Table I.

Multivariate logistic regression (GEE) analysis of factors associated with reporting willingness to cease selling drugs during follow-up (n = 868).

Characteristic	Adjusted Odds Ratio* (95% CI)	p value
Age
Per year older	1.02 (1.01 – 1.03)	0.003
Gender
Male vs. Female	0.99 (0.77 – 1.26)	0.909
Recent Crack Use
Yes vs. No	2.00 (1.44 – 2.79)	< 0.001
Recent Injection Crystal Methamphetamine
Yes vs. No	0.62 (0.47 – 0.83)	0.001
Recent Public Injection
Yes vs. No	1.95 (1.55 – 2.43)	< 0.001
Police Presence Affects Drug Purchase Location
Yes vs. No	1.53 (1.22 – 1.91)	< 0.001

Note: GEE = Generalized Estimating Equation; CI = Confidence Interval

^*Adjusted for ethnicity, residency in Vancouver’s downtown eastside, homelessness, amount of money spent on drugs per day, having experienced physical assault, having committed physical assault, involvement in the commercial sex trade, drug treatment, being stopped or detained by police, injection heroin use, injection cocaine use, and HIV serostatus