Advertisement

INTERACTION EFFECTS IN HEALTH STATE VALUATION STUDIES: AN OPTIMAL SCALING APPROACH

  • Marcel F. Jonker
    Correspondence
    Corresponding author: Dr. Marcel F. Jonker, Erasmus University Rotterdam, PO Box 1738, 3000DR Rotterdam, The Netherlands
    Affiliations
    Erasmus Choice Modelling Centre, Erasmus University Rotterdam, The Netherlands

    Erasmus School of Health Policy & Management, Erasmus University Rotterdam, The Netherlands
    Search for articles by this author
  • Bas Donkers
    Affiliations
    Erasmus Choice Modelling Centre, Erasmus University Rotterdam, The Netherlands

    Erasmus School of Economics, Erasmus University Rotterdam, The Netherlands
    Search for articles by this author
Open AccessPublished:October 30, 2022DOI:https://doi.org/10.1016/j.jval.2022.10.008
      This paper is only available as a PDF. To read, Please Download here.

      OBJECTIVE

      To introduce a parsimonious modelling approach that enables the estimation of interaction effects in health state valuation studies.

      METHODS

      Instead of supplementing a main effects model with interactions between each and every level, a more parsimonious optimal scaling approach is proposed. This approach is based on the mapping of health-state levels onto domain-specific continuous scales. The attractiveness of health states is then determined by the importance-weighted optimal scales (i.e. main effects) and the interactions between these domain-specific scales (i.e. interaction effects). The number of interaction terms only depends on the number of health domains. As a result, interactions between dimensions can be included with only a few additional parameters.

      EMPIRICAL APPLICATIONS

      The proposed models with and without interactions are fitted on three valuation datasets from two different countries, i.e. a Dutch latent-scale discrete choice experiment (DCE) dataset with N=3,699 respondents, an Australian time-trade-off (TTO) dataset with N=400 respondents, and a Dutch DCE with duration dataset with N=788 respondents.

      RESULTS

      Important interactions between health domains were found in all three applications. The results confirm that the accumulation of health problems within health states has a decreasing marginal effect on health state values. A similar effect is obtained when so-called N3 or N5 terms are included in the model specification, but the inclusion of two-way interactions provides superior model fits.

      CONCLUSIONS

      The proposed interaction model is parsimonious, produces estimates that are straightforward to interpret, and accommodates the estimation of interaction effects in health state valuation studies with realistic sample size requirements. Not accounting for interactions is shown to result in profoundly biased value sets, particularly in stand-alone DCE with duration studies.

      Keywords