1st Seattle Symposium Health Care Data Analytics

Confronting statistical challenges of using electronic health record data to conduct health research

September 28–30, 2014 │ Seattle, Wash.

Biostatisticians and other scientists from across the country gathered September 28–30 in Seattle to discuss health research and methods that involve large health care databases. The 1st Seattle Symposium on Health Care Data Analytics featured innovative research on pragmatic clinical trial design, inference and prediction using electronic health record data, and drug and vaccine safety surveillance. Experts who are involved in national initiatives that use large health care databases to answer important research questions discussed methodological challenges encountered in this setting and share ideas for addressing them.

Thanks to all of our speakers and attendees for making this event a huge success! Please help us improve future meetings by filling out a brief evaluation.

Questions? Please send them to seattlesymposium@ghc.org.

View conference agenda (PDF) and speaker bios (PDF).

View attendees and contact information (PDF).

Download presentations:
Monday, Sept. 19
8 – 9:30 a.m.SESSION 1: Welcome & overview
Health care databases for national research and surveillance initiatives (PDF)
Denise Boudreau, RPh, PhD

The study of cardiovascular health outcomes in the era of claims data (PDF)
Bruce M. Psaty, MD, PhD

Bias and problems with adjustment for bias in database studies of influenza vaccine effectiveness (PDF)
Lisa Jackson, MD, MPH

10 a.m. – Noon│Session 2: Pragmatic clinical trials

Pragmatic clinical trials: Methods for learning within health care delivery systems (PDF)
Patrick Heagerty, PhD

Power calculation for cluster-randomized trials via resampling (PDF)
Ken Kleinman, ScD

Pragmatic Solutions for Pragmatic Problems: Examples from the STOP CRC Trial (PDF)
William M. Vollmer, PhD

1:30 – 3 p.m.│Session 3: Inference and prediction using electronic health record data

With big data comes big responsibility: Using health care data to emulate randomized trials (PDF)
Miguel Hernán, MD, DrPH

Adjustment of selection bias in secondary analysis of electronic health record data (PDF)
Sebastien Haneuse, PhD

Statistical methods for misclassified outcomes and exposures in data from electronic medical records (PDF)
Rebecca Hubbard, PhD

3:30 – 5 p.m.│Session 3 (cont.): Inference and prediction using electronic health record data

Development and validation of a predictive model for preemptive pharmacogenetic testing (PDF)
Jonathan S. Schildcrout, PhD

A statistical framework for individualizing health (PDF)
Scott Zeger, PhD

Tuesday, Sept. 30
8 – 10 a.m. │Session 4: Safety surveillance

Gaps and opportunities: Methodologic challenges in post-market safety surveillance. (Why observational studies are here to stay, and why we need to get them right.) (PDF)
Susan Gruber, MPH, PhD

Accounting for uncertainty in observational database studies (PDF)
David Madigan, PhD

Adapting sequential methods for active medical product safety surveillance in a distributed health care database setting (PDF)
Andrea Cook, PhD

Discussion: Robert Platt, PhD (PDF)

10:30 – 11:45 a.m. │Panel Discussion: Complementary contributions of randomized clinical trials, traditional observational studies, and health care database studies

Drawing inferences about effects of medical treatment (PDF)
Susan S. Ellenberg, PhD

Healthcare data analytics: Classical challenges, contemporary solutions (PDF)
Sarah M. Greene, MPH

Some observations from the Women’s Health Initiative clinical trial and cohort study (PDF)
Ross L. Prentice, PhD