Translational Informatics (TI) Community

Session Description

Translational Informatics (TI) provides a variety of tools to support data-driven methods for understanding biomedical data. This session will highlight two applications of TI to support the use of real-world data.

Jasmine Luzum, PharmD, PhD, will discuss the use of genetic or polygenic risk scores to distinguish patients who had survival benefit from beta-blockers for the treatment of heart failure with reduced ejection fraction from those that did not. She will also review the methodology used to develop polygenic risk scores and their potential clinical utility in pharmacogenomics.

Michael Williams will describe how large-scale longitudinal medication data collected during the course of routine patient care can enable post-market pharmacokinetic (PK) and pharmacodynamic (PD) studies to identify patient characteristics affecting the PK/PD profile which are not studied during the standard drug approval process. Detailed medication dose information is often found in free text among vast irrelevant data and can be obscured by shorthand. Although these medication data can be extracted using a natural language processing (NLP) system, there is still considerable labor involved in validating, processing, and preparing PK/PD datasets. Further, EHR sourced medication data can require assumptions about dosing which are unjustifiable in certain analyses. All of these challenges are magnified by the observational and incomplete nature of EHR data. In this talk we will give an overview of EHR2PKPD, a PK/PD data building system using EHR data. This system consists of several modules that form a complete pipeline of NLP and data building algorithms to generate analysis-ready population PK/PD datasets from free-text and structured clinical notes. Michael will discuss the challenges of EHR sourced medication data, how they have built a system to address these challenges, and the limitations which are yet unresolved for implementation of EHR-sourced medication data in PK/PD research.

TI leaders will also provide an update on the Community and engage with the audience to discuss the role TI may play in other ASCPT Networks and Communities and how to become more involved in the TI Community.

Learning Objectives

1. Demonstrate how genetic/polygenic risk scores can be applied in clinical pharmacology research.

2. Understand the opportunities and challenges of using real-world data to support pharmacometric analyses.

Session Schedule

3:45 PM - 3:55 PM ET
Introduction and TI Community Update
Sara Quinney, PharmD, PhD, Mohamed Shahin, PhD, & Caitrin McDonough, PhD

3:55 PM - 4:10 PM ET
Application of Genetic Risk Scores in Research: Beta-Blocker Survival Benefit in Heart Failure with Reduced Ejection Fraction
Jasmine Luzum, PharmD, PhD

4:10 PM - 4:22 PM ET
Leveraging Medication Data from the Electronic Health Record: Opportunities, Challenges and the EHR2PKPD Toolset
Michael Williams, BS

4:22 PM - 4:45 PM ET
Panel Discussion and Live Q&A
All Speakers