Meta-Analysis of Clinical Trials With Sparse Binary Outcomes Using Zero-Inflated Binomial (ZIB) Models

Abstract

In meta-analysis of clinical trials, standard statistical methods run into problems when the proportions of safety events are small. Motivated by the dataset used in a published analysis of cardiovascular safety in Rosiglitazone trials, this article proposes using a zero-inflated binomial model to handle the zero-event trials. The maximum likelihood estimates of the model parameters are obtained using the expectation and maximization algorithm. Via simulation studies, it is shown that the proposed methods provide estimates of odds ratios with less bias and variation, compared with both the Mantel–Hanszel method with continuity correction and Peto’s method. The proposed methods are applied to the Rosiglitazone trials. Supplementary materials for this article are available online.

Keywords:

• EM algorithm
• Fixed-effects model
• Rare events

Supplementary Materials

The online supplement contains additional simulation results.

Acknowledgments

The authors thank the two anonymous referees for their constructive comments that greatly contributed to improving the article.