OBJECTIVES: To compare two cost-effectiveness evaluations submitted independently by two pharmaceutical compagnies for the same treatment combination in first-line renal cell carcinoma and discuss the impact of different methodological choices on conclusion by the French Health Authority (HAS).

METHODS: The two submitted models were compared in terms of structuring choices (objective, population, interventions, time horizon and perspective), modelling choices and main assumptions (clinical data and indirect treatment comparison (ITC), utility and costs), results and sensitivity analyses. HAS’s critical opinion and conclusion on cost-effectiveness were compared and discussed.

RESULTS: The HAS expressed 7 important methodological concerns in one dossier and 5 in the other one. Of these, 4 were identical in the two dossiers and were related to similar choices in both models (on model structure, comparators, global survival data and external validation). Differences were observed on 3 other methodological items (post-progression treatments, adverse events, disutilies). While one dossier concluded with an incremental-cost-effectiveness ratio (ICER) that was invalidated by the HAS due to major global uncertainty, in the other dossier the combination was dominated and the HAS retained this inefficiency although surrounded by an important uncertainty. Beyond methodological choices, it seems that differences in HAS’s conclusions can be related to implemented data impacting the level of uncertainty. Particularly in this example, it seems that the use of more mature clinical data and the availability of individual patient data for one comparator in the ITC may have mitigated the uncertainty associated with other modelling choices.

CONCLUSIONS: This example shows the issues associated with the submission of distinct cost-effectiveness dossiers in case of treatments association. It raises the interest of submitting a common dossier to facilitate decisions of public authorities in access to therapeutic innovations.