A Framework for Probabilistic Assessment of New Medical Technologies

摘要

In this paper, we present a framework for early assessment of new medical technologies based on probability and systems analysis. Its purpose is to better inform decision-making during the design and development stages of a new device, where evidence is limited and classical statistical methods are not yet applicable. Decisions made by device manufacturers during this period have significant implications downstream, as they affect not only future system performance, but also the likelihood of obtaining regulatory approval and of successful product commercialization. Furthermore, industry could save a significant amount of money by identifying early those technologies that may not work. We define first a classification of potential data sources for Bayesian analysis and suggest a method for aggregation of prior distributions. Next, we present a model that captures key variables and relationships and that evolves with the design and development of the device. We illustrate this model with the example of a new cardiology catheter for minimally-invasive surgical treatment.