Status: completed Duration: Mar 2013 - Dec 2016
Marginal structural models (MSMs) are increasingly used to address confounding issues in biomedical reserarch. This meta-epidemiological study seeks to explore whether modern MSM-based analyses of ‘real world’ observational data can be used to reliably guide health care decision making. We aim to identify potential factors or characteristics of ‘real world’ data sources, analytical approaches, or clinical topics which affect the reliability of observational data analyses. If possible, a list of criteria that affect the reliability of ‘real world’ observational data should be created that will inform future guideline development, health care decision making, and research.