Examining the Potential for Bias in IPD Meta-Analysis Results

Richard D. Riley, Jayne F. Tierney, Lesley A. Stewart

Research output: Chapter in Book/Report/Conference proceedingChapter


Obtaining individual participant data (IPD) can reduce many of the potential biases facing conventional meta-analyses of published aggregate data. This chapter describes how researchers can examine the potential impact of biases on IPD meta-analysis results by using sensitivity analyses, funnel plots, and a combination of IPD and aggregate data. Publication bias is a well-known threat to the validity of traditional meta-analyses based on aggregate data. IPD meta-analysis projects have strong potential to reduce publication bias in meta-analysis by obtaining data for unpublished trials. The factors that cause between-trial heterogeneity in treatment effect may also cause small-study effects and thus asymmetry in a funnel plot. Availability bias can occur in an IPD meta-analysis project if IPD are unavailable for some trials and this unavailability is related to the trial results. Unavailability of IPD may be related to trial quality. Selective outcome reporting occurs where trial outcomes are entirely excluded upon publication.

Original languageEnglish
Title of host publicationIndividual Participant Data Meta-Analysis
Subtitle of host publicationa Handbook for Healthcare Research
PublisherWiley-Blackwell Publishing Ltd
Number of pages15
ISBN (Electronic)9781119333784
ISBN (Print)9781119333722
Publication statusPublished - 22 Apr 2021

Bibliographical note

Publisher Copyright:
© 2021 John Wiley & Sons Ltd. All rights reserved.

Cite this