Proceedings of the 16th annual conference of INEBRIA

© The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. A1 Active components of a web‐based personalised normative feedback: a dismantling study Andre Bedendo, Jim McCambridge, Jacques Gaume, Altay A. L. Souza, Maria L. O. Souza‐Formigoni, Ana R. Noto Department of Psychobiology, Universidade Federal de Sao Paulo, Sao Paulo, Brazil; Department of Health Sciences, University of York, York, United Kingdom; Alcohol Treatment Centre, Lausanne University Hospital, Lausanne, Switzerland; Department of Psychobiology, Universidade Federal de Sao Paulo, Sao Paulo, Brazil Correspondence: Andre Bedendo ‐ andrebedendo@gmail.com Addiction Science & Clinical Practice 2019, 14(Suppl 1):A1


Background:
The sub-population of patients with alcohol use disorders (AUD) is usually excluded from studies on Screening and Brief Intervention (SBI), although SBI might be effective in increasing motivation for behavior change. For that reason, the evidence is scarce among the general population and non-existent for pregnant women. The aim of this exploratory and secondary analysis of data (from an efficacy-randomized study) is to present the results of the acceptability of SBI among a group of pregnant women with criteria for AUD. Methods: 23 pregnant women with criteria for AUDs were identified in a probabilistic sampling of pregnant women who attended the Public Health Centers of Mar del Plata, Argentina, during 2016 (n = 893). Every participant received BI and referral to treatment. Screening was performed with the AUDIT (scores 16 were considered positive) and acceptability was assessed with four ad hoc questions. Consumption and related problems were evaluated three months later. Results: Of all the women who were contacted again after three months (n = 10), only one increased her AUDIT score, due to the number of standard units consumed per occasion, although episodes of binge drinking decreased. Of the nine participants who decreased their AUDIT scores, seven reported abstinence. All the participants stated that the questions were easy to answer; most of them stated that they learned something new and that they had shared the contents of the interview with others. Conclusions: Despite limitations, these results suggest that SBI may be well accepted among pregnant women with criteria for AUD.
Background: Homelessness and substance use often coexist, with each issue exacerbating the other. Both are prevalent among emergency department (ED) patients. Concurrent screening and intervention to prevent homelessness might enhance the effectiveness of ED-based SBIRT. Materials and Methods: We conducted interviews with a random sample of New York City (NYC) public hospital ED patients who screened positive for past year unhealthy alcohol or drug use [using single-item screening questions (Smith et al. 2009(Smith et al. , 2010]. Adult patients were eligible if they spoke English or Spanish, were medically stable, and not in prison/police custody. Using patient identifiers, data were linked to the NYC shelter administrative database, which captures 90% of NYC shelters. The primary outcome was shelter entry within 6 months of the baseline ED visit, among patients who were not already homeless at baseline. Results: Interviews were conducted with 1,262 unique ED patients with unhealthy alcohol or drug use who were not currently homeless. 8.5% had a shelter entry within the next 6 months. Self-judged risk of using a shelter in the next 6 months rated as "somewhat" or "very likely" had 53.3% sensitivity and 26.5% PPV for future shelter entry. A brief homelessness risk screening tool-developed via predictive modeling plus stakeholder feedback-comprising 3 yes/ no questions (shelter use in past year, applied for shelter in past 3 months, lifetime incarceration history), with an affirmative answer to any question considered a positive screen, had 85.0% sensitivity and 19.1% PPV. Conclusions: A brief screening tool identified ED patients with unhealthy substance use who were at risk for near-term homeless shelter entry with accuracy similar to screeners developed for other populations, and exceeding the sensitivity of self-assessed risk. If replicated, this screening tool or similar tools could be used to identify which patients with unhealthy substance use may need targeted homelessness prevention services.

A11 Exploring a Complex Relationship: A Qualitative Study of Substance Use and Homelessness
Amanda Jurewicz 1 , Deborah Padgett 2 , Ziwei Ran 2 , Donna G. Castelblanco 1 , Ryan P. McCormack 1 , Lillian Gelberg 3 , Donna Shelley 1 , Kelly Doran 1 1 NYU School of Medicine, New York City, NY, USA; 2 NYU Silver School of Social Work, New York City, NY, USA; 3 University of California, Los Angeles, CA, USA Correspondence: Kelly Doran -kelly.doran@nyumc.org Addiction Science & Clinical Practice 2019, 14(Suppl 1):A11 Background: Emergency department (ED) patients commonly face problems with both substance use and homelessness. Research has suggested a bi-directional relationship between substance use and homelessness, but most prior research has been quantitative and cross-sectional. Better understanding this relationship could inform the design of more responsive ED-based substance use interventions, including those that also address homelessness. Materials and Methods: We conducted in-depth, one-on-one interviews with ED patients who had become homeless within the past 6 months. Using a semi-structured interview guide, we asked patients about their pathways into homelessness and the relationship between their substance use and homelessness. Interviews, on average lasting 42 min, were digitally recorded and professionally transcribed. Transcripts were coded line-by-line by 2-3 investigators, who discussed and refined codes in an iterative fashion. The codes then formed the basis for thematic analysis and consensus discussions. ATLAS.ti was used to assist with data organization. Results: Of the 31 patients interviewed, 54.8% reported unhealthy alcohol use and 41.9% drug use in the past year; for others, substance use was only in the past. Five themes emerged: (1) substance use often contributes to homelessness as an upstream factor, through varied intermediary factors (e.g., job loss, family discord); (2) homelessness affects substance use variably, both increasing (e.g., due to depression) and decreasing substance use (e.g., due to lack of time); (3) substance use and homelessness sometimes share precipitants, often related to interpersonal factors; (4) substance use creates practical and environmental barriers relevant to homelessness (e.g., avoiding shelters that might trigger relapse); (5) homelessness can both promote and hinder entry into substance use treatment (e.g., may motivate "change"). Conclusions: Substance use and homelessness are intertwined in complex ways. ED-based substance use interventions should consider the high prevalence of homelessness and the variable ways in which homelessness affects substance use and vice versa. Background: Gambling disorder is a rare but serious disease, and most affected individuals do not seek treatment. Especially adolescents and young adults show a high prevalence of subclinical gambling involvement, indicating that early interventions measures might be promising. Aim of this study is to define suitable subgroups in a vocational student setting. Methods: An unselected and non-treatment-seeking sample (n = 6,781) has systematically been screened proactively in vocational schools in Schleswig-Holstein, Germany. Students with at-risk or pathological gambling behavior (n = 1,809) according to the Stinchfield questionnaire were asked to participate in three in-depth telephone interviews, one at baseline, followed by two interviews after Addict Sci Clin Pract 2019, 14(Suppl 1):27 approximately 10 and 20 months. A subsample of 405 potential participants were contacted for the baseline interview. Stability of gambling involvement and associated socio-demographic variables were analyzed. Results: The telephone assessments resulted in 309 valid baseline interviews (response rate 78.7%), 268 in the first and 227 in the second follow-up. Of the baseline sample, 43.4% (n = 134) showed at least subclinical gambling involvement (2 or more DSM-5 criteria). Participants with at risk/pathological gambling were significantly more often male (96.3%), had a migration background (72.4%), were single (66.4%), and had a lower school education (88.0%) compared to participants without gambling problems. Regarding the trajectory of gambling involvement, 44.3% reported deterioration of their gambling behavior over time or maintained at least subclinical symptoms from baseline to second follow-up. Conclusions: Students in vocational schools show elevated levels of problematic gambling patterns and can be successfully approached in this setting. Data show that symptoms of pathological gambling are stable in this population and therefore should be addressed using prevention measures. Response rates are comparable to other studies in the field of substance-related Brief Interventions (BI). Implementing BI targeting pathological gambling in vocational schools therefore seems to be a promising strategy. Background: Problematic and pathological Internet use is an important current topic in research and treatment of addictions. Especially adolescents and young adults constitute a vulnerable group of users. Therefore, the effectiveness