PURPOSE: Early stage health economic modelling is increasingly promoted to support the process of medical device development in value-based health systems. While early stage health economic models are often used to support market access and pricing, several other methods are available for managing a product portfolio. This workshop will introduce constrained optimization methods and multi-criteria portfolio selection modeling (MCPS) for portfolio management. DESCRIPTION: Medical device development carries high-risks in terms of the technical, procedural, and clinical objectives associated with the medical device. Health care industry therefore needs to carefully optimize their research & development project portfolios in terms of the benefits, risks, and costs. Modelling and optimization are frequently used to explicitly manage these portfolios as well as to rationalize investments in medical device development. Optimization methods are often used by health care systems to identify the optimal allocation of resources across interventions subject to a variety of different types of constraints. Among the most widely used methodologies are linear and stochastic programming, as discussed by the ISPOR Task Force “Optimization Methods in Health Care Delivery” in a recent Value in Health publication. Although less common, MCPS, a form of multi-criteria decision analysis, can also provide a structured approach to manage a product portfolio, but does not explicitly employ constrained optimization algorithms. The present workshop will first introduce the background for portfolio management and optimization of resources in device development (Maarten IJzerman), followed by an illustration of MCPS (Praveen Thokala). The workshop will then introduce constrained optimization methods (Jon Tosh) and demonstrate where constrained optimization and MCPS can be used to identify the optimal allocation of resources over a range of medical products (William Padula). The workshop will conclude with an audience discussion about the opportunities and challenges of using these different methods for portfolio selection.