Abstract:
|
Improving the efficiency of clinical trial design and analysis is important in clinical development, especially for diseases with small populations. Borrowing of historical data under the assumption of subject exchangeability between the current and historical studies may improve the precision of treatment effect estimate and the power of a study. However, there have existed concerns about the exchangeability assumption as there are always differences between the current and historical data. Recent research has showed that the application of covariate-adjusted Bayesian hierarchical method in historical data borrowing can enhance the appropriateness of the exchangeability assumption when the differences between the current and historical data can be explained by patient's baseline demography and disease characteristics. In this presentation, we will demonstrate the application of different covariate-adjusted borrowing methods in clinical development for rare diseases such as spinal muscular dystrophy with simulation results.
|