of brief interventions (BI) to reduce problematic Internet use should be evaluated. Method: Vocational students were screened proactively with the Compulsive Internet Use Scale (CIUS) as well as with other impairment measures. Those with a CIUS score higher than 21 were asked for permission to be approached for a telephone interview. In case that at least two DSM-5 criteria for Internet use disorders were fulfilled during the in-depth interview, participants were randomized into an intervention and a control group. The intervention group received up to three counseling sessions based on Motivational Interviewing. After five and ten months follow-up assessments were conducted. Results: A total of 8,606 students were screened of which more than one-third (n = 3,142) showed problematic Internet usage. This subgroup significantly showed higher impairment in daily tasks and duties. In addition, approximately 80% were concerned to use certain applications too much. Among the 1,481 screening-positive subjects eligible for study participation, 934 interviews could be realized (Response rate 67%). Problematic or pathological Internet use was discovered in 55% (n = 507) of the interviews. In this ongoing study, vocational students' accessibility via mobile phone proved to be challenging. To realize one BI session several contacts were necessary. Conclusion: Vocational schools are an appropriate setting for offering brief interventions for pathological internet use due to elevated prevalence rates. However, motivation of students to participate in counseling sessions was limited. Brief interventions should be adapted for this target group for example by using smartphone applications. Trial registration: NCT03646448. At the request of operations partners, SPARC was rolled out alongside a Behavioral Health Integration initiative. Prior to SPARC, 19% of PC patients completed alcohol screening; there was no standardized assessment for alcohol use disorder (AUD). The intervention had three components: front-line PC support by practice coaches, electronic health record (EHR) tools, and performance feedback. Practice coaches had weekly QI meetings with each clinic's implementation team for ~ 6 months ("active implementation"). Coaches addressed clinical knowledge gaps, modeled destigmatizing language, and collaborated on EHR tools and performance metrics development. Following active implementation, operations partners continued quarterly QI meetings with PC teams (without coaches). We report findings from performance metrics for all 22 clinics at the end of SPARC (7/2018), and sustainment eight months later (3/2019): % completing alcohol screening among PC patients; and % completing standardized assessment of DSM-5 AUD symptoms among patients with high-risk alcohol screening scores. Results: There were 37,093 PC patient visits across the 22 clinics in 7/2018, and 44,954 in 3/2019. Alcohol screening rates were 88% in 7/2018 and 89% in 3/2019. AUD assessment rates of high-risk patients were 64% in 7/2018, and 70% in 3/2019. Conclusion: Enhanced practice coaching can lead to sustained improvements. Based on sustainment and staff/leader satisfaction, this implementation model has become a "gold standard" for this health system. Addict Sci Clin Pract 2019, 14(Suppl 1):27

A14
Background: To implement brief intervention for unhealthy alcohol use, SPARC tested state-of-the-art implementation strategies-practice coaching, electronic health record (EHR) decision support, and performance monitoring and feedback-in 22 clinics of Kaiser Permanente Washington (KPW) from 01/15 to 07/18 using a stepped wedge design. Primary results showed that the intervention significantly increased EHR-documented brief intervention, but rates were very low (5%). This presentation uses data from a state-wide patient experience survey conducted in the middle of the SPARC trial to report on and compare rates of patient-reported receipt of brief intervention at sites surveyed before, during or after active implementation. Methods: From 08/17 to 11/17, the Washington Health Alliance survey included questions assessing heavy episodic drinking (HED) and a question assessing receipt of alcohol-related advice (a key component of brief intervention). Sites were categorized into 3 groups, based on their randomly-assigned start date for the SPARC trial: those surveyed before, during or after implementation. For each group of sites, we calculated the percent of surveyed patients who reported alcohol-related advice ("% patient-reported brief intervention") among those reporting any HED, and compared % patient-reported brief intervention at sites surveyed before, during and after implementation, using Chisquare and test for trend. Results: Five sites were surveyed before SPARC implementation, 3 during and 13 after; % patient-reported brief intervention ranged 13.3% to 55.6% across sites. Rates of patient-reported brief intervention in groups of sites surveyed before, during and after implementation, respectively, were 40.9%, 47.9% and 39.2% (p-values for comparisons all > 0.05). Conclusions: Although rates of patient-reported alcohol-related advice were higher than those based on EHR documentation, no differences in rates of patient-reported brief intervention were observed before during and after SPARC implementation. As in the main trial, results support further quality improvement efforts to ensure patients with unhealthy alcohol use receive brief intervention. Background: Hazardous drinking (HD) is a major public health problem in India. However, healthcare access is limited by the shortage of healthcare professionals. Extensive global evidence demonstrates the effectiveness of technology-delivered BIs in reducing alcohol consumption. Our study aims to increase healthcare access for HD, by designing a contextually-appropriate mobile-based BI, and evaluating its acceptability, feasibility, and preliminary impact. Methods: Through a systematic review and in-depth interviews with experts and intended recipients, initial content areas for the intervention were derived. These were presented in a Delphi survey to 30 international experts, who rated each area on a five-point Likert scale. At the end of this two-stage iterative process, content areas that reached group consensus were synthesized to inform the intervention development. The draft intervention was then delivered in a case series to participants who screened positive for HD on the Alcohol Use Disorder Identification Test (AUDIT). At one-month follow-up, in-depth interviews were conducted to understand the acceptability and feasibility of the intervention. The preliminary impact was examined through changes in drinking parameters measured using the Timeline-Follow-Back (TLFB). Results: 26 content areas were derived from the systematic review and interviews, and 22 of those met Delphi consensus. The intervention is currently being delivered in the case series, and findings on acceptability, feasibility and impact will be ready for presentation at the conference. Preliminary follow-up interviews (n = 11) have indicated a preference for push messages and an app-based delivery. Higher number of messages was cited as an engagement deterrent, with three messages per week considered ideal. Conclusion: The content and delivery of the intervention will be iteratively refined during the case series, and the final package will be pilot tested through a randomised control trial. If demonstrated to be effective, the intervention will change the landscape of interventions for HD in resource-constrained settings. Background: Although BI has shown to be effective among university students in high income countries, little research has been done in Latin-America. Furthermore, evidence examining moderators of intervention efficacy is scarce. Certain characteristics that make alcohol more easily available to students, such as living outside of parental control or having economic autonomy to spend money on alcohol, could moderate BI effectiveness. The objective of this study is to evaluate the moderator role of the living arrangements and the employment situation on BI effectiveness. Materials and methods: Participants were 473 students from Mar del Plata National University (60% women, 40% men; between 17 and 46 years old (M = 20.34, SD = 3.9)). Prospective participants were screened and those with high-risk alcohol consumption in the last 12 months were randomly assigned to a control group or BI. After 3 months, they were re-assessed. The measures were: effectiveness (i.e. decrease in AUDIT scores (yes/no)), employment situation (work: yes/ no) and living arrangements (living with family: yes/no). Fisher´s exact test was used to analyze the moderator effects of living arrangements and employment situation on effectiveness. Logistic regression analyses were performed in order to control the possible effect of age. Results: 76% of students lived with their families, while 24% lived alone or with friends; 42% of the students were employed. Living with family moderated (increased) BI effectiveness (9,310, p = 0.01). Similarly, not having employment (i.e. being supported by family) also moderated (increased) BI effectiveness (7,611, p = 0.02). These moderator effects were not accounted for by age. Conclusions: Living arrangement and employment moderated effectiveness of BI, suggesting that restricted access to alcohol may improve the effectiveness of interventions among university students.

