Accounting for the Drug Life Cycle and Future Drug Prices in Cost-Effectiveness Analysis

Abstract

Economic evaluations of health technologies typically assume constant real drug prices and model only the cohort of patients currently eligible for treatment. It has recently been suggested that, in the UK, we should assume that real drug prices decrease at 4% per annum and, in New Zealand, that real drug prices decrease at 2% per annum and at patent expiry the drug price falls. It has also recently been suggested that we should model multiple future incident cohorts. In this article, the cost effectiveness of drugs is modelled based on these ideas. Algebraic expressions are developed to capture all costs and benefits over the entire life cycle of a new drug. The lifetime of a new drug in the UK, a key model parameter, is estimated as 33 years, based on the historical lifetime of drugs in England over the last 27 years. Under the proposed methodology, cost effectiveness is calculated for seven new drugs recently appraised in the UK. Cost effectiveness as assessed in the future is also estimated. Whilst the article is framed in mathematics, the findings and recommendations are also explained in non-mathematical language. The 'life-cycle correction factor' is introduced, which is used to convert estimates of cost effectiveness as traditionally calculated into estimates under the proposed methodology. Under the proposed methodology, all seven drugs appear far more cost effective in the UK than published. For example, the incremental cost-effectiveness ratio decreases by 46%, from £61, 900 to £33, 500 per QALY, for cinacalcet versus best supportive care for end-stage renal disease, and by 45%, from £31,100 to £17,000 per QALY, for imatinib versus interferon-α for chronic myeloid leukaemia. Assuming real drug prices decrease over time, the chance that a drug is publicly funded increases over time, and is greater when modelling multiple cohorts than with a single cohort. Using the methodology (compared with traditional methodology) all drugs in the UK and New Zealand are predicted to be more cost effective. It is suggested that the willingness-to-pay threshold should be reduced in the UK and New Zealand. The ranking of cost effectiveness will change with drugs assessed as relatively more cost effective and medical devices and surgical procedures relatively less cost effective than previously thought. The methodology is very simple to implement. It is suggested that the model should be parameterized for other countries.