The prevalence of mental health disorders is greatest in those aged under 25 years, with 26% having a disorder, including 13% with a substance use disorder.1 In addition, a substantial proportion of young adults use substances in a risky manner without reaching the formal diagnostic criteria for a disorder — more than 40% of 18–19-year-olds use alcohol in a risky or high-risk fashion every month.2 Despite the prevalence of these problems, the use of treatment services is limited, with only 35% of those with a diagnosis in the previous 12 months having accessed mental health services for that disorder.1
Between 1998 and 2006–07, the proportion of Australians with access to the internet at home rose from 16% to 64%,2 and a further 25% reported having used the internet at other locations.2 The 15–24-years age group has the highest proportion of internet users. Across all ages, most users (58%–80%) report that they search the web for health information.3,4 Many adolescents think that the internet is a useful source of information on topics that are hard to discuss,5 and online health information is regarded as trustworthy and relevant by both sexes and across socioeconomic groups.6 Therefore, the internet provides an alternative vehicle for delivering health interventions. It may also provide a means of delivering services to those who are unable or unwilling to access conventional health services.
There is extensive evidence that brief interventions are an effective means of treating some substance use problems, especially for those who do not reach the criteria for dependence.7-10 Techniques from brief interventions, traditionally delivered face-to-face, can also be delivered electronically.11 However, the quality of electronic interventions remains in doubt. An assessment of 294 health behaviour change websites found that only 8.1% fulfilled the basic “5 A’s” (advise, assess, assist, anticipatory guidance, arrange follow-up) thought to be required to initiate change in behaviour.12,13 None of the alcohol-related websites met all five guidelines.13 In addition, many of the early publications on web-based interventions were descriptive or reported on feasibility or on studies that were methodologically weak.14,15 A recent review by Bewick and colleagues on web-based interventions for alcohol consumption16 found that only one intervention had used a randomised design that included a control of the highest standard according to their criteria.17
Further studies have been published since May 2006 (the cut-off date for Bewick et al’s review16), and there have been developments in web delivery across a range of common mental health problems. In addition, interventions and content designed for adults may not be appropriate or equally effective in younger groups. Given the importance of this age group in the initiation of substance use, their high level of use of internet resources, and willingness to access online health information, we aimed to review the current literature on interventions designed to target adolescents and young adults.
We searched MEDLINE, PsycINFO and Current Contents in February 2009 using the search strategy: “Substance abuse or alcohol or drug or tobacco and ([internet or web] and [intervention or RCT])”. The search was limited to English-language results and yielded 391 reports. Titles and abstracts were checked, and potentially eligible papers accessed for final assessment by one of us (R J T). References in eligible articles and reviews, as well as key journal sites (eg, Journal of Medical Internet Research), were used to identify further potentially relevant reports.
The inclusion criteria for the review were that studies had to use a randomised design to compare a web-based intervention with at least a no-treatment control. The eligible age groups were adolescents (typically targeted via interventions delivered through schools) or young adults (ie, specifically targeting tertiary students or other people aged 25 years or less). Outcomes had to include a measure of consumption of the target substance, not just change in attitude.
From a population perspective, for prevention programs to be useful in reducing health problems, they must not only have demonstrated efficacy and effectiveness but also the potential to be scalable, as well as having available resources to allow the program to be widely replicated while maintaining treatment fidelity.18 We therefore restricted our focus to include only interventions that used fully automated treatment programs and excluded those that required additional elements, such as in-person motivational interviews, due to the difficulty of delivering these at a population level. Interventions delivered via a stand-alone computer or CD-ROM were similarly excluded.
To avoid overestimating the magnitude of effects in studies with repeated measures, effect sizes (d) were calculated as between-group differences at follow-up divided by the pooled standard deviation from the baseline data.19 Where baseline data were unavailable, pooled standard deviations from the outcome results were used. In studies reporting medians and ranges, we estimated means and standard deviations using the method of Hozo and colleagues.20
Data were combined using Meta-Analysis software, version 5.3 (Ralf Schwarzer, Berlin, Germany), employing a random effects model. Where outcomes were significantly heterogeneous, the cluster analysis software incorporated in the same program was used to identify potential outliers.
