Meta-analysis of Binary Data Using Profile Likelihood

Illustrating the use of meta-analysis in many real-world trial examples, this work focuses on the analysis and modeling of a meta-analysis with individually pooled data (MAIPD). It explores alternatives to the profile likelihood method, including approximated likelihood and multilevel models, and shows how the nonparametric profile maximum likelihood estimator can be computed via the EM algorithm with a gradient function update. The authors explain how to test for and determine the amount of heterogeneity in an MAIPD. They also highlight the application of the software program CAMAP for analyzing an MAIPD and offer the software for free online.
• Chapman & Hall/CRC Interdisciplinary Statistics
Statistical Theory & Methods
Statistics for the Biological Sciences
Reference
Professional