School-based prevention for illicit drugs use: A systematic review

Available online at www.sciencedirect.com

Preventive Medicine 46 (2008) 385 – 396 www.elsevier.com/locate/ypmed

Review

School-based prevention for illicit drugs use: A systematic review Fabrizio Faggiano a,b,⁎, Federica D. Vigna-Taglianti a,b , Elisabetta Versino c , Alessio Zambon c , Alberto Borraccino c , Patrizia Lemma c a

Department of Experimental and Clinical Medicine, Avogadro University, Via Solaroli 17-28100 Novara, Italy b Piedmont Centre for Drug Addiction Epidemiology, Grugliasco, Italy c Department of Public Health, University of Torino, Italy Available online 3 December 2007

Abstract Objective. To evaluate the effectiveness of school-based interventions in preventing or reducing drug use. Methods. The search strategy was conducted according to the Cochrane Collaboration method, and applied to MEDLINE, EMBASE, ERIC, PSYCHINFO, Cochrane Library, ACP Journal Club, Cochrane Drugs and Alcohol Group Register. RCTs and CCTs evaluating school-based interventions designed to prevent substance use were reviewed. Data were extracted independently by two reviewers. Quality was assessed. Interventions were classified as skills, affective, and knowledge focused. Results. 29 RCTs were included; 28 were conducted in the USA; most were focused on 6th–7th grade students. Compared with usual curricula, skills-based interventions significantly reduce marijuana use (RR = 0.82; 95% CI: 0.73, 0.92) and hard drug use (RR = 0.45; 95% CI: 0.24, 0.85), and improve decision-making skills, self-esteem, peer pressure resistance (RR = 2.05; 95% CI: 1.24, 3.42) and drug knowledge. Compared with usual curricula, affective interventions improve decision-making skills and drug knowledge, and knowledge-focused programs improve drug knowledge. Skills-based interventions are better than affective ones in improved self-efficacy. No differences are evident for skills vs. knowledgefocused programs on drug knowledge. Affective interventions improve decision-making skills and drug knowledge to a higher degree than knowledge-focused programs. Conclusion. Skills-based programs help to deter drug use. Well designed, long-term randomised trials, and evaluation of intervention components are required. © 2007 Elsevier Inc. All rights reserved. Keywords: Drug addiction; Prevention; School-based programs; Systematic review; Effectiveness; RCT; Cochrane collaboration

Contents Introduction . . . . . . . . . . . . Methods. . . . . . . . . . . . . . Literature search and inclusion Data collection and extraction Outcomes . . . . . . . . . . . Quality assessment . . . . . . Statistical analysis . . . . . . Excluded studies . . . . . . . Included studies. . . . . . . . Results . . . . . . . . . . . . . .

. . . . . . . . criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

386 386 386 387 387 387 387 388 388 388

⁎ Corresponding author. Department of Experimental and Clinical Medicine, Avogadro University, Via Solaroli 17-28100 Novara, Italy. Fax: +39 0321 620421. E-mail address: [email protected] (F. Faggiano). 0091-7435/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ypmed.2007.11.012

386

Skills vs. usual curricula. . . . . Affective vs. usual curricula . . . Knowledge vs. usual curricula. . Skills vs. affective . . . . . . . . Skills vs. knowledge. . . . . . . Affective vs. knowledge . . . . . Interactive vs. passive techniques The role of peers . . . . . . . . Discussion . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . References . . . . . . . . . . . . . .

