Introduction to the Design & Analysis of Observational Studies of Treatment Effects

Speaker(s)

Faculty: Linus Jönsson, PhD, MD, MSc, Section for Neurogeriatrics, Department of Neurobiology Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden Bradley Martin, PharmD, RPh, PhD, Division of Pharmaceutical Evaluation and Policy, University of Arkansas for Medical Sciences College of Pharmacy, Little Rock, AR, USA

Retrospective studies require strong principles of epidemiologic study design and complex analytical methods to adjust for bias and confounding. This course will provide an overview of the structures of commonly encountered retrospective data sources with a focus on large administrative data, as well as highlight design and measurement issues investigators face when developing a protocol using retrospective observational data. Approaches to measure and control for patient mix, including patient comorbidity and the use of restriction and stratification, will be presented. Linear multivariable regression, logistic regression, and propensity scoring analytic techniques will be presented and include examples using SAS code that can later be used by participants. This course is an introductory course designed to prepare participants to take intermediate and advanced observational research courses.

Code

005