The search strategy yielded 13 studies involving tertiary students and one with young employed adults (Box 1). All 14 studies targeted alcohol consumption, with one study also addressing other types of health behaviour (physical inactivity, low fruit and vegetable intake17). One study provided outcome data as change scores subdivided by sex;32 we did not include these in the meta-analysis phase.
The overall effect size for the outcomes (summarised in Box 2) was d = − 0.22 (SE, 0.06; 95% CI, − 0.34 to − 0.10), but with significant heterogeneity (Q = 249.03, df = 55; P < 0.00001). In light of the range of different types of outcome measures used and the different levels of baseline alcohol exposure, potential sources of heterogeneity were sought by separately analysing the results for three key outcome measures (average quantity of alcohol, frequency of heavy or binge-drinking events, and alcohol-related social problems) and for those studies that separately reported outcomes for people who were non-drinkers at baseline.
Ten studies reported on the quantity of alcohol consumed (Box 2; variables 8, 12, 18, 23, 28, 36, 42, 48, 50, 53). Overall, those who received the interventions had a lower level of alcohol consumption at follow-up than those in the control groups, with a mean difference of d = − 0.12 (SE, 0.05; 95% CI, − 0.22 to − 0.02), with the effect being homogeneous (Q = 7.36, df = 9; P = 0.600).
Seven studies reported on the frequency of heavy or binge drinking (Box 2; variables 1, 10, 20, 24, 29, 37, 44; variable 43 was an early outcome in the same study as 44 and was not included in our analysis). Young adults receiving the interventions had a lower frequency of heavy or binge drinking than controls (d = − 0.35; SE, 0.15; 95% CI, − 0.64 to − 0.06), although this measure still showed significant heterogeneity (Q = 29.74, df = 6; P = 0.00004). An inspection of a cluster plot showed the effects for the studies by Bersamin et al21 (d = − 0.99) and Kypri et al27 (d = − 0.80) to be outliers.
Six studies reported on alcohol-related social consequences, as assessed by measures such as the Rutgers Alcohol Problem Index36 or Alcohol Problems Scale37 (Box 2; variables 3, 21, 30, 38, 52, 55), with the overall effect size being d = − 0.57 (SE, 0.21; 95% CI, − 0.98 to − 0.15). These data showed significant heterogeneity (Q = 24.20, df = 5; P = 0.0002). A cluster analysis identified three clusters, with the smallest (variable 21) and largest (variable 38) effect sizes separate from the remaining variables.
Two studies reported outcomes separately for those who were non-drinkers at baseline.21,24 The overall effect for this subpopulation was d = − 0.001 (SE, 0.06; 95% CI, − 0.12 to 0.12), an effect not significantly different from zero (Z = − 0.016; P = 0.499). Our analysis did not include 751 people in Croom et al’s study who were non-drinkers at baseline and follow-up, as disaggregated data were not available for these people. The overall effect size for all participants in the study, regardless of baseline alcohol status, was d =− 0.02.24
Our systematic search yielded two eligible studies targeting adolescents (Box 3). Both studies targeted adolescent smoking: one randomly assigned schools to intervention or control groups,38 and the second used individual randomisation.39 Due to these methodological differences and differences in the target populations (all students v only smokers), these data were not combined for meta-analysis.
Our search strategy revealed two categories of web-based intervention for problematic substance use in young people. Web-based interventions targeting alcohol use by young adults, predominantly tertiary students, appear to be effective for alcohol problems in current drinkers, but there is insufficient evidence to support their use in preventing the development of alcohol-related problems among those who do not drink alcohol. The second category was web-based interventions addressing smoking cessation in adolescents and school children; however, there are currently insufficient data to assess the utility of such interventions in this group.
Brief in-person interventions for non-treatment-seeking individuals are an effective means of reducing alcohol consumption, with effect sizes in the range of 0.14 to 0.67 (with positive values showing better outcomes).8 Brief interventions have also been found to be effective among adolescents, with an overall effect size of 0.275.40 Thus, the effects reported here for web-based interventions are consistent with the magnitude of effects obtained from in-person interventions. There is preliminary evidence that computer-based interventions are cost-effective (eg, compared with cognitive behaviour therapy for depression).41 There is also evidence to support the scalability of web-based interventions,42 a potential benefit compared with in-person interventions. Nevertheless, it should be noted that some of the interventions reviewed here21,24 are more time-consuming than typical brief in-person interventions, which are designed to be delivered in less than four sessions.8 Cost–benefit analyses comparing web-based with in-person interventions are required.
