WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT • The clinical evidence level of orphan medicines in European market authorization submissions is low: few randomized controlled trials, short treatment follow‐up, rarely hard clinical endpoints. • The success rate at centralized market authorization is lower for orphan drugs (62.9%) than for non‐orphan medicines (70.7%). • The burden of disease of rare diseases is high, affecting the lives of at least 30 million patients in the European Union (EU). WHAT THIS STUDY ADDS • As opposed to market authorization, orphan drugs gain reimbursement more easily than non‐orphan innovative drugs. • Lower quality of evidence of clinical efficacy and safety, more uncertainty on cost‐effectiveness and higher product prices are accepted for orphan drugs. • There is a need for collaboration between the European Commission, competent for market authorization, and the EU member states, competent for reimbursement, in assessing the therapeutic risks and benefits of orphan drugs, to reduce the evidence gap post marketing. AIM We analysed the Belgian reimbursement decisions of orphan drugs as compared with those of innovative drugs for more common but equally severe diseases, with special emphasis on the quality of clinical evidence. METHODS Using the National Health Insurance Agency administrative database, we evaluated all submitted orphan drug files between 2002 and 2007. A quality analysis of the clinical evidence in the orphan reimbursement files was performed. The evaluation reports of the French ‘Haute Autorité de Santé’, including the five‐point scale parameter ‘Service Médical Rendu (SMR), were examined to compare disease severity. Chi‐squared tests (at P < 0.05 significance level) were used to compare the outcome of the reimbursement decisions between orphan and non‐orphan innovative medicines. RESULTS Twenty‐five files of orphan drugs and 117 files of non‐orphan drugs were evaluated. Twenty‐two of 25 (88%) submissions of orphan drugs were granted reimbursement as opposed to 74 of the 117 (63%) non‐orphan innovative medicines (P= 0.02). Only 52% of the 25 orphan drug files included a randomized controlled trial as opposed to 84% in a random control sample of 25 non‐orphan innovative submissions (P < 0.01). The duration of drug exposure was in most cases far too short in relation to the natural history of the disease. CONCLUSIONS Orphan drug designation predicts reimbursement despite poor quality of clinical evidence. The evidence gap at market authorization should be reduced by post‐marketing programmes, in which the centralized regulatory and the local reimbursement authorities collaborate in an efficient way across the European Union member states.

The European Transparency Directive requires that pricing and reimbursement decisions must be taken in a transparent, objective and verifiable way with respect of strict timelines. The Belgian competent authority integrated on January 1st, 2002 Evidence-Based Medicine (EBM) principles in the reimbursement evaluation. The present work describes the procedures and investigates whether the introduction of the EBM principles indeed affects the decision and whether it compromised the respect of strict timelines. The reimbursement decision for all new submissions except for generic drugs is preceded by an evaluation of the relative therapeutic value. There were 1285 submissions handled within the period 2002-2004 of which 159 (12.4%) related to new molecular entities. For the 824 files with valuation of the therapeutic value, the reimbursement decision was positive in 80.8% of cases. The percentage of positive decisions was dependent on the type of submission with the lowest percentages for new molecular entities and submissions for new indications (64%-71%). Line extensions and generics received a positive decision in nearly all cases (> 95%). Proof of added value by at least 1 positive superiority trial against active comparator is a requirement for obtaining a price premium: this was granted in less than 50% of the 67 submissions claiming such superiority and the odds for a negative reimbursement decision increased significantly if the applicant failed to prove added value: O.R. = 9.1 (2.3 - 35.6), indicating that clinical evidence of added therapeutic value clearly facilitates reimbursement. The introduction of the new procedure did not jeopardize the timelines. Introducing EBM principles had a significant impact on reimbursement decisions in Belgium by facilitating reimbursement with a price premium of new drugs with added value addressing unmet medical needs.

Our findings reveal that the three similar EU MS under study agree on the REA-outcome of only half of the new drugs. Differences in applied methodology (e.g. inferences from study results, handling of uncertainty) between MS may exist. We suggest that a joint REA approach would benefit from a common understanding and application of the scientific assessment process using shared guidelines.