Methodological quality of clinical trials evaluated in early benefit assessment: orphan versus non-orphan drugs

Tags: Poster
Preukschat D1, Eyding D1
1MDS e.V. (Medical Advisory Service of Social Health Insurance), Germany

Background: After approval, the public decision maker, the Federal Joint Committee (G-BA), conducts an Early Benefit Assessment (EBA) on every new drug on the market in Germany to determine the certainty and degree of the new drug’s added benefit, which then influences the drug’s reimbursement.

By law, an added benefit is assumed for orphan drugs as 'certain' by approval. The EBA remains to determine its degree based on the approval studies, whereas non-orphans need to show their added benefit (certainty and degree) over standard therapy defined by the G-BA.

Concerns have arisen that the orphans’ added benefit might not be founded well enough.

Objectives: To compare methodological quality of clinical trials evaluated in EBAs of orphan versus non-orphan drugs.

Methods: All EBAs completed up to June 2015 were included. The G-BA website was used to obtain all relevant documents (EBAs and G-BA’s decisions). Methodological aspects were extracted (e.g. comparator, treatment allocation, blinding, risk of bias) and contrasted with the certainty and degree of the added benefit.

Results: EBAs for orphan drugs were based on lower quality evidence than non-orphan EBAs. Orphan trials were less likely to be double-blind or randomized and were assessed more often as having a high risk of bias. Nonetheless, all orphans – by law – had a 'certain' added benefit, whereas non-orphan drugs rarely were classified with 'certain' added benefit but in nearly half of the cases as 'hint' or 'indication', versus nearly half with 'no added benefit'. Regarding the degree of the added benefit the classification was indicative of the underlying evidence’s lower quality: a high proportion of orphans was classified as 'not quantifiable', whereas less than 10% of the non-orphans were regarded as “not quantifiable”.

Conclusion: A 'certain' added benefit for orphans in all cases appears not to be justified by the lower quality evidence for them compared to the non-orphans. Incentives for orphans should not be based on lowered methodological standards.