational trial involving 17 countries.

Main Results:

• Standard CEA methods applied to multinational IPD showed significant variability across countries, potentially leading to misleading conclusions.
• The proposed Bayesian modeling approach produced more appropriate overall cost-effectiveness estimates and quantified sampling uncertainty effectively.
• "Shrinkage estimates" from the model facilitated appropriate quantification of country-specific cost-effectiveness, balancing information across countries.

Conclusions:

• Bayesian bivariate hierarchical modeling offers a robust framework for analyzing multinational cost-effectiveness data.
• This approach provides more accurate overall and country-specific cost-effectiveness estimates compared to standard methods.
• The methods presented offer a generalizable framework for analyzing economic data collected across diverse geographical locations.