September 9, 2016: Million Veteran Program

September 9, 2016: Million Veteran Program


Million Veteran Program: A Mega-Cohort Within a Healthcare System


J. Michael Gaziano, MD, MPH, Scientific Director of Massachusetts Veterans Epidemiology Research Center (MAVERIC); Principal Investigator of the Million Veteran Program; and Professor of Medicine, Harvard Medical School


Veterans; Million Veteran Program; VA; Population health; Pragmatic clinical trial; Biorepository; MAVERIC; Phenotyping; Genetic analysis; Big data; Self-reported survey; Electronic health records

Key Points

  • The Veterans Affairs (VA) Healthcare System is ideally suited to nested, large-scale population research such as the Million Veteran Program (MVP). The VA has a stable, willing Veteran population of 8 million, with diverse healthcare expertise, EHR infrastructure, and data reaching back 20 years.
  • MVP is a national, voluntary research program funded by the VA’s Office of Research & Development with a goal of partnering with Veterans receiving care in the VA system to study how genes affect health.
  • MVP will enroll up to 1 million Veteran volunteers into an observational mega-cohort, collect health and lifestyle information, and collect blood for storage in a biorepository. The MVP data sources include self-reported surveys, genetics, VA healthcare system, National Death Index, CMS, and the Department of Defense.
  • The program is using a new semi-automated approach for phenotyping that combines features of manual and automated phenotype development.

Discussion Themes

When dealing with complex data (for example, radiology images), it’s important to understand the “anatomy” of the data. Report structures can vary, with some having regional distinctions.

How do you handle protected health information embedded in text fields?

Do participants have access to the genetic results of the study?

Because this is a national resource, we want as many investigators as possible to use it. We want the data to be centralized and bring investigators to the data. We want the construct to be flexible so that data curation and imputation are available to the next investigator.

For More Information

Visit the Million Veteran Program (MVP) website at:

#MillionVeteranProgram, #pctGR
@MillionVeteran, @DeptVetAffairs, @PCTGrandRounds, @Collaboratory1, @PCORnetwork