Program Description
Meta-Analysis is the gold standard statistical approach to combine the results of multiple studies and to examine sources of heterogeneity and potential biases. This course will review fixed-effect and random-effects models that underlie the combination of study results in meta-analysis; the use of study-level predictors in meta-regression; assessment of small-study effects and related reporting biases; and sensitivity analyses to bias. Examples will cover meta-analysis of randomized trials and of observational studies. Throughout the course, participants will apply each model by using the R software.
Competencies
Participants will learn to:
- Utilize the fixed-effect and random-effects methods of combining effect sizes;
- Describe different ways to measure between-study heterogeneity;
- Describe the strengths and weaknesses of random-effects as compared to fixed effect meta-analysis;
- Assess the potential impact of small-study effects and related reporting biases on a combined effect size estimate;
- Perform meta-regression modeling and describe the limitations of meta-regression
Intended Audience
The target audience includes biostatisticians, data analysts, and quantitative researchers from academia, the pharmaceutical industry, and other government institutions.
Required knowledge/pre-requisites
Basic knowledge of study design and regression modeling, and a basic working knowledge of R are necessary to be successful in the course. Participants must bring a laptop to the class sessions.
Discounts available—visit our FAQs page to learn more.
Low-cost, on-campus housing is also available. Contact us for more information.
Additional Information
Detailed description of sessions
Session 1: Meta-analysis models
Learning objectives:
- Understand and apply the fixed-effect model
- Understand and apply DerSimonian and Laird random-effects model
- Use different estimators of between-trial variance
- Create forest plots
Session 2: Analyzing non-standard data
Learning objectives:
- Continuous outcomes
- Prevalence data or absolute risks
- Correlated data
- Observational studies: sensitivity analysis using credibility ceilings
Session 3: Exploring between-trial heterogeneity
Learning objectives:
- Subgroup analyses
- Meta-regression analyses
Session 4: Addressing small-study effects and reporting bias
Learning Objectives:
- Funnel plots
- Regression-based methods to adjust for small-study effects
- Copas selection model
Networking reception, interaction with folks in other courses, knowledgeable + honest + experienced faculty.
