Handling uncertainty: a plea to systematically assess the real-world cost-effectiveness of expensive outpatient medicines.

Summary Objective: Outpatient medications with high expenditures are not structurally monitored on their realworld cost-effectiveness at a post-reimbursement moment in the Netherlands. Concurrently, an increased uncertainty is observed on key reimbursement criteria at the time of decision making. Aim of the study was to investigate the feasibility and desirability of a periodic re-assessment procedure for reimbursed expensive outpatient medicines in the Netherlands. Methods: Imatinib, pegfilgrastim, and adalimumab were analysed as case studies once at the time of reimbursement decision making (labelled t=0) and once at present in 2011 (t=1). For t=0, grey literature was obtained from the website of the Dutch reimbursement agency. For t=1, a systematic literature review of economic evaluation studies was performed in the databases of MEDLINE via PubMed; the Cochrane Library; the Centre for Reviews and Dissemination, including the NHS Economic Evaluation Database; and the Cost-Effectiveness Analysis Register (www.cearegistry.com) from inception to May 2011. Bibliographies of related articles were assessed and the three marketing authorisation holders contacted directly to identify additional (un-)published studies. Based on the Dutch pharmacoeconomic guidelines, a data-collection form was compiled. Numerical scores were assigned to individual studies on their relevance and quality for the Dutch setting. Incremental costeffectiveness ratios were converted and inflated to 2010 Euros (€), the base year of costs in this study. Results: At t=0, uncertainty prevailed for the criteria effectiveness and quality of life in 5 of the 8 reimbursement advices available. In the extreme, the absence of appropriate data rendered the assessment of a reimbursement criterion impossible. Concurrently, economic considerations played no visible role in the final reimbursement decision. At t=1, 49 full economic evaluation studies qualified for inclusion and appraisal. The amount and timeliness of studies suggest that a re-assessment of drugs after ≤4 years would be very possible. They were of fair relevance to be considered for the Dutch setting. Reimbursement seemed, ceteris paribus, cost-effective in all officially assessed indications. Use in other (unregistered) conditions or therapy lines were more often regarded as cost-ineffective. Four indications in imatinib had neither reimbursement advice nor published economic evaluations.Results: At t=0, uncertainty prevailed for the criteria effectiveness and quality of life in 5 of the 8 reimbursement advices available. In the extreme, the absence of appropriate data rendered the assessment of a reimbursement criterion impossible. Concurrently, economic considerations played no visible role in the final reimbursement decision. At t=1, 49 full economic evaluation studies qualified for inclusion and appraisal. The amount and timeliness of studies suggest that a re-assessment of drugs after ≤4 years would be very possible. They were of fair relevance to be considered for the Dutch setting. Reimbursement seemed, ceteris paribus, cost-effective in all officially assessed indications. Use in other (unregistered) conditions or therapy lines were more often regarded as cost-ineffective. Four indications in imatinib had neither reimbursement advice nor published economic evaluations. Discussion: Today’s dichotomous reimbursement system is insufficient to meet the changed needs of decision making under uncertainty. The cost-effectiveness review of international economic evaluations was based on second-best data with validity concerns. Publication and funding biases cannot be excluded. Using international studies with heterogeneous designs, methods, resource utilisation, and costs bears inherent limitations. Outcomes can merely suggest improvements without optimising healthcare resource allocation for the Netherlands. A formalised re-assessment procedure for expensive outpatient medicines with real-life cost-effectiveness data seems desirable. The national reimbursement authority should demand outcomes research from manufacturers for all reimbursed expensive medications. Considerations to be taken into account include, inter alia, a transparent process, clearly defined criteria and definitions, stakeholder involvement, and political will to enforce the possible consequences of a review, which consist of continuous or conditional reimbursement, drug price alterations, or delisting.