OBJECTIVES: Budget Impact Analysis (BIA) is required in the French Health Technology Assessment framework for products with high projected sales revenues. Duration of treatment (DoT) in oncology is a key input for BIA. The objective is to explore methods to consider DoT in BIA for nivolumab, in second line advanced renal cell carcinoma (2L RCC).

METHODS: Following French guidelines, a prospective BIA was conducted for a 3-year time horizon comparing two scenarios: market with and without the introduction of nivolumab. Three methods were explored to implement DoT for three annual incident cohorts (IC): The ‘Overall Mean’ method (OMM) implemented a 3-year mean DoT to each IC. The ‘Incremental Means’ method (IMM) applied mean DoT of 0-3, 0-2 and 0-1 year intervals, to the 1^st, 2^nd and 3^rd IC respectively. The ‘Restricted Means’ method (RMM) applied restricted means for years 1, 2 and 3 for the 1^st IC, years 1 and 2 for the 2^nd IC, and year 1 for the 3^rd IC.

RESULTS: Both OMM and IMM consider total cost for each IC in their year of entrance in the model, thus not spreading costs through time horizon. OMM considered treatment duration for 2^nd and 3^rd IC beyond the selected time horizon, consequently overestimating drug acquisition costs (DAC) and budget impact of nivolumab. With IMM, DAC were overestimated in Year 1 (Y1) but underestimated in Year 3 (Y3): -25% in Y3 vs. Y1. RMM enabled spreading treatment duration through the time horizon, thus increasing DAC from one year to the next (+14% in Y3 vs. Y1). IMM and RMM resulted in identical budget impacts for the 3-year horizon, but inferior to that of OMM.

CONCLUSIONS: RMM appeared to consider costs within the selected time horizon and to correctly distribute them across the years. RMM was favorably evaluated through the French HTA in 2L RCC.