Journal of Health Communication, 14:612–630, 2009 Copyright # Taylor & Francis Group, LLC ISSN: 1081-0730 print=1087-0415 online DOI: 10.1080/10810730903089622

Online Illegal Drug Use Information: An Exploratory Analysis of Drug-Related Website Viewing by Adolescents

Downloaded By: [Donaghy, Maryann][George Washington University] At: 16:45 4 December 2009

STEVEN BELENKO Temple University, Philadelphia, Pennsylvania, USA

KAREN L. DUGOSH, KEVIN LYNCH, AMY A. MERICLE, AND MICHELE PICH Treatment Research Institute, University of Pennsylvania, Philadelphia, Pennsylvania, USA

ROBERT F. FORMAN Alkermes, Inc., Cambridge, Massachusetts, USA; University of Pennsylvania, Philadelphia, Pennsylvania, USA Given the uncertain effects of antidrug media campaigns, and the ease of finding online illegal drug information, research is needed on the Internet role in disseminating drug information to youths. This exploratory study analyzes National Survey of Parents and Youth (NSPY) data on drug website viewing among 12–18 year olds (N ¼ 7,145). Approximately 10.4% reported drug-related website exposure: 5.4% viewed only websites that communicated how to avoid drugs or bad things about drugs (antidrug websites); 1.7% only viewed websites that communicated how to use drugs and good things about drugs (prodrug websites); and 3.2% viewed both types of websites. The low rates of viewing antidrug websites occurred despite efforts in the National Youth Antidrug Media Campaign (NYAMC) to encourage youths to visit such websites. Prodrug website viewers had used inhalants and been offered marijuana, perceived little risk in trying marijuana, intended to use marijuana, had close friends who used drugs, reported low parental monitoring, and had been exposed to antidrug media messages. Viewing antidrug websites was related to gender, income, likelihood of using marijuana in the next 12 months, having close friends who use drugs and talking to friends about avoiding drugs, parental monitoring, and drug prevention exposure. Prior prevention exposure increased drug website viewing overall, perhaps by increasing general curiosity about drugs. Because adolescents increasingly seek health information online, research is needed on how they use the Internet as a drug information source, the temporal relationships of prevention exposure and drug website viewing, and the effects of viewing prodrug websites on drug risk. This research was supported in part by the National Institute on Drug Abuse and Westat, under subcontract No. s7120 to the Treatment Research Institute at the University of Pennsylvania (Principal Investigator Steven Belenko, PhD). We also are grateful to the National Survey of Parents and Youth (NSPY) Users Group and AMSAQ for technical support and guidance on use of the Restricted Use Files of the NSPY. Address correspondence to Steven Belenko, Professor, Department of Criminal Justice, Temple University, 5th Floor Gladfelter Hall, 1115 West Berks Street, Philadelphia, PA 19122, USA. E-mail: [email protected]

612

Downloaded By: [Donaghy, Maryann][George Washington University] At: 16:45 4 December 2009

Drug Website Viewing Among Adolescents

613

Adolescent drug use remains a substantial social and public health problem (Johnston, O’Malley, Bachman, & Schulenberg 2006; Substance Abuse and Mental Health Services Administration [SAMHSA], 2006) and is associated with delinquency, poor school performance, and social problems (Belenko & Logan, 2003; Hawkins, Catalano, & Miller, 1992; Loeber & Farrington, 1998). Although some effective, targeted school-based prevention models exist (Ellickson, McCaffrey, Ghosh-Dastidas, & Longshore, 2003; Griffin, Botvin, Nichols, & Doyle, 2003; Sun, Skara, Sun, Dent, & Sussman, 2006), recent evidence indicates that broadly targeted prevention efforts such as antidrug media campaigns may have limited effects on drug use or attitudes, beliefs, and intentions about drug use (Czyzewska & Ginsburg, 2007; Fishbein, Hall-Jamieson, Zimmer, von Haeften, & Nabi, 2002; Westat & Annenberg, School for Communication, 2003), and may increase risk of marijuana initiation (Czyzewska & Ginsburg, 2007; Hornik, 2006). To reduce adolescent drug use, we must improve our understanding of the individual, peer, family, and community factors affecting drug initiation, perceived risks of using drugs, escalation from experimental to regular use, and desistance. Such factors may be affected by how adolescents receive and process such information from school-based drug prevention programs, peers, parents, print media, and, increasingly, the Internet. Nearly all youths 8–18 years old have gone online (96%), 74% have a home Internet connection (Rideout, Roberts, & Foehr, 2005), and 61% go online on a typical day (Kaiser Family Foundation, 2004). This article presents exploratory analyses of the extent of drug-related website viewing and the characteristics of adolescents viewing prodrug and antidrug websites. Its universal accessibility makes the Internet a potentially powerful platform for disseminating prevention information to adolescents, but it also can undermine prevention messages. A study of teens aged 15–17 found that 76% had gone online to research health topics, including drug=alcohol abuse (25%) and smoking (23%; Kaiser Family Foundation, 2001). Among teens researching health topics online, 53% said they talked with a parent or other adult about what they found, and 41% reported changing their behavior because of online information. Recent studies suggest an emergence of websites advertising the sale of illegal drugs and promoting illegal drug use (Forman, 2003; Forman, Marlowe, & McLellan, 2006; Volkow, 2004). Such ‘‘prodrug’’ websites extol the positive aspects of psychoactive drug use, promote increased use, provide information about growing marijuana, and downplay the harms associated with drug use (Forman, 2006). Popular search engines such as Google provide adolescents with ready access to prodrug information (Forman, Woody, McLellan, & Lynch, 2006); inputting drug search terms yields many websites promoting drug use or offering to sell scheduled drugs without a prescription. Monthly Google searches in 2003–2004 using the term ‘‘codeine’’ found that 17% to 27% of the first 100 links offered codeine without a prescription; and 59% of the first 100 links for ‘‘Vicodin’’ were nonprescription websites (Forman, Woody et al., 2006). One-third of websites identified through the search terms ‘‘marijuana’’ or ‘‘Ecstasy’’ are prodrug, as are 39% of ‘‘Vicodin’’ and 38% of ‘‘Xanax’’ websites (Pich et al., 2007). Thus, young people searching the Internet for drug information are likely to encounter prodrug as well as antidrug messages, or be linked to websites offering nonprescription sale of controlled substances. Many parents recognize the potentially negative impacts of the Internet: 61% of parents have rules about their children’s Internet use, a similar percentage check which websites their children have visited, and 41% have installed website filters

