|
Abstract:
|
Vaccine efficacy is often assessed as the reduction of infection or disease rate in the vaccinated group as compared to the unvaccinated group based on an exact conditional binomial test. Because of low incidence rate, the required sample size is often very large, which poses big challenge for study conduct and enrollment. Bayesian framework provides a natural avenue to utilize historical information, if there is evidence to suggest similarity of the responses between the historical and current studies. Schoenfeld et al. (2009) proposed a Bayesian hierarchical model using adult data to increase power for a pediatric trial with continuous endpoints. In this talk, we first consider a hierarchical conditional Binomial model within a Bayesian framework to incorporate historical information. Secondly, we propose a Beta prior distribution which is equivalent to the power prior to borrow partial historical information quantitatively. Simulations are conducted to evaluate the statistical properties such as power and type-I error. The methods are demonstrated by an example to show the improved efficacy estimation through borrowing information from a historical study.
|
