Differences in time to injection onset by drug in California: Implications for the emerging heroin epidemic

Abstract

BackgroundHeroin use is increasing in the US. Heroin use may predispose users towards injection routes of drug administration as compared to other illicit substances. ObjectiveTo explore the relationship between heroin use and drug injection, we compared time from first use to first injection (referred to as time to injection onset by drug [TTIOD]) of heroin, methamphetamine/speed, cocaine, and crack cocaine among people who inject drugs (PWID). MethodsAge of first use and first injection by drug was collected from 776 PWID. Survival analyses were used to determine TTIOD and to examine demographic factors associated with TTIOD. Cox regression analysis was used to determine demographic factors associated with drug-specific injection onset. ResultsThe eventual injection onset rate by the drug was 99% for participants who used heroin, 85% for participants who used methamphetamine/speed, 80% for participants who used powder cocaine, and 38% for participants who used crack cocaine. Hazard ratios for injection use within one year of first use by drug were: 1.37 (median survival time [MST] = 0.61 years) for heroin, 0.66 (MST = 1.10 years) for methamphetamine/speed, 0.50 (MST = 2.93 years) for powder cocaine, and 0.12 (MST = 39.59 years) for crack cocaine. Demographic differences in TTIOD were found for each drug. Demographic differences were found for eventual injection by drug for all substances except heroin. ConclusionAmong PWID, heroin use was associated with a more rapid transition to injection and a higher rate of eventual heroin injection regardless of demographics. More robust, innovative efforts to reduce heroin use and prevent injection initiation are urgently needed.