What matters to society when making decisions about reimbursing drugs for prostate cancer: a discrete choice experiment.

[INTRODUCTION] The use of surrogate outcomes to support treatment efficacy and cost-effectiveness in localized prostate cancer has the potential to shorten time to access for new medicines, but it may be associated with greater uncertainty. We sought to understand how members of society might trade off drug efficacy, safety, cost-effectiveness, and evidence strength when asked to consider funding new treatments for localized prostate cancer.

[METHODS] A discrete choice experiment among the Australian general population was conducted. Treatment choices were described by cost-effectiveness, government cost, adverse events, patients expected to benefit from treatment, nature of benefits, completeness of data available, and the extent of follow-up. Responses were analyzed using mixed logit and latent class models to assess heterogeneity.

[RESULTS] The survey was completed by 1003 participants. Respondents favored interventions that benefited a high proportion of patients (OR = 2.44, 95% CI = 2.10 to 2.69), were cost-effective (OR = 0.89, 95% CI = 0.82 to 0.98), low cost to government (OR = 0.75, 95% CI = 0.68 to 0.82), and with long follow-up time (OR = 1.25, 95% CI = 1.12 to 1.39). Respondents showed aversion to ongoing studies (OR = 0.81, 95% CI = 0.75 to 0.88). Respondents were willing to accept less certain evidence if associated with a lower cost-effectiveness ratio (ICER). Latent class analysis revealed 2 subgroups with different preference patterns.

[CONCLUSIONS] The analysis found preference for interventions with maximum benefits while minimizing costs. Respondents preferred studies with longer follow-up time for reimbursement decisions despite consequent treatment delays. Notably, respondents were willing to accept the uncertainty associated with surrogate outcomes-reflected through study length and completeness of evidence-provided the intervention demonstrated a favorable ICER.