Societal benefits of surrogate outcomes to support reimbursement decisions: The case of prostate cancer.

372 Background: Incorporating surrogate outcomes for survival into economic evaluations for localized prostate cancer (LPC) has the potential to reduce the time needed to subsidise new therapies but may increase uncertainty regarding true benefit. The ICECaP study demonstrated that metastasis-free survival (MFS) is a strong surrogate of overall survival (OS) in LPC. We sought to model the benefits to society of earlier funding decisions, and to understand the context of societal trade-offs between aspects of drug value and time to reimbursement decision-making in LPC. Methods: Societal benefits were modelled using a 4-health state (biochemical recurrence (BCR), BCR-free, metastasis and death) partitioned survival model to evaluate the cost-effectiveness of abiraterone plus prednisolone (AAP) plus standard of care (SoC) compared to SoC alone in LPC. Survival outcomes were dependent on the ICECaP surrogacy relationship. Treatment pathways and costs reflected Australian practice and quality of life was informed by the literature. The model evaluated the difference between deciding to reimburse AAP now, based on MFS, versus wait for OS 3 years. Societal trade-offs were captured using a population based discrete choice experiment exploring the influence on preference for drug reimbursement of cost-effectiveness, cost to Government, adverse events, proportions expected to benefit, nature of benefits, the completeness of data and the time of follow-up. Preferences were analysed using mixed logit analyses. Results: The analysis of societal benefits showed that using MFS as a surrogate for OS could prevent 117 patients from recurrence, 30 from metastases and save 63 lives in the short-term when AAP is compared with SoC for LPC patients in Australia. Delaying the reimbursement decision for 3 years results in a loss of 4.48 years, 0.47 quality adjusted life years, and increase the cost to $11,558 per patient over the lifetime. The analysis of societal trade-offs reflected results from 1,003 community members and showed an inclination to favour reimbursing interventions that benefit a high proportion of patients (OR=2.10 CI 95%=1.91, 2.32), have a lower cost-effectiveness ratio (OR=0.81 CI 95%= 0.78, 0.88) and an aversion for evidence from ongoing studies (OR= 0.83 CI 95%= 0.78, 0.88). Conclusions: Using MFS as surrogate for funding decision in LPC is likely to produce benefits to society in terms of improved short and long-term outcomes. This contrasts with societal preferences to delay reimbursement decisions to increase certainty. This suggests, improved cost-effectiveness could be used to compensate for the uncertainty arising from the use of surrogate outcomes in reimbursement decisions. The use of evidence-based surrogate outcomes in economic evaluations to support reimbursement decisions can reduce the time it takes new interventions to become available improving the benefits to society.