Cost-effectiveness of varicella and zoster vaccination in England & Wales: importance measures for correlated parameters

Abstract

This study aims to explore irregular relationships between input and outcomes in cost-effectiveness analysis, accounting for interdependencies between many input parameters. We used a dynamic infectious disease transmission model investigating the cost-effectiveness of varicella and zoster vaccination as a case study. Incremental costs and effects were used separately as the outcomes of interest, rather than the ratio between these, and a single R² was calculated for groups of interdependent input parameters. Partial R² was obtained as measure for the marginal proportion of variance explained. The most important input parameters were related to zoster epidemiology and disease transmission. In case of linear relationships, R² is an easy to obtain importance measure, also in the presence of interdependent input parameters.