Bayesian dose ‐finding phase I trial design incorporating historical data from a preceding trial

In this study, we propose to adaptively incorporate historical data into prior distributions assumed in a new dose‐finding trial. Our proposed approach aims to appropriately borrow strength from a previous trial to improve the maximum tolerated dose determination in another patient population. We define a “historical‐to‐current (H‐C)” parameter representing the degree of borrowing based on a retrospective analysis of previous trial data. In simulation studies, we examine the operating characteristics of the proposed method in comparison with 3 alternative approaches and assess how the H‐C parameter functions across a variety of realistic settings.
Source: Pharmaceutical Statistics - Category: Statistics Authors: Tags: SPECIAL ISSUE PAPER Source Type: research