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

Introduction Since drug addiction is commonly described as a chronic, relapsing disease (Leshner, 1997), the best way to tackle it should be with prevention interventions. Schools serve as appropriate settings for illicit drugs use prevention programs for three reasons: first, four out of five drug users begin before adulthood, next schools offer the most systematic and efficient way of reaching a large number of young persons every year, and finally schools can adopt and enforce a broad spectrum of educational policies (UNODCCP, 2003). Therefore, most programs implemented are school based. According to Nancy Tobler (1986) programs can be divided into those based on (i) knowledge-only interventions, where description of the effects of drug use sets out to build negative attitudes toward drugs and hence decrease drug use; (ii) affectiveonly, for example self-esteem or self-awareness building interventions which are based on the assumption that psychological factors place people at risk of use; (iii) peer-based interventions, namely refusal skills and social life skills programs, the former focused on resistance skills or peer role models and the latter on inter-personal or intra-personal skills, both founded on the assumption that peer pressure can lead to drug use; (iv) knowledge plus affective interventions, in which knowledge is combined with affective education to provide values and build decision-making patterns; and (v) alternative approaches (activities and competence), such as interventions encouraging alternative activities to drug use or those aimed at enforcing control abilities. These approaches cannot be chosen based merely on theoretical considerations. The chain of causation of drug addiction is poorly understood as social and psychological factors, susceptibility, information about hazards and many other factors are involved (Hawkins et al., 1992, 2002). The choice must be pragmatic: “what matters is what works.” Moreover a positive association between an intervention, for example life skills based, and the reduction of drug use, from a scientific point of view would also confirm the role of the risk or protective factor aimed by the intervention (in the example, the life skills), whereas a negative result may reflect either a mistaken identification of the factor involved or the inefficacy of the intervention.

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

388 391 391 392 392 392 392 392 393 394 394 394

Many studies have evaluated the efficacy of drug use prevention programs, and several reviews have summarised their results (White and Pitts, 1998, 1997; Kroger, 1994; Hansen, 1992), including those regarded as the most relevant (Tobler and Stratton, 1997; Tobler et al., 2000). None however undertook a comprehensive assessment of the quality of study design, types of intervention, different outcomes, length of followup, and indications of which intervention type is the most effective. This paper presents a systematic review of the effectiveness of school-based interventions in reducing the incidence of first time usage and the prevalence of users, as well as intermediate variables, such as knowledge, and the development of specific skills or attitudes and behaviours. Methods Literature search and inclusion criteria All RCTs evaluating any intervention program vs. a control condition (usually curricular activities or another school-based drug prevention program) designed to prevent substance use in a school setting were considered. The search strategy was elaborated according to the Cochrane Collaboration rules. The following sources were searched: Medline (1966–February 2004), Embase (1988–February 2004), ERIC (1988–February 2004), Sociological Abstracts (1963–2000), Psychinfo (1967–February 2004), Cochrane Central Register of Controlled Trials (CCTR) (1st update 2004), ACP Journal Club (1991–February 2004), Cochrane Database of Systematic Reviews (DSR) (1st update 2004), Database of Abstracts of Reviews of Effects (DARE) (1st update 2004), Register of the Cochrane Drug and Alcohol Review Group (CDAG) (1st update 2004). Specific search methods were used for each database. No language restrictions were adopted. The following terms (limited to human studies) were included: substance-related disorders, addiction, abuse, use, drug dependence, drug addiction, illicit drug, narcotics, morphine, cannabis, heroin, heroin dependence, hashish, marijuana, ecstasy, MDMA, psychedelic agent, hallucinogens, cocaine, crack cocaine, lysergic-acid, LSD, designer drugs, Centers for Disease Control and Prevention, primary prevention, prevention, education program, health education, counselling, peer group, activities of daily living, psychological adaptation, adolescent psychology, interpersonal relations, social adjustment, social behaviour and life skills. Review articles and studies were then scanned to identify other significant studies. To gather unpublished research/results, research teams plus 18 authors of the included and excluded studies were contacted which resulted in six authors sending published and unpublished references. The target populations were primary or secondary school pupils and studies of special populations were excluded. The search identified 9657 reports (Fig. 1).

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

387

Fig. 1. Identification of eligible randomised controlled trials (RCTs).

