Raminta Daniulaityte

PREDOSE: A semantic web platform for drug abuse epidemiology using social media

OBJECTIVES The role of social media in biomedical knowledge mining, including clinical, medical and healthcare informatics, prescription drug abuse epidemiology and drug pharmacology, has become increasingly significant in recent years. Social media offers opportunities for people to share opinions and experiences freely in online communities, which may contribute information beyond the knowledge of domain professionals. This paper describes the development of a novel semantic web platform called PREDOSE (PREscription Drug abuse Online Surveillance and Epidemiology), which is designed to facilitate the epidemiologic study of prescription (and related) drug abuse practices using social media. PREDOSE uses web forum posts and domain knowledge, modeled in a manually created Drug Abuse Ontology (DAO--pronounced dow), to facilitate the extraction of semantic information from User Generated Content (UGC), through combination of lexical, pattern-based and semantics-based techniques. In a previous study, PREDOSE was used to obtain the datasets from which new knowledge in drug abuse research was derived. Here, we report on various platform enhancements, including an updated DAO, new components for relationship and triple extraction, and tools for content analysis, trend detection and emerging patterns exploration, which enhance the capabilities of the PREDOSE platform. Given these enhancements, PREDOSE is now more equipped to impact drug abuse research by alleviating traditional labor-intensive content analysis tasks. METHODS Using custom web crawlers that scrape UGC from publicly available web forums, PREDOSE first automates the collection of web-based social media content for subsequent semantic annotation. The annotation scheme is modeled in the DAO, and includes domain specific knowledge such as prescription (and related) drugs, methods of preparation, side effects, and routes of administration. The DAO is also used to help recognize three types of data, namely: (1) entities, (2) relationships and (3) triples. PREDOSE then uses a combination of lexical and semantic-based techniques to extract entities and relationships from the scraped content, and a top-down approach for triple extraction that uses patterns expressed in the DAO. In addition, PREDOSE uses publicly available lexicons to identify initial sentiment expressions in text, and then a probabilistic optimization algorithm (from related research) to extract the final sentiment expressions. Together, these techniques enable the capture of fine-grained semantic information, which facilitate search, trend analysis and overall content analysis using social media on prescription drug abuse. Moreover, extracted data are also made available to domain experts for the creation of training and test sets for use in evaluation and refinements in information extraction techniques. RESULTS A recent evaluation of the information extraction techniques applied in the PREDOSE platform indicates 85% precision and 72% recall in entity identification, on a manually created gold standard dataset. In another study, PREDOSE achieved 36% precision in relationship identification and 33% precision in triple extraction, through manual evaluation by domain experts. Given the complexity of the relationship and triple extraction tasks and the abstruse nature of social media texts, we interpret these as favorable initial results. Extracted semantic information is currently in use in an online discovery support system, by prescription drug abuse researchers at the Center for Interventions, Treatment and Addictions Research (CITAR) at Wright State University. CONCLUSION A comprehensive platform for entity, relationship, triple and sentiment extraction from such abstruse texts has never been developed for drug abuse research. PREDOSE has already demonstrated the importance of mining social media by providing data from which new findings in drug abuse research were uncovered. Given the recent platform enhancements, including the refined DAO, components for relationship and triple extraction, and tools for content, trend and emerging pattern analysis, it is expected that PREDOSE will play a significant role in advancing drug abuse epidemiology in future.

Predictors of transition to heroin use among initially non-opioid dependent illicit pharmaceutical opioid users: A natural history study

BACKGROUND Increases in illicit pharmaceutical opioid (PO) use have been associated with risk for transition to heroin use. We identify predictors of transition to heroin use among young, illicit PO users with no history of opioid dependence or heroin use at baseline. METHODS Respondent-driven sampling recruited 383 participants; 362 returned for at least one biannual structured interview over 36 months. Cox regression was used to test for associations between lagged predictors and hazard of transition to heroin use. Potential predictors were based on those suggested in the literature. We also computed population attributable risk (PAR) and the rate of heroin transition. RESULTS Over 36 months, 27 (7.5%) participants initiated heroin use; all were white, and the rate of heroin initiation was 2.8% per year (95% CI=1.9%-4.1%). Mean length of PO at first reported heroin use was 6.2 years (SD=1.9). Lifetime PO dependence (AHR=2.39, 95% CI=1.07-5.48; PAR=32%, 95% CI=-2% to 64%), early age of PO initiation (AHR=3.08, 95%; CI=1.26-7.47; PAR=30%, 95% CI=2%-59%), using illicit POs to get high but not to self-medicate a health problem (AHR=4.83, 95% CI=2.11-11.0; PAR=38%, 95% CI=12%-65%), and ever using PO non-orally most often (AHR=6.57, 95% CI=2.81-17.2; PAR=63%, 95% CI=31%-86%) were significant predictors. CONCLUSION This is one of the first prospective studies to test observations from previous cross-sectional and retrospective research on the relationship between illicit PO use and heroin initiation among young, initially non-opioid dependent PO users. The results provide insights into targets for the design of urgently needed prevention interventions.

