An Efficient Calculation Method for the Expected Value of Sample Information: Can we do it? Yes, we can

Abstract

The Expected Value of Sample Information (EVSI) allows us to quantify the economic benefit of a potential future trial or study. While this is clearly of benefit, especially when considering which trials should be funded, it has rarely been used in practise due to the immense computational power required to calculate the EVSI using traditional methods. Therefore, recently a large number of methods have been designed to reduce the computational burden required to calculate the EVSI and therefore allow practitioners to use it when designing trials and applying for funding. However, these methods are quite technical and are generally simply presented in theoretical terms. Therefore, this paper describes and demonstrates one of these recent methods using a simple toy example and realistic health economic model to aid practitioners to put this method into practise. It also investigates the practical application of this method and makes recommendations for the practitioners to achieve accurate EVSI estimates.

Gianluca Baio
Gianluca Baio
Professor of Statistics and Health Economics