Do Advanced Methods Mean More Uncertainty? Quantifying Uncertainty Induced by Using Advanced Modelling Methods
Koen Degeling, PhD, MSc, BSc, Lumen Value & Access – a Healthcare Consultancy Group Company, New York City, NY, USA, Elisabeth Anne Louise Fenwick, PhD, OPEN Health Evidence & Access, Oxford, OXF, UK and Erik Koffijberg, PhD, MSc, Health Technology and Services Research Department, University of Twente, Enschede, OV, Netherlands
Purpose: This breakout session will create awareness of the different sources of uncertainty in health economic models that utilize advanced methods such as advanced survival models, individual-level simulations, and prediction models, and provide guidance on how that uncertainty can be quantified and propagated appropriately in analyses to ensure appropriate reflection of uncertainty in outcomes. Description: Increasingly sophisticated methods are used to populate and analyze health economic models to appropriately represent clinical pathways and outcomes. Appropriately quantifying the uncertainty in these analyses is essential to inform decision making, however this is not as straightforward as it is for more conventional modeling methods. Dr Degeling will introduce the session and provide a refresher on the different types of uncertainty and their importance in health economic modeling through an interactive example (8 minutes). Dr Fenwick will discuss challenges around quantifying uncertainty in advanced survival models, such as mixture cure models (15 minutes). Dr Degeling will discuss how the different types of uncertainty apply to individual-level simulation models and provide guidance on approaches for dealing with stochastic and parameter uncertainty (15 minutes). Prof Koffijberg will discuss challenges regarding the use of prediction models in economic evaluations and the different sources of uncertainty they introduce (15 minutes). The audience will be able to ask questions after each presentation (included in presentation time), and the session will end with another interactive example along with voting questions for the audience, to evaluate whether their understanding on the topic has improved (7 minutes). Those who (aim to) develop or appraise health economic models will benefit from attending this session.
Conference/Value in Health Info
2022-11, ISPOR Europe 2022, Vienna, Austria