A18
Who are the users of the Brazilian self-help intervention program "Bebermenos" (drink less) who accepted to participate in a RCT to evaluate its effectiveness? use and related problems. Objective: To describe the profile of users of this web-based intervention. Materials and methods: From September 2016 to January 2019, 579 users who filled out the AUDIT and were considered risk users, or possibly dependent, participated in a Randomized Clinical Trial to evaluate the effectiveness of the intervention. Out of 579, 281 were randomly allocated to the experimental group (virtual Brief Intervention) and 298 to the control group (waiting list). Six months after admission, follow-up was conducted. Results: Of the total sample, most participants (61.1%) were men and 50% were between 33 and 44 years old (median = 36 years). Regarding the classification based on AUDIT scores, 17.8% were risk users (zone II), 16.4% presented harmful/hazardous drinking (zone III) and most (65.8%) were classified as possible dependence users (zone IV). The mean AUDIT total score was 22 (SD = 6.8) and participants reported having consumed about 37 (median) standard drinks in the week prior to entering the program. Regarding the Readiness to Change questionnaire scores, most participants were classified in the contemplation phase. Discussion: Although designed for at-risk users, most of the users who registered on the site already had severe alcohol-related problems. These data suggest there is a hidden population that should be under treatment for alcohol dependence, but refuse to do it, do not look for it or even do not find available treatment. Internet interventions could help these people by raising awareness of their alcoholrelated problems and encourage them to enter the action phase and look for treatment. Trial registration: ISRCTN14037475. Background: Low income countries are faced with a growing challenge of negative alcohol and substance use, yet interventions are scarce and rarely documented. Uganda is estimated to have 3,900,000 people with alcohol use disorders, yet the country's main source of treatment for addictive disorders is the National Mental Referral Hospital situated in the capital city. The aim of this paper is to highlight brief intervention as an alternative treatment for Alcohol and Substance Use Disorders (SUD), the likely challenges and potential solutions for this strategy in a low resource setting. Methods: As a way of evolving culturally appropriate services, Hope and Beyond conducted a 5 days' residential camp to treat and sensitize communities about SUD. The pilot treatment camp was held at Kisigula Health Centre (HC) II in Wakiso District; a metropolitan area that houses many city dwellers and nationals from Uganda and surrounding countries. Residents in the program catchment area were also mobilized to contribute towards logistical needs of camp participants. Camp activities included screening and assessments, detoxification and medications; psychotherapies; HIV counseling and testing, sensitization workshops, prayers/spiritual support and referrals. Challenges faced ranged from logistical to human resource constraints, yet many clients were in severe physical and mental condition and low on motivation. Results: Although the project was planned for 20 participants, 53 clients turned up and were treated for SUD, 12 health workers from nearby medical centers were trained in addiction management and sensitization was conducted in 11 Churches and 2 Mosques reaching out to over 10,000 people. Conclusion: Camp treatment as a way of brief intervention for alcohol and substance use disorders is a promising practice for alternative SUD treatment that should be adapted in low income countries but scientific studies are necessary to establish its effectiveness. Background: Brief intervention is promoted as a cost-effective measure to reduce alcohol consumption and prevent alcohol-related harm. However, there is scant evidence on its implementation in developing countries. This paper offers insight into the clinical treatment needs of alcohol and other substance users in low income countries by reporting the prevalence of alcohol and drug use disorders and the co-occurring illnesses amongst residential treatment participants in Kampala, Uganda. Methods: 53 participants (50 males and 3 females) reporting for a 5 day residential treatment camp were interviewed regarding their perception of their physical and mental wellbeing, and screened for drug use disorder(s) using the Alcohol and Substance use Screening Involvement Tool. The treatment camp was held in the Wakiso District near the capital city of Uganda, Kampala. Results: Alcohol was the most commonly consumed drug, used by 66% of participants, followed by nicotine and cannabis at 23% and 8% respectively. 43% were diagnosed with a single substance use disorder and the remainder had multiple use disorders, of which, 36% reported addiction to two drugs and 18.3% used three or more drugs. 86% reported co-occurring physical medical conditions such as fever, sexually transmitted diseases (STI) and/or a cough; 40% reported psychiatric symptoms such as psychosis, insomnia and bipolar; and 57% reported psychological symptoms such as depression and anxiety. Conclusion: Participants of brief intervention treatment for alcohol and substance use in low income settings have varying needs resulting from multiple drug use disorders and other co-occurring medical, psychiatric and psychological illnesses. Brief interventions for alcohol and substance use disorders delivered in free treatment camps should include a range of additional services to meet the participants' varying multiple and complex needs. Further research is necessary to establish culturally sensitive effective treatment approaches and modalities of brief interventions in developing countries. Background: Alcohol use has been shown to increase an individual's likelihood to experience a traumatic injury, particularly those who misuse, use at risky levels, or have a use disorder. Based on this information, screening and brief counseling interventions (BCIs) can be particularly impactful in hospital trauma center settings to address these patterns. Recently, efforts have grown to understand health disparities in order to provide the best treatment possible. This study aimed to understand how health disparity related factors, namely age, geographic location, insurance type, and sex, as well as two distinct BCIs uniquely predict changes in drinking patterns. Methods: This study was a retrospective analysis on a pre-existing dataset collected from hospitalized trauma patients to evaluate two different BCIs for patients with alcohol-related injuries. The initial study found that both quantitative BCIs and personalized BCIs were effective in reducing self-reported drinking patterns at a six-month follow up with the AUDIT screening tool. This retrospective analysis sought to add depth to these findings by understanding the impact that the aforementioned factors have on predicting these same drinking patterns. After statistically controlling for other variables, each unique Addict Sci Clin Pract 2019, 14(Suppl 1):27 health disparity factor and intervention type was tested through hierarchical regressions to determine its contribution. Results: Results demonstrated that regardless of health disparity factors or BCI type, on average, patients displayed reduction in drinking patterns at the six-month follow up. These results also indicated that though all patients experienced improvements, females were more likely to show greater changes than men in either intervention. Conclusion: Overall, this study supports and further highlights the evidence that the use of BCIs containing more innovative and empathybased approaches are appropriate for impacting positive patient change behaviors. In addition, these study results show that this more flexible approach is appropriate in sub-populations that are more likely to experience health disparity.  (Hawton et al. 2013). Addressing both substance use and suicidal ideation is critical in integrated care settings, as those who have been previously hospitalized following suicide attempts are at significantly greater risk of subsequent attempts and hospitalizations (O'Connor et al. 2015). The aim of this preliminary study is to demonstrate the results of providing Screening and Brief Interventions (SBI) for patients who have suicidal ideation (SI) or were hospitalized due to a suicide attempt. Materials and Methods: This preliminary study looked at the feasibility of SBI for substance use and suicide in a Level-1 trauma center. A sample of patients were identified based on hospitalization due to suicide attempt or SI comorbidity with substance use, and were placed in either a control group or intervention group to receive a specialized SBI. Hospital readmission rates were measured to identify the impact of these interventions. Results: Of the intervention group, none of the patients were readmitted to the hospital within thirty days due to repeated suicide attempts or SI. In contrast, within the control group, 37.5% were readmitted within thirty days of their initial hospitalization, of which 25% were hospitalizations related to additional suicide attempts or SI. Conclusions: Preliminary findings support a positive trend: patients that received specialized SBIs addressing both substance use and SI show a reduction in hospital readmissions. Given these findings, further research is warranted regarding the effectiveness of specialized SBIs in patients who have attempted suicide or active SI. Background: This study's purpose was to determine whether SBIRT interventions by professional counselors working on inpatient integrated care settings are effective treatments for alcohol and illicit drug misuse and disordered use. Effectiveness was determined by evaluating the association between interventions and subsequent hospitalizations and emergency department visits. Inpatient settings were selected for study given inconclusive results from prior research for this category of SBIRT recipient. This study controlled for type (alcohol, illicit drugs, or both) and severity of substance use, with inpatient clinical service as a clustering variable. Materials and methods: Using a difference-in-differences approach and generalized linear mixed modeling, 1,577 hospitalized patients receiving SBIRT interventions were compared to 618 patients identified for but not receiving interventions, for a single U.S. hospital over a four-year period. Utilization data were collected one year prior to and following the identifying hospitalization, along with substance use type and severity and clinical service. Propensity scores were developed from demographic, disease, and insurance indicators and used as covariates. Results: On average, patients receiving counselor-provided SBIRT interventions experienced 22% fewer subsequent hospitalizations and emergency department visits