Alexis N. Martinez, PhDc, MPH¹, Alexander H. Kral, PhD¹, Rachel L. Anderson, BA², Neil Flynn, MD, MPH², Andrea Scott¹, Mary Lou Gilbert, JD, MA³, Brian K. Finch, PhD³, and Ricky N. Bluthenthal, PhD³. (1) Urban Health Studies, UCSF, 3180 18th St. Suite 302, San Francisco, CA 94110, 415-502-5338, amartin@itsa.ucsf.edu, (2) Infectious Diseases, UC Davis, 4150 V St., Suite 500, Sacramento, CA 95817, (3) RAND Health, RAND, 1700 Main Street, PO Box 2138, Santa Monica, CA 90405

Background: In California, the rate of arrests for drug-related crimes increased 25-fold in the past 20 years. The consequences of arrest include loss of housing, employment, and/or government assistance. Objective: To determine which individual-level and neighborhood-level variables are associated with arrest in the last 6 months among injection drug users (IDU) in California. Methods: We obtained arrest data from 1,588 IDUs who attended 23 syringe exchange programs (SEP) in California between 2001 and 2003. Zip-code level neighborhood characteristics were measured using ZCTA data from the 2000 United States Census. Results: 28% of IDUs had been arrested in past 6 months. In bivariate analysis, the following neighborhood-level variables were associated (p<0.05) with arrest: ratio of household income to the poverty level, percentage of poverty, percentage of renters, percentage of public assistance, percentage of Hispanic or Latino residents, and percentage of black residents. These variables did not remain significant at the multivariate level when injection-related and socio-economic status related individual measures were introduced in the model. Being homeless, being unemployed, reporting sex work in the past 6 months, injecting heroin in the last 30 days, and injecting speed in the last 30 days were independently associated with arrest in the last 6 months. Attending an SEP in an urban area and a history of injecting drugs less than 10 years was independently associated with not being arrested in the last 6 months. Conclusion: Individual-level variables were better predictors of arrest than neighborhood-level variables in this drug-using population.

Learning Objectives:

This session will describe individual and neighborhood factors that influence arrests in California. Each participant will be able to discuss the individual-level characteristics that predict arrest, and assess the use of census data in this analysis. At the conclusion of the session, the participant (learner) in this session will be able to
• Describe the individual-level and census-level variables used in the analysis of arrests.
• Discuss the ability of census-level variables to predict arrests in a drug using population.
• Articulate the need for further research on using census-level variables in drug-using populations.

Keywords: Drug Use, Criminal Justice

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.