Incorporating Efficacy Data from Initial Trials Into Subsequent Evaluations: Application to Vaccines Against Respiratory Syncytial Virus

Abstract

Background When a randomized controlled trial fails to demonstrate statistically significant efficacy against the primary endpoint, a potentially costly new trial would need to be conducted to receive licensure. Incorporating data from previous trials might allow for more efficient follow-up trials to demonstrate efficacy, speeding availability of effective vaccines. Methods Based on the outcomes from a failed trial of a maternal vaccine against respiratory syncytial virus (RSV), we simulated data for a new Bayesian group-sequential trial. We analyzed the data either ignoring data from the previous trial (i.e., weakly informative prior distributions) or using prior distributions incorporating the historical data into the analysis. We evaluated scenarios where efficacy in the new trial was the same, greater than, or less than that in the original trial. For each scenario, we evaluated the statistical power and type I error rate for estimating the vaccine effect following interim analyses. Results When we used a stringent threshold to control the type I error rate, analyses incorporating historical data had a small advantage over trials that did not. If control of type I error is less important (e.g., in a post-licensure evaluation), the incorporation of historical data can provide a substantial boost in efficiency. Conclusions Due to the need to control the type I error rate in trials used to license a vaccine, incorporating historical data provides little additional benefit in terms of stopping the trial early. However, these statistical approaches could be promising in evaluations that use real-world evidence following licensure.

A scaled kernel density estimation prior for dynamic borrowing of historical information with application to clinical trial design

“…However, determining how to most effectively borrow historical information in a Bayesian setting is an open question, as incorrectly borrowing (ie, when the historical and current data are in disagreement) can result in inaccurate statistical inference, while failing to utilize supporting information is inefficient. To this end, several methods that allow for data‐driven borrowing of historical information to improve statistical inference in the current data analysis have been introduced (eg, References 1‐3) and are becoming increasingly used in practice across many different scientific disciplines (eg, References 4‐10). We detail several of these major methodological categories in Section 3.…”