than patients not receiving interventions, controlling for substance use type and severity. Outcomes varied significantly across inpatient clinical services. The study sample was 74% male and 73% White, with a mean age of 44.7 years. Conclusions: The study tested a novel substance use treatment model, counselor-provided SBIRT, for a population with a wide spectrum of substance use types and levels of severity. The results offer support for this process as an effective treatment model for reducing utilization of hospitalizations and emergency department visits. Given these findings, health system administrators, physicians, and community leaders may support integrating professional counselors into hospital units and other medical settings, raising the likelihood that people who need help with their substance use actually receive it. Background: As the largest profession in the healthcare workforce, nurses have important roles in moving evidence related to Screening, Brief Intervention, and Referral to Treatment (SBIRT) to action to address the global burden associated with alcohol and other drug (AOD) use. The purpose of this quality improvement project was to evaluate an online self-paced educational program, Screening, Brief Intervention, and Referral to Treatment (SBIRT) for Healthcare Providers among nurses in an ambulatory care facility with the goal of increasing their knowledge in the screening and management of patients with AOD use. Methods: A one-sample, pretest-posttest design was used in this project. The Wilcoxon Signed Rank Test was used to analyze results from the SBIRT-related knowledge test from before to after the intervention. Descriptive statistics were used to analyze data related to previous SBIRT education and confidence to deliver SBIRT in practice. Thematic analysis was used to categorize barriers to and facilitators for SBIRT implementation. Results: There was a significant increase in SBIRT-related knowledge (p < .001) from before to after the intervention. A high proportion of the nurses had no SBIRT knowledge (45%) prior to the intervention. Nurses reported high confidence levels to screen for alcohol and drugs after the intervention. Barriers to and facilitators for SBIRT implementation related to five themes: (1) time, (2) education, (3) resources, (4) receptivity and (5) interprofessional collaboration. Conclusion: It was feasible to deliver this online SBIRT education to nurses at a busy ambulatory care facility and impactful in terms of increasing SBIRT-related knowledge and confidence. To promote system-wide readiness for widescale dissemination, providing this online program to other ambulatory care clinics and other healthcare professionals is warranted. Background: Primary care clinics often struggle to choose the approach to alcohol and drug screening that is best suited to their resources, workflows, and patient populations. We are conducting a multi-site study to inform the implementation and feasibility of electronic health record (EHR)-integrated screening. Methods: In two urban academic health systems, researchers worked with stakeholders from 6 clinics to define and implement their optimal screening approach. All clinics used single-item screening questions for alcohol/drugs followed by AUDIT-C/DAST-10. Clinics chose between: (1) screening at routine vs. annual visits; and (2) staffadministered vs. computer self-administered screening. Results were recorded in the EHR, and data was extracted quarterly to describe implementation outcomes including screening rate and detected prevalence of unhealthy (moderate-high risk) use among those screened. Findings are from the first 3-12 months post-implementation at each clinic. Results: Across sites, of 84,311 patients with primary care visits, 58,492 (69%) were screened. In the 4 clinics with mature (9-12 months) implementation, screening rates ranged from 42 to 95%. Rates were lower (10-22%) in the 2 clinics that recently launched. Screening at routine encounters, in comparison to annual visits, achieved higher screening rates for alcohol (90-95% vs. 42-62%) and drugs (90-94% vs 38-60%). Staff-administered screening, in comparison to patient self-administered screening, had lower rates of detection of unhealthy alcohol use (2% vs. 15-37%). Detection of unhealthy drug use was low, ranging from 0.3 to 1.5%. Conclusions: EHR-integrated screening was feasible to implement in at least 4 of the 6 clinics; 1-year results (available Fall 2019) will determine feasibility at all sites. Self-administered screening at routine primary care visits achieved the highest rates of screening and detection of unhealthy alcohol use. Although limited by differences among clinics and their patient populations, this study provides insight into outcomes that may be expected with commonly used screening strategies in primary care. ClinicalTrials.gov identifier: NCT02963948. Background: Web-based programs for substance use have been designed, evaluated and implemented over the past two decades in high-income countries. The development of such tools in Latin America is more recent. One of the few available web-based programs is the Programa de Ayuda para Abuso de Drogas y Depresión (PAADD). Its feasibility was demonstrated through a randomized trial. The next step is to design a strategy to promote its implementation.

Methods:
This study aimed at identifying the factors involved in the implementation of technological innovation in Mexico. The level of readiness to adopt technologies for health-care provision was measured with an adapted version of the Telehealth Capacity Assessment Tool (TCAT), which considers 6 domains: organizational, technology, regulatory, financial, clinical, and workforce factors. Free training was offered at 12 substance use prevention and treatment institutions. The managers were asked to complete the TCAT before the training. Results: We received eight completed questionnaires: four from treatment centers affiliated to Psychology Schools in two Universities; one from an immune-infectious clinic, and three from Primary Care Centers for Addictions (PCCA). Additionally, professionals who were trained provided information about: internet use; academic background; experience in substance use treatment and attitudes towards the use of technology. The highest TCAT scores (3-4.5) were observed at the clinic, showing a high degree of readiness to implement web-based programs, the lowest scores (0-2.5) belong to the PCCA, where the implementation is challenging. Conclusions: The information provided by the professionals indicates a negative attitude towards technology and less success in enrolling clients in web-based program at institutions with a low TCAT score, while professionals at institutions with moderate scores were more successful at enrolling and had positive attitudes. The data is relevant to create a dissemination strategy to approach the misconceptions about web-based interventions and facilitate its acceptance as a valid therapeutic alternative. Background: We developed a fully automatized computer-based intervention to address alcohol consumption and depression simultaneously. In the present paper, we report an initial proof of concept trial. Methods: Participants were recruited via a multicenter screening program approaching adult patients from ambulatory practices and hospitals. Inclusion criteria were hazardous alcohol consumption and an episode of subclinical or clinical symptoms of depression in the past year. Patients with current severe depression or indication of alcohol dependence were excluded. In total, 132 participants were randomized to an assessment only control or an intervention group receiving six individually tailored motivational feedback letters and weekly text messages over a period of 6 months. Intervention content was constructed based on the principals of the Transtheoretical Model of behavior change. Outcome was assessed by computer-assisted telephone interviews scheduled 6, 12 and 24 months after baseline. Results: Preliminary analyses were based on data from 6-(n = 104) and 12-month (not completed, current state of April 2019: n = 107) follow-ups. Generalized estimating equation analysis adjusting for recruitment setting, age, and sex revealed a significant decrease in depression scores (p < .01) and no significant time effect for alcohol measures. After 12 months, changes in alcohol and depression measures were numerically larger in the intervention compared to the control group, with small to medium effect-sizes (Cohen's d: heavy Addict Sci Clin Pract 2019, 14(Suppl 1):27 drinking days = 0.36, mean daily consumption = 0.25, depression score = 0.29), but statistical significance was only reached for frequency of heavy drinking days (t-test, one-sided p = .03) Conclusions: The intervention and research logistic proved to be technically feasible. Based on our preliminary analysis, effects seem comparable to single focused motivational interventions among unselected samples. Thus, a future adequately powered effectiveness trial is warranted. Given the low baseline motivation to adopt healthy behaviors final conclusion on effectiveness should be postponed to the availability of long-term outcome data. Trial registration: German Clinical Trials Register DRKS00011635. Background: The public health impact of brief alcohol interventions (BAIs) might be increased by approaching an entire population rather than selected high-risk individuals only. In this study, all persons who drink alcohol were offered BAI, including those identified as having low-risk alcohol use or with greater severity. The aim was to investigate the BAI efficacy during the active intervention phase as a function of alcohol use severity. Methods: In our ongoing randomized controlled trial (http://www. drks.de/DRKS0 00142 74), we systematically screened all persons aged 18-64 years appearing in the waiting area of a local registration office over a period of two months. Those who reported alcohol use in the past 12 months (n = 1,648) were randomized to BAI or assessment only. BAI consisted of computer-generated individualized feedback letters delivered at baseline, month 3, and month 6. Latent growth modeling was used to test BAI effects through the 6 months of intervention as a function of the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) score. By the cut-off date for this analysis, 6-month assessments have been completed and two of three interventions have been delivered. Results: The trial participation rate was 67%. Three-and 6-month retention rates were 85% and 81%, respectively. Participants with lower AUDIT-C scores were more likely to participate in the trial (OR = 1.07, p = 0.010) and in multiple BAIs (OR = 1.11, p = 0.003) than those scoring high on the AUDIT-C. At month 6, BAI produced significant changes in the number of drinks per week among participants with low AUDIT-C scores (IRR = 0.83, p = 0.035). Effects decreased with increasing AUDIT-C scores (IRR = 1.04, p = 0.048).

