Designing observational studies using electronic health record data
Date
Friday, March 31 via Zoom. 2:00pm to 3:00pm
Speaker
Ruta Brazauskas, Associate Professor, Division of Biostatistics, Medical College of Wisconsin
About the speaker
Dr. Brazauskas is an Associate Professor in the Division of Biostatistics at the Medical College of Wisconsin, a biostatistician for the Center for International Blood and Marrow Transplant Research (CIBMTR) and Primary Immune Deficiency Treatment Consortium (PIDTC). She has extensive experience in designing observational research studies and prospective clinical trials and performing statistical analysis of the data from a variety of medical studies. Dr. Brazauskas teaches biostatistics at MCW and has a long track record of mentoring physicians. She provides statistical consulting services via Division of Biostatistics Consulting Service and the CTSI Biostatistics & Epidemiology Research Design (BERD) Program.
Abstract
Although clinical trials are the gold standard for evaluating experimental treatments, in settings where clinical trials may not be feasible or ethical, observational studies are the next best option. Electronic health records (EHRs) are an increasingly important source of data for observational research. Analyses of data collected for purposes other than research require careful consideration of data quality. In addition, due to the observational nature of this type of studies, there are potential selection bias or confounding factors that need to be accounted for when assessing treatment effect or comparing groups of patients. In this seminar, we will introduce various types of observational studies, common pitfalls, and common analysis approaches for observational studies. Aspects relevant to EHR-based cohort selection and outcome identification will be discussed.
NIH Funding Acknowledgment:Important Reminder – Please acknowledge the NIH when publishing papers, patents, projects, and presentations resulting from the use of CTSI resources by including the NIH Funding Acknowledgement.
