e126
Abstracts / Drug and Alcohol Dependence 171 (2017) e2–e226
Methods: This is a cross-sectional study using a probabilistic multistage cluster sample design to select 4607 Brazilian participants, 14 years of age and older from a representative sample of the household population (RR: 77%). Stata13 and Process for SPSS22 were used to estimate weighted prevalence rates and build the conditional model, which controlled for all demographic characteristics. Results: Lifetime and previous year smoked cocaine use was 1.3% and 0.7% respectively (men: 1.0%, women: 0.3%). Being a victim of at least one event of urban violence was reported by 2.6%, whilst being a perpetrator was reported by 6.1% of the population. Crack cocaine users were 4.2 (CI: 1.02–17.22) times more likely to perpetrate urban violence. They also were also 11.5 times more likely to have alcohol dependence disorder (CI: 4.06–32.64). The estimated conditional model suggests that crack cocaine use has a direct effect on urban violence (perpetration) (p < 0.000 coefficient: 0.22), which is mediated by depressive symptoms (index: 0.01, 0.003–0.005). Binge drinking significantly moderates this mediation (effect: 0.02, 0.004–0.05). Conclusions: Brazil has one of the highest rates of smoked crack cocaine in the world, with over 1 million crack cocaine users. The elevated rates of crack cocaine use combined with harmful alcohol consumption patterns are directly associated with the equally high rates of urban violence. Our conditional model sheds light on associated factors that must guide prevention strategies and mental healthcare. Financial support: CNPq Grant No 55.0024/2011-5. http://dx.doi.org/10.1016/j.drugalcdep.2016.08.349 Meta-analysis as a form of evidence: An examination of the cochrane reviews for behavioral health Stephen Magura ∗ , Miranda J. Lee, Stephanie Means Western Michigan University, Kalamazoo, MI, United States Aims: Program funders are increasingly requiring providers to implement “evidence-based programs.” Meta-analysis is viewed as the highest form of evidence for interventions in traditional evidence hierarchies. The study examined how meta-analysis is used by one of the best known registers of evidence-based programs, the Cochrane Reviews database. The study focused on reviews of behavioral health interventions. Methods: The seven and three most populated topic areas from mental health and substance abuse, respectively, were chosen for study. Three cochrane reviews were randomly selected from each topic area (total n = 30 reviews). Two expert raters rated each review by consensus on the analytic dimensions. Results: 23 of the reviews were formal meta-analyses; seven were reduced to systematic reviews due to lack of poolable data. Mean percent of identified studies in final analysis = 4% (range = <1–17%; mean no. = 16). Minimum design requirements for study inclusion were: Randomized controlled trial (RCT) (47%); Cluster RCT (14%); “controlled” clinical trials (20%); other (19%). Heterogeneity (variation in intervention effects not due to chance): I2 = 43% (range = 0–91%, SD = 36%); but I2 was not identifiable in 40% of the reviews. Formal framework for judging study quality (50%). Clinical significance of effects addressed (1 review). Forest plots of outcomes (80%). Publication bias formally addressed (17%). Conclusions about intervention effectiveness: positive (40%); mixed (37%); no effect (23%).
Conclusions: The small percent of identified studies included in final analyses may bias results. Despite the potential strength of meta-analysis, many reviews still had indefinite results. Heterogeneity of effects for interventions was substantial and the reviews gave no guidance for selecting one specific program over another. Key dimensions of concern to researchers and clinicians, such as quality of evidence, clinical significance and publication bias, were infrequently addressed. Financial support: National Institute on Drug Abuse R21DA032151. http://dx.doi.org/10.1016/j.drugalcdep.2016.08.350 Alcohol and cocaine use and drug distribution among formerly incarcerated Black men Carlos C. Mahaffey 2,∗ , Danelle Stevens-Watkins 1 , Carl Leukefeld 3 1
Spalding University, Louisville, KY, United States University of Kentucky, Lexington, KY, United States 3 Behavioral Science, University of Kentucky, Lexington, KY, United States 2
Aims: More than 75% of persons released from prison are rearrested in five years (Durose et al., 2014). Those rearrested are overrepresented by Black men charged with drug offenses. The purpose of this study is to examine alcohol and cocaine use and other factors as correlates of illegal drug distribution among Black men, N = 250. Methods: Data from drug-using Black men, one year after their release from prison, was used to conduct bivariate correlations and ordinal logistic regression. The dependent variable of interest was “How many times did you sell drugs in the past year?” with responses coded as 0 = never, 1 = once, 2 = a few (2–5 times), and 3 = a lot (≥5 times). Independent variables included age, 30 day alcohol use, one year cocaine use, drug use in their neighborhood, and their feelings that drug use causes problems with the law, family, and friends. It is hypothesized that men who used alcohol and cocaine; observed drug use in his neighborhood; and had problems would be more likely to sell drugs. Results: For every one year increase in age, men were 5% less likely (Exp B = .95, p = .02) to sell drugs. Men who used alcohol in the last 30 days were 4.2 (Exp B = 4.23, p = .03) times more likely to sell drugs than men who didn’t drink. Men who used cocaine in the past year were 2.4 (Exp B = 2.36, p = .03) times more likely to sell drugs than men who did not use cocaine. Men who reported drug use never causes problems with the law were 87% less likely (Exp B = .13, p < .01) to sell drugs than me that felt drug use always causes legal problems. Visible drug use and reporting that drug use caused family/friend problems were not significant. Conclusions: The current study adds to the literature that Black men who use alcohol and cocaine are at greater risk for engaging in illegal drug activity after release from prison and are likely to be rearrested. Future studies are needed to examine factors associated with reducing drug use and related behaviors among Black men. Financial support: NIDA: K08-DA032296, PI:StevensWatkins;T32- DA035200, PI:Rush; & R01-DA011309, PI:Leukefeld. http://dx.doi.org/10.1016/j.drugalcdep.2016.08.351