A Foundation for Linking Nontraditional Data Sources to the National Death Index: Understanding Mortality Risk Among Gang Members

David Pyrooz , University of Colorado Boulder
Richard G. Rogers, University of Colorado Boulder
Ryan K. Masters, University of Colorado Boulder
Jennifer Tostlebe, University of Colorado Boulder

We examine mortality risk among an understudied and hard-to-read but exceptionally high-risk subpopulation in the United States: gang members. Prior studies have reported mortality statistics on gang members, but this work was hindered by small sample sizes of gang members and deaths, often crudely constructed comparison groups (or none at all), and a lack of systematic data collection on mortality, including cause-of-death. We introduce the Gang Member-Linked Mortality Files (GM-LMFs) database, composed of 3,154 male gang members (mean age=20, 90% black) in the St. Louis area, derived from law enforcement gang intelligence gathered between 1993 and 2003 and linked to National Death Index records through 2016. We outline the procedures for linking a non-traditional data source to the NDI, and we report preliminary findings on baseline risk of mortality, including overall and cause-specific, and years of potential life lost among gang members.

See extended abstract

 Presented in Session 157. Violence and Health