Initiation to Pharmaceutical Opioids and Patterns of Misuse: Preliminary Qualitative Findings Obtained by the Ohio Substance Abuse Monitoring Network

Pharmaceutical opioid misuse has been recognized as a growing public health problem across the nation. To develop appropriate treatment and prevention programs, the population of pharmaceutical opioid abusers has to be well understood. This exploratory study is based on qualitative interviews with 24 people in the Dayton/Columbus, Ohio, area. Interviews were conducted for the Ohio Substance Abuse Monitoring Network, a statewide epidemiological surveillance system. Participants ranged in age from 18 to 48 years; the majority was White and male. To explain initiation and continued use of pharmaceutical opioids, participants discussed a number of different reasons, including self-medication of emotional and physical pain, legitimate medical prescriptions related to chronic pain management, social influences, recreation, and easy access to pharmaceutical opioids. On the basis of participant age and lifetime experiences with pharmaceutical opioid and other drug misuse, six user groups were identified that faced unique risks and prevention/treatment challenges. Research implications are discussed.

Illicit use of pharmaceutical opioids among young polydrug users in Ohio

This study, conducted in Columbus, Ohio, identified the predictors of current (past 30 day) illicit use of pharmaceutical opioids among young individuals (n=402) with a history of MDMA/ecstasy use who were recruited in 2002-2003 using respondent-driven sampling. To be eligible, participants had to be 18-30 years old, not involved in a drug abuse treatment, and report MDMA/ecstasy use at least once in the past 6 months. About 81% reported lifetime, and more than 31% reported current illicit use of pharmaceutical opioids. Logistic regression analysis revealed that illicit use of pharmaceutical tranquilizers and pharmaceutical stimulants were the strongest predictors of illicit use of pharmaceutical opioids. Risk of pharmaceutical opioid use was also greater among those who had higher levels of depressive symptoms and reported current use of hallucinogens and inhalants. Our findings suggest that the non-medical use of pharmaceutical opioids is a part of polydrug use practices that often incorporate illicit use of other prescription drugs. The observed relationship between depressive symptoms and illicit use of pharmaceutical opioids may have important implications for prevention programming and should be examined in future research.

“Heavy Users,” “Controlled Users,” and “Quitters”: Understanding Patterns of Crack Use Among Women in a Midwestern City

Over the past two decades, the use of crack cocaine has become an enduring part of the social ecology of many cities across the United States. The purpose of this exploratory study is to describe patterns of crack use drawing on life history interviews conducted with 18 women in Dayton, Ohio, between 1998 and 2000. Ten of the women were African American, and eight were white. Age ranged between 23 and 47. The women were at very different stages of their crack-cocaine careers. We focus on understanding the social factors, life history, and everyday circumstances that participants related to their current patterns and levels of crack use. Implications for intervention are discussed.

Respondent-driven sampling to recruit young adult non-medical users of pharmaceutical opioids: Problems and solutions

Respondent-driven sampling (RDS) has been promoted as a superior method in recruiting hard-to-reach and hidden populations. Although its application has expanded enormously, there remains a need for empirical data evaluating the performance of RDS in different settings. This study describes the application of RDS to recruit a community sample (N=396) of young adults (18-23 years old) into a natural history study of non-medical pharmaceutical opioid use. Since recruitment targeted non-dependent pharmaceutical opioid users, and applied other eligibility restrictions, several modifications had to be made to make RDS work with this narrowly defined target population. RDS recruitment was less efficient than expected, and produced greater numbers of African American recruits than anticipated. Although the sampling quota was met, sample analysis revealed a lack of equilibrium in terms of ethnic composition and very strong in-group recruitment tendencies among White and African American respondents. This study contributes potentially helpful insights into the strengths and limitations of using RDS which may benefit future studies.