Conclusions:
We provided a computer-based BAI that may be particularly appropriate for the large but understudied group of persons with low severity. Twelve-month data need to be included in analyses before we can draw more definite conclusions about its efficacy in the population as a whole. Trial registration: German Clinical Trials Register DRKS00014274. Background: Population-based screening of alcohol use disorder (AUD) are crucially needed for public health planning. Evidence-based measurements are needed, but empirical studies comparing different self-reported measures to a gold standard are scarce. This study aimed at identifying a valid screening tool. Methods: This Swiss controlled study collected data among young men from the ongoing Cohort Study on Substance Use and Risk Factors, using a stratified random sample selection (n = 233). AUD was diagnosed using the Diagnostic Interview for Genetic Studies (gold standard). Self-reported measures included criteria of AUD, alcoholrelated consequences, and previous twelve-month alcohol use. We tested psychometric performances of the self-reported measures deriving sensitivity and specificity from receiver operating characteristics curves and using all possible subsets of questions. Results: Taken separately, none of the self-reported measures displayed good psychometric properties, maximizing sensitivity and specificity. This was true for the self-reported AUD (cut-off of two or more symptoms: sensitivity = 92.3%, specificity = 45.8%; cut-off of four or more symptoms: sensitivity = 60.3%, specificity = 87.1%) and alcohol use (cut-off of 10 drinks per week: sensitivity = 85.9%, specificity = 55.5%; cut-off of 21 drinks per week: sensitivity = 38.5%, specificity = 93.6%). The best model combined 8 self-reported AUD criteria and 4 alcohol-related consequences. With a cut-off of 3, this screening tool displayed good sensitivity (83.3%) and specificity (78.7%). Conclusions: These findings provided important insights among young men in current debate in the alcohol field: heavy alcohol use was not a suitable single criterion to assess AUD and consequences were important to identify a valid assessment. Even if alcohol use is not part of the final screening tool, it should not be neglected, as it is responsible of a large burden of disease and detrimental health consequences.
Background: Empirical studies on the quality of self-reported alcohol use as measures of excessive drinking compared to objective measures, such as ethyl glucuronide (EtG), are needed. In addition, associations of EtG with risky single occasion drinking (RSOD, i.e., ≥ 6 drinks on a single occasion) have been scarcely investigated. This study tested whether self-reported measures of alcohol use and RSOD allow detecting excessive chronic drinking as assessed by EtG. Methods: Data were collected among young Swiss men, recruited in the ongoing Cohort Study on Substance Use and Risk Factors using a stratified random selection. Assessments included self-reported measures of alcohol use (previous twelve-month and previous-week) and RSOD. Capillary blood was collected to determine EtG (n = 227). Data were analyzed using receiver operating characteristics curves, using EtG (cut-off of 30 pg/mg) as the gold standard of excessive drinking. Addict Sci Clin Pract 2019, 14(Suppl 1):27 Sensitivity and specificity were computed. We also performed a multivariate logistic regression to test whether alcohol use and RSOD were uniquely associated with EtG. Results: Overall, 23.4% of the participants presented a chronic excessive drinking according to the EtG cut-off of 30 pg/mg. For previous twelve-month alcohol use, a cut-off > 15 drinks per week yielded acceptable psychometric performance (sensitivity = 75.5%, specificity = 78.7%). No cut-off maximized sensitivity and specificity for previous-week alcohol use. Weekly RSOD detected EtG with acceptable psychometric properties (sensitivity = 75.5%, specificity = 70.1%). Sensitivity and specificity were respectively maximized for monthly RSOD (sensitivity = 94.3%) and daily RSOD (specificity = 98.9%). In the multivariate logistic regression, both previous twelve-month alcohol use with a cut-off of 15 and weekly RSOD were significantly associated with EtG (respectively p < .001 and p = .022). Conclusion: Self-reported measures of RSOD and of previous twelvemonth alcohol use were acceptable measures of excessive drinking for population-based screening. Self-reported RSOD appeared as an interesting screening measure to identify accurately excessive drinking among young people. Background: Systematic reviews suggest that cost-effectiveness evidence for ABI in emergency care and hospital settings is scarce and that the cost-effectiveness evidence for primary care is not based on a consistent set of economic outcomes. These reviews, as well as future economic evaluations of ABI, are hampered by the lack of a consensus on what economic outcomes should be measured in ABI evaluations. In this presentation, we present a preliminary methodology to establish a core outcome set (COS) for economic evaluations of ABI. This COS is intended to be supplemental to the COS developed for ABI trials by the INEBRIA Research Measurement Standardization Special Interest Group (RMS-SIG). Methods: We present a rapid review of the ABI economic evaluation literature as a first step towards developing an ABI economic COS. Our review began by first mining the existing Outcome Reporting in Brief Intervention Trials: Alcohol (ORBITAL) systematic review database to assess outcomes use in previous economic evaluations of ABI trials. We then supplemented the ORBITAL review with a rapid review specifically designed to identify any gaps in our literature database. Results: We find that ABI economic evaluations seldom use consistent measures, but an increasing number of studies report quality adjusted life years (QALYs) in addition to measures of social costs. Studies suggest that the RMS-SIG should consider measures of: health state utility as derived from health-related quality of life; health care use; injuries and accidents, including motor vehicle accidents; crime and criminal justice involvement; employment, workplace productivity, and absenteeism; and, for adolescent studies, educational outcomes such as school attendance and matriculation. Conclusion: To support the development of a rigorous evidence base for the economic benefits of ABI, the RMS-SIG should develop a core set of economic outcome measures that build on the ORBITAL COS. Standardization Special Interest Group (RMS-SIG), progress has been made to establish a core outcome set to assess ASBI effectiveness and efficacy. However there remains a need to identify which implementation outcome measures are most appropriate for this field. This presentation will identify which outcomes and associated measurement instruments are currently employed in implementation-focused research overall, and consider their potential applicability to ASBI. Methods: We scrutinised existing systematic reviews (Proctor 2011; Lewis 2015; Khadjesari unpublished) to identify outcomes and associated measurement instruments employed in implementationfocused research. We also searched databases of outcomes assessed in the existing ASBI trial literature (Shorter et al. in press) to determine whether any appropriate implementation outcomes/measures were included. Results: Proctor's taxonomy identifies a core set of implementation outcomes (acceptability; adoption; appropriateness; cost; feasibility; fidelity; penetration; sustainability). However whilst previous research has employed various implementation outcome measures with relevance to mental, behavioural and/or physical health (n = 154 studies), most assess intervention acceptability (n = 77) and/or adoption (n = 27), and are of relatively low psychometric quality. There is a particular need to develop instruments to assess feasibility, appropriateness and sustainability. Current ASBI trials do not collect implementation outcome data or employ relevant measurement instruments when assessing effectiveness. Conclusions: Recognised outcome taxonomies exist to support efforts to improve the quality and consistency of ASBI implementation research. However, there is an identified lack of robust instruments to support their measurement. Future research is needed to evaluate the status of ASBI implementation research, and to develop/validate instruments relevant to the field.