An important caveat must be noted. In general, the web-based interventions in this review reported short-term outcomes, usually 3 months or less, and these may not represent a meaningful change in behaviour. Only four studies reported outcomes for 4 months or longer.27,29,33,34 The outcomes for these studies (Box 2) show effect sizes ranging from 0.12 to − 2.73, suggesting that persistent change in behaviour is possible with web-based interventions. Nevertheless, confirmation of long-term impacts is needed.
Considerable heterogeneity was noted in many of the measures and across the studies. Inspection of the interventions and measures did not reveal any clear explanation for this. For example, the measure of heavy drinking in Kypri et al’s 2008 study27 that was identified as an outlier in our sub-analysis was the same measure used in their 2004 study,29 with both studies using similar interventions and similar target populations (tertiary students recruited in a health care setting). Additionally, although the target populations across all but one of the alcohol interventions26 were tertiary students, the study samples included subgroups of this population ranging from non-drinkers at baseline21 to students who had been referred for counselling after breaches of university alcohol or drug policies.25 Furthermore, both Bersamin et al21 and Croom et al24 noted that the web interventions were ineffective with young people who were non-drinkers at baseline — findings confirmed by the combined data reported here. Although all the studies used randomised designs, there was considerable diversity in the intensity of the interventions delivered, ranging from an online course24 to a 15-minute assessment and feedback session.25 Finally, not all studies provided a control group with an intervention of similar intensity as the experimental intervention. Therefore, it is unsurprising that there were overall differences in the effectiveness of interventions or the effect sizes.
A previous review of brief interventions to reduce smoking in adolescents noted an effect size of 0.037.40 This included interventions delivered at the population level43 and for existing smokers.44 Our review also identified one population-level approach and one targeting existing smokers. The population-level intervention recruited participants in both Australia and the United States. The Australian arm of the study reported a lower prevalence of smoking in the intervention group compared with the control group and a lower level of initiation of smoking by non-smokers.38 However, these findings were not replicated in the US sample, and the authors concluded that web-based interventions are likely to have little practical impact on the level of smoking by adolescents.38
Another review45 estimated that the overall quit rate for stand-alone computer-based smoking-cessation interventions for adolescents was 13%.46-48 Patten et al39 reported that none of the adolescents in their study attained complete abstinence, but the 30-day abstinence level at 36 weeks was 13% for brief office-based intervention, compared with 6% for the web-based intervention. A potential reason for the poor performance of the web-based approach is that participants accessed the website from home (on a mean of 6.8 days over 24 weeks), with 86% visiting the site at least once, but less than a third visiting the site weekly after the third week. In contrast, delivery in a school setting may encourage greater compliance (eg, 77% completing all sessions,46 90% completing two of three sessions47). Therefore, motivation of adolescents and compliance may be key factors in the effectiveness of this form of intervention.
Our review has some potential limitations. The techniques and process of meta-analysis have received detailed and at times trenchant critiques,49,50 with key concerns being the quality of the studies included (“garbage in, garbage out”), combining different measures or interventions (“apples and oranges”), including multiple measures from studies (“inflated Ns”) and publication bias (“file drawer problem”).51,52 Nevertheless, the systematic assembly of data fulfilling clear criteria has come to the forefront in summarising scientific evidence.
Further, we did not identify any studies on the effectiveness of web-based interventions with adolescent drinkers. Given that 70% of 17-year-old students report having used alcohol in the past month and more than 40% report using it in a risky fashion,53 this would appear to be an important target for future research, even though it is difficult to obtain the necessary ethics approval to conduct research in this age group, especially when investigating interventions to convert behaviour.
