|
Abstract:
|
Choosing the correct dose in early clinical development is critical to later clinical development. But small sample sizes make it difficult to accurately select the correct dose without exposing patients to a dangerous dose. Historical data can augment sample sizes to increase precision in dose escalation. In our case, we utilize a hierarchical model to augment a Bayesian interval dose-escalation design, mTPI-2, with historical data. We utilize simulations to show that this approach can provide greater accuracy when appropriate historical data is available.
|
