Online Program

Return to main conference page

All Times EDT

Thursday, October 1
Thu, Oct 1, 2:40 PM - 3:55 PM
Electronic Health Record Data: Statistical Issues and Opportunities

Electronic Health Record Data: Statistical Issues and Opportunities (309549)

Xiangqin Cui, Emory University School of Public Health 
Kristen Dang, Sage Bionetworks 
Rebecca Hubbard, UPenn School of Medicine 
Jessica Lavery, Memorial Sloan Kettering Cancer Center 
*Katherine S Panageas, Memorial Sloan Kettering Cancer Center 

Keywords: EHR, real-world data linkage

The use of patient encounter data contained within electronic health records (EHRs) has provided unique opportunities for innovative research. EHRs have been adopted globally and advances in efficient software have made it possible to capture large amounts of clinical data. These rich, big-data sources allow for an efficient low-cost approach to further study therapeutic interventions, biological mechanisms, appropriateness and quality of healthcare delivery in real-time. However, since the data compiled from EHRs are not intended for clinical research purposes, preparation and understanding of the data for secondary use are critical for analyses. Inadequate understanding and poor data standards can result in erroneous inferences in real-world comparative effectiveness data analyses which might impact clinical decision-making. Linking EHR data with other complex data sources such as genomic sequencing or administrative claims data can provide additional richness but are challenging to compile. In this session, data linkage processes will be outlined. Statistical issues and analytic challenges when using EHR data will be discussed. *Sponsored by the Caucus for Women in Statistics