Adolescence and young adulthood are the key period for initiation of substance use and the development of substance use disorders. Although the lifetime prevalence of licit drug use has remained stable, the age of initiation has fallen in more recent birth cohorts,54 with a concomitant increase in the risk of developing disorders in later life.55 Thus, there is an imperative to design and deliver interventions that address substance use by adolescents and young adults. Web-based interventions have the potential to provide interventions at a population level, with initial findings supporting their effectiveness in reducing problematic alcohol use in tertiary students and young adults.
1 Interventions for problematic substance use in young adults
2 Outcome measures and effect sizes of interventions for problematic substance use in young adults
3 Interventions for problematic substance use in adolescents
- Robert J Tait1
- Helen Christensen2
- Centre for Mental Health Research, Australian National University, Canberra, ACT.
This research was supported by the Centre for Mental Health Research at the Australian National University.
None identified.
- 1. Australian Bureau of Statistics. National Survey of Mental Health and Wellbeing: summary of results, 2007. Canberra: ABS, 2008. (ABS Cat. No. 4326.0.)
- 2. Australian Bureau of Statistics. Australian social trends, 2008. Canberra: ABS, 2008. (ABS Cat. No. 4102.0.)
- 3. Atkinson NL, Saperstein SL, Pleis J. Using the internet for health-related activities: findings from a national probability sample. J Med Internet Res 2009; 11 (1): e4.
- 4. Fox S. The engaged e-patient population. Washington, DC: Pew Internet & American Life Project, 2008. http://www.pewinternet.org/~/media//Files/Reports/2008/PIP_Health_Aug08.pdf.pdf (accessed Mar 2009).
- 5. Lenhart A, Rainie L, Lewis O. Teenage life online: the rise of the instant-message generation and the Internet’s impact on friendships and family relationships. Washington, DC: Pew Internet & American Life Project, 2001. http://www.pewinternet.org/~/media//Files/Reports/2001/PIP_Teens_Report.pdf.pdf (accessed Apr 2009).
- 6. Borzekowski DL, Rickert VI. Adolescent cybersurfing for health information: a new resource that crosses barriers. Arch Pediatr Adolesc Med 2001; 155: 813-817.
- 7. Bien TH, Miller WR, Tonigan JS. Brief interventions for alcohol problems: a review. Addiction 1993; 88: 315-335.
- 8. Moyer A, Finney JW, Swearingen CE, Vergun P. Brief interventions for alcohol problems: a meta-analytic review of controlled investigations in treatment-seeking and non-treatment-seeking populations. Addiction 2002; 97: 279-292.
- 9. Stead LF, Bergson G, Lancaster T. Physician advice for smoking cessation. Cochrane Database Syst Rev 2008; (2): CD000165.
- 10. Russell MA, Wilson C, Taylor C, Baker CD. Effect of general practitioners’ advice against smoking. Br Med J 1979; 2: 231-235.
- 11. Walters ST, Neighbors C. Feedback interventions for college alcohol misuse: what, why and for whom? Addict Behav 2005; 30: 1168-1182.
- 12. Fiore MC, Bailey WC, Cohen SJ, et al. Treating tobacco use and dependence: clinical practice guideline. Rockville, Md: US Department of Health and Human Services, 2000.
- 13. Evers KE, Cummins CO, Prochaska JO, Prochaska JM. Online health behavior and disease management programs: are we ready for them? Are they ready for us? J Med Internet Res 2005; 7 (3): e27.
- 14. Copeland J, Martin G. Web-based interventions for substance use disorders: a qualitative review. J Subst Abuse Treat 2004; 26: 109-116.
- 15. Kypri K, Sitharthan T, Cunningham JA, et al. Innovative approaches to intervention for problem drinking. Curr Opin Psychiatry 2005; 18: 229-234.
- 16. Bewick BM, Trusler K, Barkham M, et al. The effectiveness of web-based interventions designed to decrease alcohol consumption — a systematic review. Prev Med 2008; 47: 17-26.
- 17. Kypri K, McAnally HM. Randomized controlled trial of a web-based primary care intervention for multiple health risk behaviors. Prev Med 2005; 41: 761-766.
- 18. Flay BR, Biglan A, Boruch RF, et al. Standards of evidence: criteria for efficacy, effectiveness and dissemination. Prev Sci 2005; 6: 151-175.
