Pediatric populations, comprising individuals aged 18 or younger, and rare disease populations pose significant challenges in clinical trial design due to limited participant numbers, patient sensitivity, and insufficient natural history data. Addressing these challenges often involves data extrapolation, leveraging existing data from adults to inform pediatric trials. Bayesian hierarchical modeling is increasingly recognized as a valuable tool for combining information across disparate sources, such as adult and pediatric datasets. This manuscript introduces an extension of an existing statistical model to enhance the efficiency of borrowing strength from multiple historical trials under consistent assumptions. A quantitative method is developed to improve the borrowing of historical information. The methods are illustrated with a simulation study motivated by real case study in pediatric clinical trial, and practical considerations are provided regarding the selection of prior distributions. This work aims to provide comprehensive insights and practical guidance for leveraging historical data effectively in clinical trial design.
Read full abstract