Methamphetamine Use in Dayton, Ohio: Preliminary Findings from the Ohio Substance Abuse Monitoring Network

Abstract Since the 1990s, methamphetamine manufacturing and use have been spreading into midwestern and southern United States. However, most research on methamphetamine use has been conducted in the western United States. This exploratory study is based on qualitative interviews with 26 people recruited in Dayton, Ohio, for the Ohio Substance Abuse Monitoring Network, a statewide epidemiological surveillance system. Participants ranged in age from 21 to 57 years. The majority were White and male. The participants reported increases in methamphetamine availability and identified powder-type locally manufactured methamphetamine as the most commonly seen form of the drug. The participants reported extensive drug use histories, and were introduced to methamphetamine in the context of their involvement with pharmaceutical stimulants, crack, powder cocaine, and/or club drug scene. A large proportion of the participants continued to use crack or powder cocaine in addition to methamphetamine. Many primary methamphetamine users felt they were able to moderate methamphetamine use and fulfill their daily responsibilities related to work and family life. Ranking results and consensus analysis revealed that participants shared high agreement about drug-related risks, and perceived methamphetamine as a serious drug, although not as risky as heroin, crack or OxyContin®. Implications for treatment and prevention are discussed.

Sources of Pharmaceutical Opioids for Non-Medical Use among Young Adults

Abstract The study uses qualitative and quantitative data to describe sources of pain pills for illicit use among young adult (18- to 23-year-old) users. Respondent-driven sampling was used to recruit 383 individuals in the Columbus, Ohio, area. The sample was almost 50% Caucasian and about 55% male. Qualitative interview participants (n = 45) were selected from the larger sample. Qualitative data suggest that pharmaceutical opioid availability was so pervasive that most individuals did not have to venture outside of their immediate social networks to find people who sold or shared pills. Participants emphasized differences between those who are actively involved in obtaining pills and those who play a more passive role. Active involvement was described as going out searching for pills and paying money to obtain them. In contrast, passive role included obtaining pills when somebody offered or shared them free of charge. Multiple logistic regression analysis indicates that a more active role in obtaining pharmaceutical opioids was related to being White, more frequent use of pharmaceutical opioids, extended-release oxycodone use, and using pharmaceutical opioids to get high, as opposed to self-treating a health problem. The study results can help inform drug use epidemiology, interventions, and policy.

“To Numb Out and Start to Feel Nothing”: Experiences of Stress Among Crack-Cocaine Using Women in a Midwestern City

The study uses qualitative interviews conducted with 19 crack using women to explore their experiences of stress and their views regarding the relationship between stress and drug use. Fifteen of the women participated in follow-up interviews conducted 5–7 years after the baseline. Life history interviews unveiled a pattern of close connection between the intensity of womens drug use and the level of stress they experienced in relation to their past adversities and current life circumstances. The majority of the women viewed stress as an important causal explanation of their drug use. Tensions related to romantic relationships, traumatic childhood, motherhood failures, unabated grief, and humiliating experiences of “crack life” were discussed as the most common sources of psychosocial stress. Most women had very limited positive coping resources and skills. Crack use was perceived as a very common, although highly maladaptive, way to deal with stress. Implications for interventions are discussed.

Predictors of Depressive Symptomatology Among Rural Stimulant Users

Abstract This study examined sociodemographic and drug-related predictors of depressive symptoms among a rural, multistate sample of not-in-treatment stimulant drug users (n = 710). Participants were recruited using respondent-driven sampling in Ohio, Arkansas, and Kentucky. The Patient Health Questionnaire (PHQ-9) was used to measure symptoms of depression. Moderate to severe depressive symptomatology was reported by 43.0% of the sample. Cumulative logistic regression analysis showed that daily and nondaily crack use as well as the daily use of cocaine HCI increased the odds of depressive symptoms. Methamphetamine use had no significant association with depression. The daily use of marijuana, the illicit use of tranquilizers, light/moderate cigarette smoking, and injection drug use also increased the risk of depressive symptoms. Living in Kentucky or Ohio (compared to Arkansas), having unstable living arrangements, and being White, female, and older were related to higher odds of depressive symptoms. These results suggest that a host of drug and nondrug factors need to be considered when addressing depressive symptoms in stimulant users.