Prior Elicitation as input to Phase 3 PTS Calculation

April 18, 2024

Speaker: Niki Arya (AstraZeneca)

Abstract:

We utilized expert elicitation to determine the prior assumptions needed to calculate the probability of technical success (PTS) for a planned large outcome study. The primary endpoint for the large outcome study is a composite endpoint with subcomponents that are major events (deaths, hospitalizations, etc.). A presentation was put together summarizing the results from previous trials with our drug and other drugs within the same medication class. In these trials, a similar composite endpoint and the individual subcomponents were analyzed in slightly different populations to help shape prior assumptions about the effects of our drug on the primary endpoint proposed in our large outcome study. This presentation was shared with a panel of 5 experts in this field who then used these results to predict the expected treatment effect on the primary composite endpoint and each of its subcomponents in the population proposed for the planned large outcome study. These predictions were then pooled together and weighted appropriately based on the expertise of the panelist to form a prior distribution which was then used to determine the PTS for the planned large outcome study. In the presentation for the Society of Decision Professionals conference, we will go into the details of the expert elicitation process, how each expert’s predictions were incorporated into the prior assumptions, and how the prior assumptions were used to determine the PTS.