Development of a COS for implementation studies on alcohol brief interventions
Background: School-based health centers (SBHCs) have emerged as important clinical settings in the US for expanding access to healthcare for underserved adolescents. SBHCs could hold promise as sites in which to deliver brief intervention (BI) for substance use. Methods: Participants were adolescents aged 14-18 who screened positive for risky cannabis and/or alcohol use on the CRAFFT screener at two urban school-based health centers (SBHCs). A sample of adolescents were enrolled in a randomized trial of computer-vs. nurse practitioner-delivered BI (N = 300). Additionally, in the year prior to launching the trial, we enrolled an assessment-only cohort of adolescents using the same recruitment protocol and inclusion/exclusion criteria (N = 50). Participants completed assessments at baseline, 3-, and 6-month follow-up. The current study compared outcomes for the BI conditions with the historical assessment-only cohort. Frequency of cannabis, alcohol, unprotected sex, and sex while intoxicated at follow-up were examined using negative binomial regression, controlling for participant sex, age, clinic site, and baseline value of the outcome. Results: There were no significant differences between computer-and nurse practitioner-delivered BI conditions on reported past-30-day frequency of cannabis, alcohol, unprotected sex, or sex while intoxicated. At 3-month follow-up, the pooled BI conditions had lower past-30-day frequency of alcohol use (IRR = .43; 95% CI = .29, .64; p < .001) and cannabis use (IRR = .74; 95% CI = .57, .97; p = .03) than the assessmentonly cohort. At 6-month follow-up, the pooled BI conditions had lower frequency of alcohol use (IRR = .58; 95% CI = .34, .98; p = .04) and sex while intoxicated (IRR = .42, 95% CI = .21, .83; p = .01) than the assessment-only cohort. Conclusions: Although we found no differences between two approaches to delivering BI at SBHCs on the outcomes considered, on average, participants who received a BI reported greater behavioral risk reductions than participants in a recent historical cohort that received no intervention. Trial registration: NCT02387489. Background: The proposed presentation will include an overview of an on-going stepped-wedge randomized trial of adolescent SBIRT being conducted at 4 rural health centers in the United States. The study examines the effectiveness of delivering the full range of provider interventions using the FaCES (Facilitating Change for Excellence in SBIRT) approach. FaCES includes a prescribed set of responses (anticipatory guidance, abbreviated BI, full BI) based on S2BI screening results. Adolescent patients, age 12-17 years, will receive either the FaCES intervention or standard care, depending on when their provider is randomized to begin delivering the intervention. Methods: As of March 31, 2019 a total of 621 patients had been recruited into the study across 4 rural U.S. health centers. Of those, 381 standard care control condition participants had also completed their 3-month follow-up interview, which included a re-administration of the S2BI assessing past 90 day substance use. Paired-samples t-tests were conducted to compare baseline with follow-up self-reported past 90 day use of tobacco, marijuana, and alcohol. Results: Among the 381 standard care participants in the follow-up sample, 52% were female, 67% were white, and 34% were Hispanic. Significantly higher rates of alcohol use in the past 90 days were reported at follow-up (M = .366, SD = .66) than at baseline (M = .276, SD = .54); t(379) = − 3.4845, p = .0006. No significant differences were noted in past 90 day use rates of either tobacco or marijuana. Conclusions: In order to interpret the effectiveness of adolescent SBIRT interventions it is important to include a standard care arm to track changes in substance use, which can fluctuate rapidly during adolescence. Early data indicate that reported alcohol use may increase for patients receiving standard care over the 3-month follow-up period. Background: Polysubstance use is common among people living with HIV infection (PLWH) and substance use disorder (SUD) but its effects are under-studied. We aimed to (1) identify polysubstance use patterns over time with latent class analysis, and (2) assess their associations with HIV disease severity. Methods: We studied a prospective cohort of 233 PLWH who also had SUD. Latent class analysis identified polysubstance use patterns based on the Alcohol Use Disorders Identification Test (consumption) and past 30-day use of cannabis, cocaine, opioids, and tranquilizers. We categorized changes in substance use patterns over 12 months and tested associations between those changes and CD4 cell count and HIV viral suppression at 12 months in linear and logistic regressions, adjusting for demographics. Results: At baseline, three patterns (classes) were identified: 18% did not use any substance (NONE), 63% used mostly cannabis and alcohol (CA), and 19% used mostly opioids, cocaine, tranquilizers, cannabis and alcohol (MULTI). At 12 months, 61% were in the same class. Forty percent decreased the number of substances used (MULTI to CA, either to NONE) or remained as NONE; 43% were in CA both times; and 17% increased (NONE to CA or either to MULTI, including remaining MULTI). Adjusted mean CD4 count was lower among participants increasing substance use (mean [95% CI] 446 ) and among those in Background: Drug use is common in people with alcohol use disorder (AUD) and opioid use can preclude use of naltrexone to treat it. The aim of this study was to describe substance use, and specifically opioid use, in adults with AUD, who were eligible to start naltrexone during hospitalization. Methods: Adult inpatients with AUD (DSM5) and at least one pastmonth heavy drinking day (HDD) who had no naltrexone contraindications were enrolled in the Alcohol Disorder hOsPital Treatment (ADOPT) randomized trial comparing oral and extended-release naltrexone at discharge. AUD was assessed by the following: AUD and Associated Disabilities Interview Schedule-5 (AUDADIS-5); past 30-day alcohol use by the Timeline Followback; and past 3-month other drug use by the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Results: Of 821 patients screened who met AUD criteria, 11% were excluded due to opioid use. Among the first 176 participants enrolled, 82% were men, 48% black, 42% white; 13% were Hispanic; mean age was 50 ± 10 years. Participants reported mean 11 ± 11 standard (14 g) drinks/day, 20 ± 10 HDDs/month, mean percent HDD 68 ± 32. Almost half (48%) reported cannabis use and 27% reported cocaine use; 41% and 31%, respectively, had a moderate or high risk ASSIST specific substance involvement score. Conclusion: Illicit drug use is common among medically hospitalized patients with alcohol use disorder. However, opioid use specifically only excludes a small minority of potential patients from receiving naltrexone for their AUD. Nevertheless, as drug use and disorder may affect prognosis and treatment selection, it should be considered in treatment planning. Background: According to the behavioral economics framework, substance use is more likely when constraints on use are minimal and when there are important constraints on access to substance-free reinforcers. For example, alcohol is a potent reinforcer, but its consumption is sensitive to constraints on access (including drink price) and the presence of alternative reinforcers. The alcohol purchase task (APT) presents a scenario and asks participants how many drinks they would purchase and consume at different prices. It has been used among students and small clinical samples and has not been tested using longterm prospective design. Methods: We administered the APT to a large sample of 4790 Swiss young men from the general population. Among those, 4326 (90.3%) were successfully followed-up 4 years later [mean age 21.4 and 25.4 (sd = 1.3)]. Parameters derived from the APT at baseline were used to predict weekly drinking, monthly binge drinking, maximum drinks in one occasion, alcohol-related consequences, and DSM-5 alcohol use disorder criteria. Results: Intensity (planned consumption when drinks are free) and Omax (maximum alcohol expenditure) were significantly correlated with all outcomes (r range: 0.25-0.37 for Intensity, 0.17-0.28 for Omax, all p < 0.001). Breakpoint (price at which consumption was suppressed) and Elasticity were significantly, but weakly correlated with outcomes (r range: 0.09-0.12 for Breakpoint, − 0.07 to 0.11 for Elasticity, all p < 0.001). Pmax (price at which demand became elastic) was not a significant predictor. Regression analyses controlling for baseline value of outcome showed consistent findings. Conclusions: Behavioral economics measures are useful in characterizing alcohol demand in young men from the general population and have long-term predictive value. Integrating behavioral economics components in BI models has been proposed but seldom tested. Potential for this approach will be discussed.

