Abstract Body

Misuse of prescription opioids (PO) increased dramatically in the U.S. since the early 2000s and may lead to increases in heroin use and drug injection. Understanding the role of PO in injection initiation is key to preventing injection-related harms, including HIV and hepatitis C. We assessed factors associated with PO abuse prior to first injection (prior PO abuse) among people who inject drugs (PWID).

PWID ages ≥18 years were recruited for the 2015 National HIV Behavioral Surveillance using respondent-driven sampling. Data on prior PO abuse (‘hooked on painkillers before you injected for the very first time’) among PWID who injected opioids (heroin, PO; alone or in combination) were available in 16 cities. We estimated Poisson regression models with generalized estimating equations clustered on recruitment chain and adjusted for sampling design covariates to assess factors associated with prior PO abuse. We report adjusted prevalence ratios (aPR) and 95% confidence intervals (CI).

Of 7,454 PWID, 2,208 (30%) reported prior PO abuse. Prior PO abuse was higher among PWID who began injecting in more recent years (Figure). PWID reporting prior PO abuse compared to other PWID were more likely to be younger (mean age 34 vs. 47 years, p<0.0001), female (32% vs 26%, aPR 1.22, CI 1.13-1.31), non-Hispanic white (65% vs. 33%, aPR 2.14, CI 1.85-2.47), have high school education or higher (74% vs 68%, aPR 1.21, CI 1.12-1.30) and receptively shared syringes (44% vs 33%, aPR 1.28, CI 1.19-1.37). PWID who reported prior PO abuse were less likely to test HIV-positive, even after controlling for age and race/ethnicity (2% vs 6%, aPR=0.59, 95%CI 0.41-0.84). Common sources of first ever PO were prescription by a physician (41%), purchased from friends, family or others (36%), and given by friends (29%). Mean time between first ever PO use and first injection was 5 years (SD=6.4, median 3).

Prior PO abuse was substantially higher among PWID who began injecting during the opioid epidemic (i.e., since 2000). PWID with prior PO abuse differed in socio-demographic characteristics and despite lower HIV prevalence, were more likely to engage in practices that increase risk of infection. Efforts to prevent HIV transmission and other blood-borne infections would benefit from injection prevention interventions for people who abuse PO and increased understanding of barriers and facilitators to effective prevention delivery for those already injecting.