Integrating Dynamic Pricing Into Cost-Effectiveness Models: Implications for US-Based Drug Evaluations
Author(s)
Dominique Seo, MPH1, Alice Kate G. Cummings Joyner, BA, MS2, Warren Stevens, BA, MSc, PhD2, Richard Chapman, MS, PhD1, Larragem Raines, MS1;
1Center for Innovation & Value Research, Alexandria, VA, USA, 2Medicus Economics, Boston, MA, USA
1Center for Innovation & Value Research, Alexandria, VA, USA, 2Medicus Economics, Boston, MA, USA
Presentation Documents
OBJECTIVES: Incorporating dynamic pricing into cost-effectiveness models is crucial to accurately capture changes in drug prices over time. Building on a previously published open-source value model for major depressive disorder (MDD), we added a functional component to allow the price of a ‘new’ placeholder drug to vary annually following launch.
METHODS: We calculated incremental cost-effectiveness ratios (ICER) for 5- and 10-year time horizons for a hypothetical new therapy across four lines of treatment under 2 different scenarios, 1) assuming static pricing and 2) dynamic pricing, each compared to standard progression through four distinct treatments with static prices: (SSRI, SNRI, SNRI + atypical antidepressant, and SNRI + antipsychotic). Input parameters (other than drug costs) were based on predefined values and applied uniformly across scenarios. Assumed price for the hypothetical new drug was $14,000 per annum in the static scenario, with assumed discounts of 32% in year 2 increasing to 77% by years 8+ in the dynamic scenario.
RESULTS: Dynamic pricing substantially reduced the ICER for the new therapy compared to the standard basket, with ICER for the new therapy dropping from $221,000 using static pricing to $124,000/QALY using dynamic pricing over a 5-year time horizon, falling below a $150,000 willingness-to-pay threshold. For the 10-year time horizon, dynamic pricing lowered the ICER from $199,000 to $82,000/QALY.
CONCLUSIONS: Dynamic pricing can significantly influence the cost-effectiveness of new therapies by reducing treatment costs over time, to reflect genericization of drugs. This shift highlights the importance of incorporating dynamic pricing into cost-effectiveness analyses to better reflect real-world conditions. While pricing is a straightforward variable to model dynamically, including price trajectory functions over time adds another layer of uncertainty. Researchers may also consider extending dynamic modeling to other inputs, such as effectiveness measures, that are usually considered to be static but may change over time.
METHODS: We calculated incremental cost-effectiveness ratios (ICER) for 5- and 10-year time horizons for a hypothetical new therapy across four lines of treatment under 2 different scenarios, 1) assuming static pricing and 2) dynamic pricing, each compared to standard progression through four distinct treatments with static prices: (SSRI, SNRI, SNRI + atypical antidepressant, and SNRI + antipsychotic). Input parameters (other than drug costs) were based on predefined values and applied uniformly across scenarios. Assumed price for the hypothetical new drug was $14,000 per annum in the static scenario, with assumed discounts of 32% in year 2 increasing to 77% by years 8+ in the dynamic scenario.
RESULTS: Dynamic pricing substantially reduced the ICER for the new therapy compared to the standard basket, with ICER for the new therapy dropping from $221,000 using static pricing to $124,000/QALY using dynamic pricing over a 5-year time horizon, falling below a $150,000 willingness-to-pay threshold. For the 10-year time horizon, dynamic pricing lowered the ICER from $199,000 to $82,000/QALY.
CONCLUSIONS: Dynamic pricing can significantly influence the cost-effectiveness of new therapies by reducing treatment costs over time, to reflect genericization of drugs. This shift highlights the importance of incorporating dynamic pricing into cost-effectiveness analyses to better reflect real-world conditions. While pricing is a straightforward variable to model dynamically, including price trajectory functions over time adds another layer of uncertainty. Researchers may also consider extending dynamic modeling to other inputs, such as effectiveness measures, that are usually considered to be static but may change over time.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
EE447
Topic
Economic Evaluation
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, SDC: Mental Health (including addition)