Data collection and extraction

Outcomes

The 2216 abstracts retrieved were read by two reviewers to evaluate the relevance of the study and decide whether it should be included (Fig. 1). Thereafter, the full text of the 678 studies not excluded was obtained and independently assessed by two reviewers, in order to determine definitive inclusion. Of the 65 provisionally included reports (corresponding to 40 RCTs), 24 (21 RCTs) were excluded for methodological reasons. Data from the final pool of 41 reports included (29 RCTs) were then independently extracted by two reviewers using a standardised checklist. Disagreements were settled by a third reviewer. Fourteen authors were contacted by e-mail and invited to provide supplementary information. Eight replied and three provided the data requested which allowed for integration in the analysis. When no reply was received after 6 months, the studies concerned were re-evaluated for inclusion. The interventions and control arms of the studies were classified by the main component of the program as (i) skills focused, aimed to enhance students' abilities in general, refusal, and safety skills; (ii) affective focused, aimed to modify inner qualities (personality traits such as self-esteem and self-efficacy, and motivational aspects such as the intention to use drugs); (iii) knowledgefocused programs, aimed to enhance knowledge, effects, and consequences of drug use; and (iv) usual curricula. Table 2 reports, for multimodal interventions, the details of the components and other characteristics of the program. The interventions were also classified according to the type of the teaching method: (i) interactive programs, in which participants are actively involved in the activities: e.g., role playing, group discussion, peer leaders' involvement and (ii) didactic programs, consisting mainly of teacher-led lectures. Lastly, the studies were classified according to those involved in their administration: (i) teachers; (ii) external educators; and (iii) peers.

The following outcomes were considered: drug knowledge, drug attitudes, acquirement of personal skills (self-esteem, self-efficacy, decision-making skills, peer pressure resistance, assertiveness), peers/adults drug use, intention to use drugs, and use of drugs.

Quality assessment The quality of the studies included was assessed by two reviewers. Scores were assigned, according to the CDAG's check list (Amato et al., 2007), based on randomisation, allocation concealment, blinding, attrition, similarity of groups at baseline and balanced treatment among groups. Studies were then grouped in three classes: A: low risk of bias (scores 9–11); B: moderate risk of bias (scores 6–8); C: high risk of bias (scores 0–5). Disagreements were settled by a third reviewer. The results are illustrated in Table 1.

Statistical analysis Data were analysed with RevMan software (Clarke and Oxman, 2004). A standardised effect size was calculated for each outcome within each study. Wherever possible, a summary of relative risks and 95% confidence intervals were calculated with a random effects model (Deeks et al., 2001); for continuous outcomes measured in different ways a standardised mean difference (SMD) between groups was calculated to summarise results across studies. When two or more studies were included in the meta-analysis, a test of heterogeneity was applied. The study by Furr-Holden et al. (2004) included the adjustment for cluster effects and for confounding factors. Since the number of subjects included in the

388

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

Table 1 Methodological quality of randomised controlled trials included in the review Study

Attrition

Similarity of groups

Total Quality Cluster score class effect adjustment

Bernstein and Woodall (1987) Botvin et al. (1984, 1990b) Botvin et al. (1990a, 1995a) Botvin et al. (1994, 1995b) Botvin et al. (1997) Botvin et al. (2001) Clayton et al. (1991, 1996), and Lynam et al. (1999) Cook et al. (1984) Corbin et al. (1993) Dent et al. (2001) (Ellickson and Bell (1990), Bell et al. (1993), Ellickson et al. (1993a,b) Ellickson et al. (2003) Furr-Holden et al. (2004) Hansen et al. (1988) Hansen and Graham (1991), and Palmer et al. (1998) Hecht et al. (1993) Hurry and McGurk (1997) Jones et al. (1990) Jones et al. (1995) Kim et al. (1989) Malvin et al. (1985) Moskowitz et al. (1984) Ringwalt et al. (1991) Sexter et al. (1984) Sigelman et al. (2003) Snow et al. (1992), and Gersick et al. (1988) Sussman et al. (1998, 2002b) Sussman et al. (2002a, 2003) Werch et al. (1991)