- 19. Dunlap WP, Cortina JM, Vaslow JB, Burke MJ. Meta-analysis of experiments with matched groups or repeated measures designs. Psychol Methods 1996; 1: 170-177.
- 20. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005; 5: 13. doi: 10.1186/1471-2288-5-13.
- 21. Bersamin M, Paschall MJ, Fearnow-Kenney M, Wyrick D. Effectiveness of a Web-based alcohol-misuse and harm-prevention course among high- and low-risk students. J Am Coll Health 2007; 55: 247-254.
- 22. Bewick BM, Trusler K, Mulhern B, et al. The feasibility and effectiveness of a web-based personalised feedback and social norms alcohol intervention in UK university students: a randomised control trial. Addict Behav 2008; 33: 1192-1198.
- 23. Chiauzzi E, Green TC, Lord S, et al. My student body: a high-risk drinking prevention web site for college students. J Am Coll Health 2005; 53: 263-274.
- 24. Croom K, Lewis D, Marchell T, et al. Impact of an online alcohol education course on behavior and harm for incoming first-year college students: short-term evaluation of a randomized trial. J Am Coll Health 2009; 57: 445-454.
- 25. Doumas DM, McKinley LL, Book P. Evaluation of two Web-based alcohol interventions for mandated college students. J Subst Abuse Treat 2009; 36: 65-74.
- 26. Doumas DM, Hannah E. Preventing high-risk drinking in youth in the workplace: a web-based normative feedback program. J Subst Abuse Treat 2008; 34: 263-271.
- 27. Kypri K, Langley JD, Saunders JB, et al. Randomized controlled trial of web-based alcohol screening and brief intervention in primary care. Arch Intern Med 2008; 168: 530-536.
- 28. Saunders JB, Aasland OG, Babor TF, et al. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption — II. Addiction 1993; 88: 791-804.
- 29. Kypri K, Saunders JB, Williams SM, et al. Web-based screening and brief intervention for hazardous drinking: a double-blind randomized controlled trial. Addiction 2004; 99: 1410-1417.
- 30. Moore MJ, Soderquist J, Werch C. Feasibility and efficacy of a binge drinking prevention intervention for college students delivered via the Internet versus postal mail. J Am Coll Health 2005; 54: 38-44.
- 31. Neighbors C, Lee CM, Lewis MA, et al. Internet-based personalized feedback to reduce 21st-birthday drinking: a randomized controlled trial of an event-specific prevention intervention. J Consult Clin Psychol 2009; 77: 51-63.
- 32. Saitz R, Palfai TP, Freedner N, et al. Screening and brief intervention online for college students: the ihealth study. Alcohol Alcohol 2007; 42: 28-36.
- 33. Walters ST, Vader AM, Harris TR, et al. Dismantling motivational interviewing and feedback for college drinkers: a randomized clinical trial. J Consult Clin Psychol 2009; 77: 64-73.
- 34. Walters ST, Vader AM, Harris TR. A controlled trial of web-based feedback for heavy drinking college students. Prev Sci 2007; 8: 83-88.
- 35. Wechsler H, Lee JE, Kuo M, et al. Trends in college binge drinking during a period of increased prevention efforts. Findings from 4 Harvard School of Public Health College Alcohol Study surveys: 1993–2001. J Am Coll Health 2002; 50: 203-217.
- 36. White HR, Labouvie EW. Towards the assessment of adolescent problem drinking. J Stud Alcohol 1989; 50: 30-37.
- 37. McGee R, Kypri K. Alcohol-related problems experienced by university students in New Zealand. Aust N Z J Public Health 2004; 28: 321-323.
- 38. Buller DB, Borland R, Woodall WG, et al. Randomized trials on Consider This, a tailored, internet-delivered smoking prevention program for adolescents. Health Educ Behav 2008; 35: 260-281.
- 39. Patten CA, Croghan IT, Meis TM, et al. Randomized clinical trial of an Internet-based versus brief office intervention for adolescent smoking cessation. Patient Educ Couns 2006; 64: 249-258.
