Meta-analysis of Rare Events
Open AccessMeta-analysis (MA) is a quantitative approach that provides a systematic way of integrating the findings of individual research studies. However, for the analysis of rare events such as adverse events in drug safety studies, the conventional MA methods that are built on asymptotic distributions tend to produce biased results. In the first part of this thesis, we propose a novel method for individual participant data meta-analysis (IPDMA) of incidence rate (IR) in the context of excess zero counts. This exact likelihood method is built on a Poisson-Gamma hierarchical model. We thoroughly compare the performance of the proposed method and several other methods via a comprehensive simulation study, and conclude that the proposed method is preferable in terms of percent bias, root mean square error and empirical coverage probability. We further apply the proposed method to a real-world study. In the second part of this thesis, we focus on aggregate data meta-analysis (ADMA) of incidence rate ratio (IRR) in the context of rare events. In ADMA, most widely used exact methods are built on generalized linear mixed-effects models (GLMMs). However, these methods are mainly discussed for binary outcomes such as odds ratio and relative risk. There is a lack of systematic review of these methods for IRR. To fill this gap, we review and compare the performance of some existing GLMMs through a simulation study in the context of rare events for IRR. All models are illustrated using several real-world studies.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.