Online Program Home
My Program

Abstract Details

Activity Number: 582 - Random Effects and Mixed Models
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 2:00 PM to 3:50 PM
Sponsor: Biometrics Section
Abstract #330756 Presentation
Title: A Multi-Level Mixed-Effects Model for Individual Participant Data Meta-Analysis with Outcomes from an Exponential Family
Author(s): Ying Zhang* and Vernon M Chinchilli
Companies: Penn State College of Medicine and Penn State College of Medicine
Keywords: mixed model; meta-analysis; multi-level model; individual participant data

Individual participant data (IPD) meta-analysis that combines and analyzes raw data from studies has been suggested to be more powerful and flexible compared with meta-analysis based on summary statistics. We propose a one-stage method for IPD meta-analysis with outcome variables from an exponential family, which is a statistical model that is a combination of a generalized linear mixed-effect model and a multi-level model. The new model contains fixed effects and random effects for each level, such as participant and study level, and it allows each study to have different designs, different lengths of follow-ups, and different sets of covariates. We have derived the estimators of fixed-effect parameters and variance-covariance parameters. To evaluate the proposed model, we performed a simulation study in which we generated multicenter clinical data to mimic clinical studies investigating a treatment effect and then applied the proposed model, 3-level mixed-effects models. Compared with naïve models, the proposed model generally improved the precision, as indicated by smaller biases of fixed-effect parameters, and provided more accurate estimates of variance-covariance parameters.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program