0% = 3

Unclear = 0 6

B

–

9.6–24% = 3

Yes = 1

7

B

–

2 5 – 3 9 . 6 % Yes = 1 =2 16–40% = 2 Yes = 1

6

B

–

6

B

–

13% = 3 30.6% = 2 7–35% = 2

Yes = 1 Yes = 1 Yes = 1

6 6 6

B B B

– Yes –

10–17% = 3 23% = 3 37.1% = 2 18–40% = 2

Unclear = 0 Yes = 1 Yes = 1 Yes = 1

6 7 6 6

B B B B

– – Yes –

8.8% = 3

Unclear = 0 6

B

Yes

16% = 3

No = 0

6

B

Yes

30–39% = 2 22% = 3

No = 0 No = 0

5 6

C B

– Yes

Unclear = 0 10.9–17.9% =3 0% = 3 0% = 3 Unclear = 0 20–25% = 3 15% = 3

Unclear = 0 3 Yes = 1 7

C B

– –

Yes = 1 Yes = 1 Unclear = 0 Unclear = 0 No = 0

7 7 3 6 6

B B C B B

– – – – –

9.4% = 3

No = 0

6

B

–

Unclear = 0 7.2% = 3

Unclear = 0 3 Yes = 1 7

C B

– –

8.9–20.7% Yes = 1 =3

7

B

–

33% = 2

Yes = 1

6

B

–

44.6% = 0

Yes = 1

4

C

Yes

11% = 3

Yes = 1

7

B

–

Fourteen studies (Bernstein and Woodall, 1987; Botvin et al., 1984, 1990b, 1997; Clayton et al., 1991, 1996; Lynam et al., 1999; Cook et al., 1984; Ellickson and Bell, 1990; Bell et al. 1993; Ellickson et al., 1993a,b; Hansen et al., 1988; Hansen and Graham, 1991; Hecht et al., 1993; Dent et al., 2001; Werch et al., 1991; Palmer et al., 1998) did not present data suitable for the inclusion in the meta-analyses and their results are reported in the Results section in a narrative way. The effect of the low-quality studies on the overall results was determined by a sensitivity analysis, with either inclusion or exclusion of the class C studies. Publication bias was not assessed. The methods used are described on a greater detail in Faggiano et al. (2005).

Excluded studies Twenty-one (published in 24 reports) studies were excluded. Six (Ambtman et al., 1990; Bry, 1982; Rollin et al., 1995, 1994; Kim et al., 1993; Losciuto and Ausetts, 1988; Shope et al., 1998, 1996; Villalbi et al., 1993) were due to a failed randomisation process. Three (Botvin et al., 2000b; Griffin et al., 2003; Snow et al., 1997) presented secondary analysis of sub-samples of original trials (Botvin et al., 1990a, 1995a; Gersick et al., 1988; Snow et al., 1992; Botvin et al., 2001). In the studies by Graham et al. (1990), Eggert et al. (1994), and Thompson et al. (1997), three subsequent cohorts of students were pooled for the analysis, but the programs were different across the cohorts. In one study (Donaldson et al., 1994) the initial random assignment to the groups was not taken into account at the analysis stage and one study (Olton, 1985) gave no control group data. In two studies (Dent et al., 1998; Calafat et al., 1984) the randomisation procedure was unclear. In the study by De La Rosa Lopez (1995) the randomisation units were too limited to assure the validity of the method. In one study (O'Donnell et al., 1995) the randomisation procedure was only applied to a sub-sample. In the study by Short (1998) subjects assigned to intervention and control groups were subsamples of different populations. The study by Duncan et al. (2000) did not present any criteria for selecting the students. In the study by Schinke et al. (2000) the students enrolled were Native Americans and since the intervention was focused on their culture and traditions making results difficult to generalise.

Included studies A total of 29 studies (41 reports) were included (Table 2). Fourteen did not present data useful for the inclusion in the meta-analyses, because of either their statistical models or limited reporting of data from these models. Most studies (n = 18) investigated 6th and 7th grade students. In 18 studies the evaluation was based on post-test assessment; 13 provided data at 1-year follow-up. Few studies provided data for longer periods. All but one RCT were conducted in the USA, which took place in the UK. Most studies evaluated programs focused on skills (n = 25); affective programs were measured by six studies, and six included a knowledge-focused arm. Programs using interactive techniques were assessed in 27 studies. Administrators were external educators in 20 studies, teachers in 10, peer leaders in four, and policemen in two.

