PNS328 REAL OPTIONS ANALYSIS FOR MORE OPTIMALLY INFORMED REIMBURSEMENT DECISIONS

Abstract

Informing reimbursement decisions is associated with two major issues: i. It lacks a united assessment of cost-effectiveness and budget impact (BI); and ii. it lacks incorporation of timing of decisions based on the level of uncertainty. We aim to develop a real options analysis (ROA) based method that amends these issues and therefore allows for more optimal reimbursement decision-making. To achieve this, we use a Dutch perspective and an oncology case study. Net Monetary Benefit (NMB) is the main outcome and is calculated as: NMB = ((WTP – ICER) * incremental effectiveness) * (BI / treatment cost per patient). Nivolumab was selected as case-study. Data on the ICER was derived from the reimbursement dossier whilst BI data was generated using a validated BI prediction model. For WTP, three methods for the influence of BI on WTP were used. For ROA implementation, we assumed that the true BI could be observed after one month. We compared traditional ‘now or never’ decisions to the option to wait for more data. For some scenarios, waiting for 10 months of data was the optimal decision as risk due to uncertainty in the first 10 months outweighed immediate benefit (NMB). The different methods describing the relationship between WTP and BI had great influence on NMB (- €42 million for a fixed WTP vs €69 million for a WTP method based on a real-world Dutch reimbursement decision). We unified assessment of cost-effectiveness and BI by means of NMB and then incorporated timing using ROA. We demonstrated that our ROA based method can be used to inform on the timing of reimbursement decisions. ROA could therefore be suitable for providing early guidance on flexible and adaptive reimbursement decisions, deemed essential in the current landscape of ever higher uncertainty at market access of new costly drugs.