Modeling the initiation of others into injection drug use, using data from 2,500 injectors surveyed in Scotland during 2008-2009

Am J Epidemiol. 2015 May 15;181(10):771-80. doi: 10.1093/aje/kwu345. Epub 2015 Mar 18.

Abstract

The prevalence of injection drug use has been of especial interest for assessment of the impact of blood-borne viruses. However, the incidence of injection drug use has been underresearched. Our 2-fold aim in this study was to estimate 1) how many other persons, per annum, an injection drug user (IDU) has the equivalent of full responsibility (EFR) for initiating into injection drug use and 2) the consequences for IDUs' replacement rate. EFR initiation rates are strongly associated with incarceration history, so that our analysis of IDUs' replacement rate must incorporate when, in their injecting career, IDUs were first incarcerated. To do so, we have first to estimate piecewise constant incarceration rates in conjunction with EFR initiation rates, which are then combined with rates of cessation from injecting to model IDUs' replacement rate over their injecting career, analogous to the reproduction number of an epidemic model. We apply our approach to Scotland's IDUs, using over 2,500 anonymous injector participants who were interviewed in Scotland's Needle Exchange Surveillance Initiative during 2008-2009. Our approach was made possible by the inclusion of key questions about initiations. Finally, we extend our model to include an immediate quit rate, as a reasoned compensation for higher-than-expected replacement rates, and we estimate how high initiates' quit rate should be for IDUs' replacement rate to be 1.

Keywords: incarceration rate; initiator characteristics; injection drug users; injector incidence; replacement rate.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Female
  • Health Surveys
  • Humans
  • Incidence
  • Male
  • Prisoners / statistics & numerical data*
  • Regression Analysis
  • Risk-Taking
  • Scotland / epidemiology
  • Selection Bias
  • Substance Abuse, Intravenous / epidemiology*