The V AMMCS International Conference
Waterloo, Ontario, Canada | August 18-23, 2019
AMMCS 2019 Plenary Talk
Quality of Electronic Health Databases for Real-World Evidence Studies
Lisa Lix (University of Manitoba)
Canada is a world leader in the creation, linkage, and use of population-based electronic health databases to support healthcare decision making. However, there are challenges in using electronic health databases to improve our understanding of health. These databases were developed to monitor patient care and manage the healthcare system; they were not intended to be used for research. Accordingly, the information contained in electronic health databases is often of poor quality for real-world studies; it lacks standardization and documentation, and frequently contains inaccurate and inconsistent information. In this talk, methods to combine electronic health data from multiple sources or time periods to improve the accuracy of health measures will be discussed, including imputation models, latent variable models, and various approaches to data validation. Activities to automate data quality evaluations will be described. Examples from provincial and national chronic disease research and surveillance projects will be used to illustrate the challenges and opportunities.
Dr. Lisa Lix is Professor and Tier I Canada Research Chair in Methods for Electronic Health Data Quality in the Department of Community Health Sciences, Max Rady College of Medicine, University of Manitoba. She is also Director of the Data Science Platform in the George & Fay Yee Centre for Healthcare Innovation (CHI), a collaboration between the Winnipeg Regional Health Authority and the University of Manitoba. Her team of 30+ faculty, staff and trainees have expertise in biostatistics, bioinformatics, and clinical database development; they focus on providing methodological research, training, and consulting services to strengthen patient-oriented research in Manitoba. Dr. Lix's areas of research expertise include methods to address bias and error in electronic health databases, statistical methods for patient-reported outcomes, and longitudinal data analysis. She is a prolific researcher who has published more than 325 scholarly papers and mentored more than 40 trainees. Dr. Lix is a member of the Board of Directors of the Statistical Society of Canada, Program Chair for the Society's 2019 meeting in Calgary, Co-Chair of the Data Quality Working Group for the Canadian Chronic Disease Surveillance System, and Program Chair-Elect for the 2020 Joint Statistical Meetings Health Policy Statistics Section.