- 40. Tait RJ, Hulse GK. A systematic review of the effectiveness of brief interventions with substance using adolescents by type of drug. Drug Alcohol Rev 2003; 22: 337-346.
- 41. McCrone P, Knapp M, Proudfoot J, et al. Cost-effectiveness of computerised cognitive-behavioural therapy for anxiety and depression in primary care: randomised controlled trial. Br J Psychiatry 2004; 185: 55-62.
- 42. Christensen H, Griffiths K. Reaching standards for dissemination: a case study. Stud Health Technol Inform 2007; 129: 459-463.
- 43. Kentala J, Utriainen P, Pahkala K, Mattila K. Can brief intervention through community dental care have an effect on adolescent smoking? Prev Med 1999; 29: 107-111.
- 44. Colby SM, Monti PM, Barnett NP, et al. Brief motivational interviewing in a hospital setting for adolescent smoking: a preliminary study. J Consult Clin Psychol 1998; 66: 574-578.
- 45. Sussman S. Effects of sixty six adolescent tobacco use cessation trials and seventeen prospective studies of self-initiated quitting. Tob Induc Dis 2002; 1: 35-81.
- 46. Aveyard P, Cheng KK, Almond J, et al. Cluster randomised controlled trial of expert system based on the transtheoretical (“stages of change”) model for smoking prevention and cessation in schools. BMJ 1999; 319: 948-953.
- 47. Pallonen UE, Velicer WF, Prochaska JO, et al. Computer-based smoking cessation interventions in adolescents: description, feasibility, and six-month follow-up findings. Subst Use Misuse 1998; 33: 935-965.
- 48. Riley W, Jerome A, Behar A, Zack S. Feasibility of computerized scheduled gradual reduction for adolescent smoking cessation. Subst Use Misuse 2002; 37: 255-263.
- 49. Eysenck HJ. An exercise in mega-silliness. Am Psychol 1978; 33: 517.
- 50. Oakes M. The logic and role of meta-analysis in clinical research. Stat Methods Med Res 1993; 2: 147-160.
- 51. Bangert-Drowns RL. Review of developments in meta-analytic method. Psychol Bull 1986; 99: 388-399.
- 52. Rosenthal R, Rosnow RL. Essentials of behavioral research: methods and data analysis. New York: McGraw-Hill, 1984.
- 53. White V, Hayman J. Australian secondary school students’ use of alcohol in 2005. Canberra: Drug Strategy Branch, Australian Government Department of Health and Ageing, 2006.
- 54. Degenhardt L, Lynskey M, Hall W. Cohort trends in the age of initiation of drug use in Australia. Aust N Z J Public Health 2000; 24: 421-426.
- 55. Anthony JC, Petronis KR. Early-onset drug use and risk of later drug problems. Drug Alcohol Depend 1995; 40: 9-15.
Abstract
Objective: To conduct a systematic review of randomised trials of web-based interventions for problematic substance use by adolescents and young adults.
Data sources: An extensive search conducted in February 2009 of computer databases (MEDLINE, PsycINFO, Current Contents) and manual searches of key references.
Study selection: Randomised comparisons of fully automated web-based interventions specifically targeting adolescents and young adults (ie, typically school or tertiary students, ≤ 25 years old) versus other interventions.
Data synthesis: 16 relevant studies were identified, and data were extracted from 13 of the 14 reporting on alcohol use by young adults. The alcohol interventions had a small effect overall (d = − 0.22) and for specific outcomes (level of alcohol consumption, d = − 0.12; binge or heavy drinking frequency, d = − 0.35; alcohol-related social problems, d = − 0.57). The interventions were not effective (d = − 0.001) in preventing subsequent development of alcohol-related problems among people who were non-drinkers at baseline. Due to methodological differences, data from the two studies reporting on tobacco interventions among adolescents were not combined.
Conclusions: Based on findings largely from tertiary students, web interventions targeting alcohol-related problems have an effect about equivalent to brief in-person interventions, but with the advantage that they can be delivered to a far larger proportion of the target population. Web-based interventions to prevent the development of alcohol-related problems in those who do not currently drink appear to have minimal impact. There are currently insufficient data to assess the effectiveness of web-based interventions for tobacco use by adolescents.