A39 Behavioral economics indices predict alcohol use and consequences in young men at 4-year follow-up -a target for brief intervention?
Background: Many students come to university relatively inexperienced with alcohol, which may increase alcohol-related consequences. Several interventions exist to combat this problem. However, recent research suggests these interventions may not be as successful as initially thought. The current study investigates the use of a relatively new web-based intervention, grounded in Deviance Regulation Theory (DRT), aimed at increasing alcohol Protective Behavioral Strategies (PBS). Methods: College freshmen participants (n = 157) were randomly assigned to one of three conditions: a positive message about individuals who use PBS, a negative message about individuals who do not use PBS, or an attention control condition. Participants then completed weekly assessments examining alcohol-related behaviors for six weeks. Participants also reported norms of PBS use each week. Results: Findings replicated previous research, showing actual PBS use increases across time among those with initially high PBS norms who also received a negative message about non-PBS use (b = 0.120, p = .022). Further, there was an increase in PBS norms across time (b = 0.023, p = .019). The growth in the effectiveness of the negative message was related to increases in PBS norms across time (r = .129, p < .001). There were no immediate effect of the positive message. However, within-subjects analyses showed that within a given week, the positive message was associated with increased weekly PBS among those with low weekly PBS norms, a finding consistent with DRT prediction. Conclusion: These results suggest that DRT works by (a) increasing PBS use across time among those who receive a negative message by also increasing PBS norms and (b) increasing PBS use at the event level as a function of current PBS normative beliefs. The results indicate an in-the-moment DRT intervention may be beneficial for firsttime-in-college students. Background: Young people (18-24 years) are more likely to 'binge drink' and increase their health risks than other age groups. There is growing evidence that health staffs' attitudes affect alcohol service delivery. This research aims to identify alcohol use amongst UK final-year pharmacy undergraduates and explore if alcohol consumption and other characteristics are related to perceptions to supporting those with alcohol problems. Methods: Mixed-methods approach was used to: screen students' drinking using the Alcohol Use Disorders Identification Test (AUDIT); measure attitudes to support patients using a modified version of the Short Alcohol and Alcohol Problems Perception Addict Sci Clin Pract 2019, 14(Suppl 1):27 Questionnaire (SAAPPQ); and explore students' perceptions toward supporting drinkers through a focus group. Results: 54 students (44 female, age-range 21-25 years) from 106 participated (51% response). Mean AUDIT was 3.46 [S.D. ± 4.58]; 19% (N = 10) were higher risk (AUDIT ≥ 8), mean AUDIT = 11.4 [S.D. ± 3.01]; and 82% (N = 44) were low risk drinkers (AUDIT ≤ 7), mean AUDIT = 1.66 [S.D. ± 2.36]. Median total attitudes (most positive = 7 and least positive = 1) was 3.8 (range 4.2 to 3.5), close to neutral. Non-parametric tests identified higher-risk drinkers had significantly higher work-satisfaction (P = 0.013) and total positive attitudes toward this patient group (P = 0.013) compared to low-risk drinkers. Also, smokers (N = 8) had significantly higher work-satisfaction (P = 0.034) and total positive attitudes (P = 0.034) compared to non-smokers (N = 46). Focus group (involving 8 students) identified themes relating to work environment, social influences, alcohol education and stigma as possible factors relating to alcohol use and support. Conclusions: Most students were low-risk drinkers with neutral attitudes to supporting patients. Students wanted further alcohol education and counselling skills during their pharmacy degree to support this patient group for future clinical practice. Smokers and higher-risk drinkers had more positive attitudes. These findings require further examination, especially if experience of alcohol use and smoking may enhance knowledge and relatability to drinkers.

A44
Comorbid depression in alcohol users: refining target groups for brief alcohol interventions in medical care settings Diana Guertler 1,2 , Anne Moehring 1,2 , Kristian Krause 1 , Jennis Freyer-Adam 2,3 , Sabina Ulbricht 1,2 , Gallus Bischof 4 , Hans-Jürgen Rumpf 4 , Anil Batra 5 , Sandra Eck 5 , Sophie Baumann 1,2,6 , Ulrich John 1,2 ; Christian Meyer 1,2 Addict Sci Clin Pract 2019, 14(Suppl 1):27 Background: Primary care providers (PCPs) face multiple barriers to offering substance use interventions, including lack of time, knowledge, and information about their patients' drug use. We developed a tablet-based Substance Use Screening and Intervention Tool (SUSIT) to assist PCPs by delivering screening results and clinical decision support for conducting brief intervention (BI) to address unhealthy drug use. The SUSIT screener is a self-administered brief screen (SUBS) and modified WHO-ASSIST. This pilot study examined whether the SUSIT increases delivery of BI during primary care visits. Methods: Adult patients completed tablet-based screening in the waiting room, and identified their drug of most concern (DOMC). Those with moderate-risk use of any drug (without high-risk alcohol or drug use) were eligible. A pre-post design compared participants enrolled during the control period to a new group of participants enrolled during the intervention period, in which PCPs received the SUSIT. All participants completed an after-visit survey documenting the elements of BI delivered by the PCP, and a 90-day timeline follow-back. Results: The 78 participants (42 control, 36 intervention) were majority male (76%), with a mean age of 46 (SD = 13). Marijuana was the most prevalent DOMC (n = 52 (66.7%)); cocaine was the second most prevalent DOMC (n = 7 (9.0%)). Mean days of use of the DOMC in the past 90 days was 38.8 (SD = 37.7). During the intervention period, PCPs used the SUSIT with 31 of 36 (86%) participants. Participants in the intervention condition were more likely to report receiving BI [(n = 33 (91.7%) vs. n = 17 (40.5%), P < 0.001]. The intervention group also received more elements of BI [median = 9.5, mean 7.8 (SD = 4.5) vs. median = 0, mean 2.7 (SD = 4.3); P < 0.001]. Conclusions: Providing drug use screening information and clinical decision support to PCPs increased the delivery of BI during routine primary care visits. Future analyses will examine changes in drug use behavior 3 months post-intervention. Trial registration: ClinicalTrials.gov registration number: 16-01074. Background: Association of the use of psychoactive substances and tuberculosis make diagnosis and treatment a barrier to reach universal coverage of the disease worldwide. Objectives: To identify the profile and the pattern of consumption of psychoactive substances of patients undergoing tuberculosis treatment in the network of basic health services; to analyse the adherence to the treatment of patients of tuberculosis who consume these substances and perform brief intervention in this clientele from the perspective of the adherence to the tuberculosis treatment. Methods: Sectional study, carried out in primary care units in the modality of the Family Health Strategy, in Rio de Janeiro, with a sample of 114 patients in the treatment of tuberculosis using the ASSIST. The exposure variable was the consumption of psychoactive substances and the outcome variable adherence to treatment. In the first phase the brief intervention was carried out, in the stages of feedback, due guidance and empathically. In the second phase after two months, a search was performed on the medical record for confirmation or non-compliance. Results: Prevalence in the male population 71.1%, median age 39 years, incomplete primary schooling 52.6%, brown skin color 42.1%, family income > 1 minimum wage, 74.5% lived with relatives. Prevalence for tobacco 28.0%, alcoholic beverages 12.3%, marijuana 5.4% and cocaine/Crack 3.5%. Regarding adherence, after two months of Brief Intervention, with a survey in the patients' charts, a higher prevalence of adherence was observed in the male population, over 40 years old, with medium/high school education, married and living in union, whites received up to 1 minimum wage, live with relatives, adhered to the treatment of tuberculosis. Conclusions: These results demonstrate the importance of brief interventions applied by health Professionals with these patients, decreasing the incidence of infected and bacilliferous patients, prone to the spread of the disease.

