No abstract

No abstract

No abstract

In 2011-12, the cost-effectiveness of imatinib, dasatinib and nilotinib for 1 st -line chronic myeloid leukaemia in the UK was evaluated by NICE. We discuss three methodological issues which strongly influence the estimated cost-effectiveness of these drugs. These issues are also important for the cost-effectiveness of many other drugs and medical devices. METHODS: We discuss the pros and cons of the following competing methods. 1) Estimation of overall survival: Method A: estimated as the cumulative duration of 1 st -, 2 nd -and 3 rd -lines of treatments. Method B: estimated from the surrogate responses: complete cytogenetic response and major molecular response; 2) Cost-effectiveness of subsequent treatments: the cost-effectiveness of 1 st -line drugs are substantially affected by the cost-effectiveness of subsequent drugs. Method A: traditional method of modelling estimated costs and QALYs of subsequent drugs. Alternatively, minimise impact of costeffectiveness of subsequent treatments by either Method B: setting per patient costs and QALYs of subsequent treatments equal between treatment arms, or Method C: cap the cost-effectiveness ratio whilst on subsequent treatments at the willingness to pay threshold; 3) Future drug prices: This is an important issue given that the patent for imatinib will expire soon, in 2016, after which its price may fall substantially. Method A: use the current list prices of all drugs in the future, as required by NICE. Method B: assume constant drug prices until patent expiry, at which time assume a fixed price cut. Assuming a modest 25% price cut on patent expiry, the ICER for nilotinib vs. imatinib increases substantially, from £36,000 to £54,000 per QALY. RESULTS: The pros and cons of the various methods are discussed. CONCLUSIONS: This study informs important methodological issues which apply to many health technologies. The study ultimately contributes to more accurate assessments of the cost-effectiveness of health technologies, and hence whether a given technology should be publicly-funded.

No abstract