What Value Do Dutch Citizens Place on Health Interventions That Provide Greater Health Gains to Lower-Income Groups? A Discrete Choice Experiment

Background: Reimbursement decisions for new health interventions focus on maximizing health gains, with limited attention to who benefits from these gains or the impact on income related health inequalities. This study aimed to examine the preferences of Dutch citizens regarding the distribution of health gains of new interventions across income groups. Methods: A discrete choice experiment (DCE) was completed by 614 Dutch adults. Respondents were presented with 12 choice tasks. In each choice task, they were asked to choose between two health interventions that differed on the following attributes: total healthy life years gained, distribution of healthy life years gained across income groups, additional costs in terms of health insurance premium increases and whether the intervention was curative or preventive. Preferences were estimated using multinomial logit (MNL) models, relative attribute importance, willingness-to-pay, and willingness-to-trade total health gains. Preference heterogeneity was examined using latent class (LC) analyses. Results: Respondents found the distribution of health gains by income the most important attribute in their decision between health interventions (relative importance [RI] = 40.5%, 95% CI: 38.3%–42.7%). Overall, respondents preferred an equal distribution of healthy life years gained across income groups (βhigher income groups = -1.427, 95% CI: -1.547–-1.307; βlower-income groups = -0.315, 95% CI: -0.395–-0.235). A health intervention should yield 14 283 (95% CI: 10 463–18, 102) additional healthy life years or reduce the yearly health insurance premium by €39.96 (95% CI: €29.03–€50.89) if it mainly favors lower-income groups. Preventive interventions were generally preferred over equally effective or more effective curative interventions (βprevention = 0.270, 95% CI: 0.204–0.336). While preferences displayed a similar direction across LCs, the classes differed in the RI assigned to the attributes. Conclusion: Our findings suggest societal support for interventions that prioritize preventive programs over equally effective or more effective curative interventions and prioritize interventions that provide equal benefits across different income groups.