Downloaded By: [Donaghy, Maryann][George Washington University] At: 16:45 4 December 2009

614

S. Belenko et al.

(Lenhart, Rainie, & Lewis, 2001). Much of this parental concern centers on access to pornography, violence, and contact with sexual predators, but easily accessible prodrug websites also increase the importance of studying drug website viewing patterns. Unfortunately, there is little research on this phenomenon, the factors that distinguish those viewing different drug websites, or how such viewing affects attitudes toward and use of marijuana and other drugs. The NSPY provides data for informing some of these issues. For example, antidrug website viewing was more common among females (10.1% vs. 7.8% of males), African Americans (11.0% vs. 7.8% of Hispanics and 8.6% of Whites), and youths who were high sensation seekers (9.8% vs. 7.8% of low sensation seekers; Westat & Annenberg, 2003). Prodrug website viewing was more common among males (5.5% vs. 4.1% of females), higher-risk youths (8.8% compared with 2.5% of low risk), and high sensation seekers (7.1% vs. 2.1% of low sensation seekers). Although an important initial step for understanding adolescent drug website viewing, these findings did not distinguish among youths who view only antidrug websites, view only prodrug websites, and both types of websites. These youth may differ in ways that could influence the effectiveness of web-based prevention strategies. For example, youths with more general curiosity about drug use or mutable intentions may explore both types of websites, while those with firmer pro or antidrug use beliefs may seek out websites supporting these beliefs. Problem behavior theory (Jessor, 1991; Jessor & Jessor, 1977) and research on initiation into substance abuse and delinquency suggests that risk factors for such behaviors often occur across multiple domains and cluster in the same individuals (Elliott, Huizinga, & Menard, 1989; Huizinga, Loeber, & Thornberry, 1993). Risk factors include low school achievement, childhood abuse, ineffective parenting, low socioeconomic status, negative peer influence, and high-risk schools and communities (Elliott et al., 1996; Farrington, 2005; Maguin & Loeber, 1996; Thornberry, Krohn, Lizotte, Smith, & Tobin, 2003). Viewing prodrug websites might be related to other risk factors for substance abuse, so it is important to explore whether such factors also might be predictors of different types of website viewing. The importance of clear parental behavioral boundaries and rules, positive communication patterns, parental attitudes toward drug use, and parental attachment also have been found to influence drug risk (Belenko & Logan, 2003; Hawkins, 1997, Hussong & Chassin, 1997; Jackson, Henriksen, & Dickinson, 1999; Kosterman, Hawkins, Guo, Catalano, & Abbott, 2000; Maguin & Loeber, 1996; Resnick et al., 1997; Sullivan & Farrell, 1999). A large body of research finds that peer behaviors and norms are linked to substance use among adolescents, including peer drug and alcohol use, peer group interactions, and peer approval of drug use (Brook, Cohen, Whiteman, & Gordon, 1992; Curran, White, & Hansell, 2000; Kandel, 1996; Keenan, Loeber, Zhang, Stouthamer-Loeber, & Van Kammen, 1995; Newcomb, 1997); these peer influences may vary by drug type and reflect both the influence of peers and that youths may seek out peers with attitudes and behaviors similar to their own (Rice, Donohew, & Clayton, 2003). Accordingly, our analyses also explore the influence of several peer and parental monitoring=oversight behaviors on drug-related websites viewing. Although we could find no published research on the factors associated with drug-related website viewing, the health communication literature suggests potential factors. For example, Sun and colleagues (2005) found that higher