A48
Prevalence of alcohol misuse problem (AMP) recognition within those meeting criteria for alcohol misuse: A meta-analysis Panagiotis Spanakis 1 , Jessica Smith 1 , Rachael Gribble 2 , Sharon Stevelink 2 , Roberto Rona 2 , Nicola Fear 2 , Laura Goodwin 1 Background: Recognition of AMP is a fundamental stage in the psychological process of taking action to change behaviour and, if necessary, seek treatment. Several studies have identified that AMP recognition is a significant correlate of help-seeking, while lack of recognition is a barrier to help. A systematic review and meta-analysis were conducted to estimate the prevalence of AMP recognition within individuals meeting validated AMP criteria. Methods: We searched PsycInfo, Web of Science, Scopus and MedLine using the keywords: problem*; recogni* OR perceive* OR perception OR self-identif*; alcohol. We identified studies that reported weighted or unweighted frequencies of individuals who meet criteria for AMP (e.g. AUDIT scores or DSM criteria) and who self-identified their problems, self-recognised a need for receiving help, or have passed the pre-contemplation stage for taking action on their problems. Studies were eligible if they were published between 2000 and 2019 in English and included an adult sample. Studies were excluded if the study used a sample of adolescents, university students, or illicit drug users. A random-effects model meta-analysis was used to estimate the pooled prevalence of AMP recognition with 95% CIs. Results: 27 papers were included in the meta-analysis (N = 77,081, on average 73% were male). Seventeen studies included participants with at least hazardous drinking and ten studies included participants with alcohol use disorder (AUD). Thirteen studies examined AMP recognition directly (self-identification), eight studies examined stage of change, and six studies examined need for help. Preliminary results showed that the pooled prevalence of AMP recognition was 42% (95 CI = 34%-51%). Conclusion: Less than half of those with AMP recognise their problems. Given the importance of AMP recognition in the process of change, future research should focus on policies and interventions that could help the affected individuals increase their self-awareness regarding their AMP. unemployment, and drinking motives including boredom, with a view to designing an Alcohol Brief Intervention for those out of work based on this information. Methods: The study recruited employed (n = 94) and unemployed (actively seeking employment) (n = 94) individuals through social media. The groups were compared on their drinking habits (Alcohol Use Disorder Identification Test), drinking motivations (Drinking Motivations Questionnaire), and recent feelings of boredom and low mood. MANOVAs were run to explore the between group differences, with adjusted regressions examining the role of potential confounders. Results: The average age was 33.05 (SD = 10.902) years, with more females (73.5%) recruited. Unemployed participants scored significantly higher on the AUDIT [F(1,164) = 8.59, p < 0.01], coping [F(1,164) = 9.80, p < 0.01], and boredom[F(1,164) = 14.83, p < 0.01] drinking motivations compared to the employed group. These motivations were significantly positively associated with higher AUDIT scores in the unemployed group when controlling for demographic, depression, and boredom scores. Conclusion: The study shows that problematic alcohol use is more common in the unemployed and demonstrates the links between alcohol use and unemployment, with boredom and coping as two key motives among the unemployed for increased drinking. Further qualitative work will seek to engage those who are out of work to further understand the link between boredom and alcohol use in the unemployed. This will then be used to develop an Alcohol Brief Intervention which will be targeted at the unemployed. Background: In order to improve interventions efficiency among Indigenous teenagers, adapting questionnaires to their cultural reality is paramount. In fact, without this cultural adaptation, the results obtained with these questionnaires could minimize or exaggerate the extent of youth's difficulties. The present study adapted the Detection of Alcohol and Drug Problems among Teenagers (Landry et al. 2004). Method: Crees (anglophones) and Attikamekws (francophones) have collaborated to the validation of this screening test. A first step constituted in a co-building process allowing for the development of an initial version of the DEP-ADO adapted to Indigenous' cultural reality. During a second step, 20 youth have filled the adapted version and have taken part in focus groups aimed to give them the opportunity to comment their understanding and usefulness of such a grid for Indigenous Youth. Finally, during a third step, a second adapted version of the test was administered to Youth from three communities (N = 421, average age 14.75) for a final validation. Results and conclusion: The main challenge was to make sure that the wording and the examples were easily understandable and make sense for Indigenous Youth. For examples: changing the questionnaire response options, give examples to help understand the issues at stake, review the phrasing of the questions. During the focus groups, youth reported that many questions or concepts were difficult to understand and needed to be reworded or clarified. Also, as some questions are more sensitive, participants reported experiencing shame and hesitating to answer accurately. Regarding the reliability, the DEP-ADO (adapted) scales indicated that alcohol scale (α tet = 0.91), drugs scale (α tet = 0.86), risk factors scale (α tet = 0.88) and global score (α tet = 0.93) had adequate reliability for Indigenous Youth. The one-factor latent structure was confirmed for all scales. Background: Less than 10% of the organizations treating addiction include parenthood support in their treatment plan. The Parent Management Training (PMT) programs have proven their efficiency among the parents who have an addiction. These programs based on probative data and relying on the best practices are, however, not easy to implement in some practice environments. Two approaches are in opposition; the one of a program type based on probative data and the one of a practice type based on probative data. How can we maintain the equilibrium between the «ingredients» associated to a program's efficiency and the implementation in all the different practice environments? Method: Developed in order to meet the specific needs of the families in which parents are addicted to alcohol or drugs, the PMT Cap sur la famille was implemented in addiction treatment centers in Quebec. However, in consideration of regional particularities, the program could not be implemented in its entirety uniformly. A consultation was carried out in order to identify the factors that could hamper the implementation of this type of program in the practice environments. In total, 17 addiction treatment centers (22 standardized interviews) have been consulted. Results and conclusion: During consultation and despite the acknowledgement of the need for a PMT program specific to the clientele in addiction, managers and addiction workers agreed that the implementation feasibility of the PMT program presents important challenges on human resources and facilities aspects. Up to now, 8 addiction treatment centers are implementing the program, each with its adapted formula. Aspects such as dosage, intensity, content and clientele reached needed to be adapted while respecting the best practices recommendations. The steps taken, the various versions of the program and the implementation quality will be discussed according to the practice based on probative data. Background: Adolescent substance use poses serious physical and mental health risks that can extend into adulthood. Parental disapproval of substance use is commonly considered an important protective factor for prevention of risky adolescent behaviors. Less is known about the role of parental disapproval among youth who have initiated risky substance use and are candidates for brief intervention. Methods: Adolescents ages 14-18 with risky alcohol and/or marijuana use were recruited from two US urban school-based health centers into a randomized trial comparing a computer versus nurse practitioner-delivered brief intervention. This secondary analysis examines the relationship between perceived parental disapproval of substance use and adolescents' frequency of marijuana and alcohol use at baseline, 3-, and 6-month follow-up. Measures of parental disapproval were aligned with questions from national epidemiological survey data. Generalized estimating equations were used to examine trajectories of alcohol and marijuana use frequency by level of perceived parental disapproval, adjusting for identified covariates.