HERC Seminar - Value of sample information as a tool for clinical trial design

Abstract

The Expected Value of Sample Information (EVSI) uses evidence about the costs and effectiveness of different interventions to calculate the economic value of specific trial designs. Despite the potential benefits of using the EVSI, in terms of finding economically valuable trial designs, practical applications have been limited due to computational difficulties and ease of interpreting the results. Recently, methods have been developed to allow for the efficient computation of the EVSI across different trial designs. This allows researchers to use the EVSI as a tool for clinical trial design. This talk will present a recent method for EVSI calculation that is based on variance calculations and Bayesian updating alongside an R package that has been developed to aid with EVSI calculations.  An example based on the HomeHealth feasibility study will be demonstrated to show how the EVSI can be used in a real world setting. We also demonstrate the capabilities of an R Shiny tool that has been developed for the presentation of the EVSI to key stakeholders.

Anna Heath is a PhD student in the department of Statistical Science at University College London. Her main research interests are in Bayesian statistical modelling for cost-effectiveness analysis and decision making within health-care systems.  Principally, she focuses on the development of fast computation methods for Value of Information measures. She has a keen interest in the practical application of value of information for research prioritisation and clinical trial design. She has also worked as a health economic consultant for pharmaceutical and consultancy companies and is a developer for the R packages BCEA and EVSI.