Downloaded By: [Donaghy, Maryann][George Washington University] At: 16:45 4 December 2009

Drug Website Viewing Among Adolescents

615

socioeconomic status and Asian ethnicity were associated with higher Internet use among seventh graders in Southern California. These researchers also found that substance abuse, less parental monitoring, and more unsupervised time were positively related to email use, chat-room use, and home Internet use. Additionally, international research suggests that problematic Internet use is associated with decreased grade point average (Bayraktar & Gun, 2007) and alcohol abuse (Ko et al., 2008). With the increasing salience of the online world for adolescents seeking health information and social connections, it is important to explore their drug website viewing patterns and the factors associated with these patterns. Specifically, we explore whether youths with different website viewing patterns differ by level of risk, peer norms, parental behaviors, prior drug use, and exposure to other types of media messages. These findings can help guide the development of additional research and hypotheses to be examined in future studies.

Methods The NSPY was developed to evaluate the 1999–2003 National Youth Antidrug Media Campaign (NYAMC; see Westat & Annenberg, 2003, for details on the NSPY sampling design and content). Briefly, a national probability household sample of youths aged 9–18 and their parents or guardians was interviewed from November 1999 to June 2001 (Round 1). Follow-up interviews were conducted from July 2001 through June 2002 (Round 2), July 2002 through June 2003 (Round 3), and July 2003 through June 2004 (Round 4). The sample was stratified to obtain sufficient numbers in each targeted age range: 9–11, 12–13, and 14–18. Analytical Sample Data for the present study were derived from a subsample of the youth and parent (dyad) files of the NSPY Restricted Use files for Rounds 1, 2, and 3; youth and parent data were linked by household ID. Most variables were contained in the youth file; only two variables (parental income and parental viewing of drug-prevention website) were used from the parent file. Children aged 9–11 and respondents in the first two waves of Round 1 were not asked Internet use questions. Because of relatively low base response rates for viewing drug-related websites, and because there were multiple data waves, we maximized the number of cases for analysis by examining website viewing at the most recent follow-up interview. For youths with all three rounds of data, we used Round 3 data to reflect the most recent patterns. For youths without a Round 3 interview, Round 2 data were selected, and for those without a Round 2 or 3 interview, Round 1 data were selected. Youths who reported no Internet use the past 6 months in their most recent interview were excluded. A total of 5,343 youths provided Round 3 data, 807 Round 2 data, and 1,121 Round 1 data. Excluding 126 cases missing drug website viewing data yielded a final sample of 7,145. To check for potential bias (if youths in later rounds were more likely to view drug-related websites), we compared website viewing across rounds. No significant differences in the proportions viewing drug websites were found over time, whether examining all youths interviewed in each round or those interviewed in all three rounds (data not shown).

616

S. Belenko et al.

Downloaded By: [Donaghy, Maryann][George Washington University] At: 16:45 4 December 2009

Measures Youth Website Viewing Patterns The youth survey included questions pertaining to recent Internet use. One general item asked how often they had used the Internet in the past 6 months. Consistent with prior analyses (Westat, 2005), we categorized youths as viewing ‘‘prodrug’’ websites if they reported visiting a website that told them ‘‘how to use drugs’’ or ‘‘good things about drugs,’’ and as viewing ‘‘antidrug’’ websites if they reported visiting a website that told them ‘‘how to avoid drugs’’ or ‘‘bad things about drugs.’’ Because some viewed both types of sites, and might have different characteristics than those viewing only pro or antidrug websites, we created a cross-classification of antidrug or prodrug site viewing that yielded a four-level nominal dependent variable: viewed neither type of website, viewed prodrug websites only, viewed antidrug websites only, viewed both types of websites. The NSPY Internet questions did not specifically ask about viewing antidrug websites specifically tied to the NYAMC. Predictors of Website Viewing Patterns There was little extant research to guide development of specific hypotheses about potential mediating or moderating factors on website viewing patterns. Considering that prodrug website viewing behaviors may be similar to other types of risk behaviors, however, we explored a number of potential covariates available in the NSPY including demographics, substance use history (alcohol, marijuana, inhalants), intentions to use marijuana and perceived risk of marijuana use, peer behavior and norms, parental monitoring and Internet use, and exposure to prevention messages. Demographics. Variables included gender, race=ethnicity, age (12–13, 14–15, 16–18), average school grades (A=A
, Bþ=B=B
, Cþ=C=C
, Dþ=D=D
), and parent-reported income (