OBJECTIVES: With the draft EU regulation from 2018 for a European Health Technology Assessment (EUHTA), the voluntary HTA co-operation within the framework of the European Network for Health Technology (EUnetHTA) shall become mandatory within the scope of a clinical joint assessment. As the EUHTA will largely rest on the evidence that is already assessed by EMA the question arises what additional insights will be brought forward by this new process.

METHODS: This analysis compares the key findings from EMA using the EPAR and from EUnetHTA using the respective assessment report. Basis for this analysis are the three most recent EUnetHTA assessments which have also been assessed by G-BA: Alectinib (Firstline NSCLC), Midostaurin (AML), Ramucirumab (Gastric cancer). Respective analyses are compared using the following criteria: relevant evidence, outcomes per category, overall conclusion.

RESULTS: EMA looks for direct evidence as well as supportive (direct) evidence, i.e. all available clinical studies on a substance in a given indication. EUnetHTA on the other hand uses the same direct evidence without supportive evidence. Instead EUnetHTA also prepares indirect comparisons and network meta analyses to draw conclusions over further relevant therapies. However it remains skeptical on their relevance due to the heterogeneity of the comparisons.

Outcomes and actual data presented for direct evidence is identical. The same holds true for the general conclusions based on the direct evidence.

CONCLUSIONS: Based on the three assessments studied in detail it becomes clear that EUnetHTA basically repeat data already presented by EMA. Indirect comparisons provide additional aspects; however relevance of this information remains questionable due to severe methodological shortcomings. Hence the additional insights provided by EUnetHTA seem rather limited. That is why it remains unclear as to (1) why EUnetHTA duplicates the data already presented in the EPAR and (2) if it would be useful to incorporate the NMAs in the approval process.