A DISCRETE EVENT MODEL FOR ASSESSING THE COST-EFFECTIVENESS OF PRECISION MEDICINE INITIATIVES

A Discrete Event Model for Assessing the Cost-Effectiveness of Precision Medicine Programs

 

 

Purpose: The promise of precision medicine is fueled by extraordinary advances in the discovery of genomic variation that predicts both disease risk and therapeutic response.  However, rapid advances in the science and technology underlying genomic medicine has not yet been met with parallel efforts to improve health economic models to aid patients, payors, clinicians and other stakeholders in understanding the value of personalized medicine. 

Methods: This methodological discussion will describe the development and deployment of a generalized cost-effectiveness framework for precision medicine using discrete event simulation (DES).  

Results: We discuss how our discrete event simulation framework addresses numerous challenges that arise in health economic evaluations of genomic medicine.  First, panel genomic testing is distinct from usual testing practice because it can be highly multiplexed, providing dense biological data that could be used over an individual's lifetime (see Figure).  We demonstrate, using stylized examples based on pharmacogenomic prescribing and dosing of clopidogrel, warfarin, and statins, that traditional cost-effectiveness approaches based on single-gene testing likely underestimate the value of genomic information conveyed by multi-gene assays.  Second, heterogeneity in preferences and responses to testing among physicians and patients has also made it difficult to identify the relevant populations for whom testing may be most valuable.  We demonstrate how our DES model provides a platform for understanding how different approaches to the application of testing – such as general testing of an entire population; targeted, prospective testing based on risk stratification; and reactive testing at the point of indication or diagnosis – affect estimates of cost-effectiveness.   Finally, we discuss methodological innovations that allow for use of a DES framework to investigate the relative importance of factors that influence the implementation of genetic testing in clinical practice.  Specifically, we discuss how value of information (VOI) methods can be used to facilitate investigations of evidentiary uncertainties, behavioral frictions, and other factors that explain sub-optimal implementation of genomic testing. 

Conclusions:  Discrete-event simulation provides a powerful platform from which to evaluate the cost-effectiveness and other implications for the design and deployment of precision medicine initiatives.