Missing Data in the Baseline Health Surveys of the All of Us Research Program and the Opportunity from Multiple Information Sources

• Activity Number: 312 - What We Know About What We Don’t Know: Overcoming Incomplete Data in Practice
• Type: Invited
• Date/Time: Tuesday, August 9, 2022 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Qingxia Chen* and Robert M Cronin and Xiaoke Feng and Lina Sulieman and Brandy Mapes and Shawn Garbett and Ashley Able and Rebecca Johnston and Mick P. Couper and Brian K Ahmedani
• Companies: Vanderbilt University Medical Center and The Ohio State University and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and Vanderbilt University Medical Center and University of Michigan and Henry Ford Health System
• Keywords: All of Us; Electronic Medical Records; Missing Data; Multiple Information Sources; Observational Medical Outcomes Partnership; Surveys

Abstract

The All of Us Research Program is building a national longitudinal cohort and collecting data from multiple information sources including physical measurements, surveys, wearables and electronic health records, and soon genomics, to advance precision medicine. It is by far one of the largest and most diverse biomedical data resource of its kind containing data from over 315,000 participants with a goal of reaching at least 1 million. As an important component, surveys will complement, validate, and augment information from other sources. However, missing responses from surveys pose a challenge to conclusions using those data. We will describe the missingness from the initial three surveys from All of Us and evaluate factors associated with missingness. Additionally, by comparing information from family health history survey module vs electronic health record data that were mapped to the observational medical outcomes partnership data model, we will illustrate the challenge and opportunity to augment missing information from multiple information sources.