American Statistical Association
Observational data are valuable in the study of cancer screening and diagnostic evaluation because the performance, utilization and outcomes of screening and diagnostic tests in routine clinical practice may differ from that observed in clinical trials. Data from electronic health records (EHR) are particularly well-suited to such studies because they reach a large, representative population and directly allow us to characterize care in routine practice. Despite this potential, many aspects of EHR data pose a challenge to valid inference including confounding by indication, misclassification of exposures and outcomes, missing data, and informative observation schemes. In this talk, I will present a few examples of opportunities and challenges I have encountered in studying cancer screening and diagnosis using EHR data. For each example, I will discuss alternative analytic approaches and introduce methodology to address some of the challenges of working with EHR data. I will conclude with a discussion of areas where future methodological development is needed to capitalize on the potential of EHR data.
|Date:||Wednesday, November 8, 2017|
|Time:||4:00 - 5:00 P.M.|
Memorial Sloan Kettering Cancer Center
Department of Epidemiology and Biostatistics
485 Lexington Avenue
(Between 46th & 47th Streets)
2nd Floor, Conference Room B
New York, New York
**Outside visitors please email email@example.com for building access.
You must be on the security list to enter the floor.