Results

All studies mentioned randomisation (score = 1) and for all studies allocation concealment was unclear (score = 1). Blinding was inadequate for all studies (score = 0). Experimental and control groups were equally treated in all studies (score = 1). Attrition scores: less than 25% of losses to follow-up = 3; 25–40% = 2; more than 40% = 0. multivariate model was not given, “adjusted numbers” of subjects in the intervention group were re-calculated for each outcome as the product of the adjusted RRs by the absolute numbers of the control group to allow the inclusion of the data in the meta-analysis.

Skills vs. usual curricula Skills-based interventions significantly reduced marijuana use as reported by four studies in the meta-analysis (RR = 0.82; CI 95%: 0.73, 0.92) (Table 3, Fig. 2 ). This result was unchanged when the study by Sussman et al. (2002a, 2003) designated as C quality class was excluded (RR = 0.81; CI 95%: 0.72, 0.91). Among studies not included in the meta-analysis, Botvin et al. (1990a, 1995a) showed a significant effect on marijuana use, whereas no significant differences were found by Ellickson and Bell (1990), Bell et al. (1993), and Ellickson et al. (1993a,b). Skills-based interventions have no effect on marijuana use when

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

389

Table 2 Characteristics of randomised controlled trials included in the review Study

Population/ Setting

Intervention

Arm focus a

Technique Administrator

Bernstein and Woodall (1987)

Alcohol and Substance Abuse Prevention Program (ASAP) Life Skills Training Program

End of the program, 8 Months

Interactive Older peers, teachers

1185 996

4 Months

Botvin et al. (1990a, 1995a)

7th Grade New York (USA)

Life Skills Training Program

Interactive Teachers

3597

6 Years

Botvin et al. (1994, 1995b)

7th Grade New York (USA)

Botvin et al. (1997)

7th Grade New York (USA)

Life Skills Training Program and culturally focused intervention Drug Abuse Prevention Program

EXP: affective CTR: usual curricula EXP: skills CTR: usual curricula EXP: skills CTR: usual curricula EXP: skills CTR: knowledge

Interactive External educators 33

Botvin et al. (1984, 1990b)

7th Grade Albuquerque (USA) 7th Grade New York (USA)

Botvin et al. (2001)

7th Grade New York (USA)

Drug Abuse Prevention Program

Clayton et al. (1991, 1996), and Lynam et al. (1999) Cook et al. (1984)

6th Grade Kentucky (USA)

Drug Abuse Resistance Education (DARE)

Junior and senior high school Milwaukee (USA) 3rd Grade Virginia (USA)

Positive Alternatives for Youth

Corbin et al. (1993)

Rehearsal plus and general information

Dent et al. (2001) (Ellickson and Bell (1990), Bell et al. (1993), Ellickson et al. (1993a,b) Ellickson et al. (2003)

9th–10th–11th Grades Project Towards Los Angeles (USA) No Drug abuse (TND) 7th Grade California Project ALERT and Oregon (USA)

7th Grade South Dakota (USA)

Project ALERT

Furr-Holden et al. (2004)

1st Grade USA

Hansen et al. (1988)

7th Grade Los Angeles (USA)

Classroom-centred and family–school partnership interventions Project SMART (Self Management and Resistance Training)

Hansen and Graham (1991), and Palmer et al. (1998) Hecht et al. (1993)

Analysis Follow-up period sample

Interactive External educators, 639 Passive Peers, teachers 456

4 Months

EXP: skills CTR: usual curricula EXP: skills CTR: usual curricula EXP: skills CTR: usual curricula

Interactive Teachers

725

3 Months

Interactive Teachers

3621

End of the program, 1 Year

Interactive Policemen

1002

End of the program, 7th–10th Grades 19–20 years of age

EXP: skills CTR: usual curricula EXP: affective, knowledge CTR: usual curricula EXP: skills CTR: usual curricula EXP: skills

Interactive External educators, teacher Interactive External educators

242

End of the program, 2 years

235 57

End of the program

Interactive External educators Interactive Teachers, peer leaders

679

1 Year

3852

12, 15, 36 and 60 Months, 24 Months

Interactive Teachers

4276

18 Months

Passive

370

6th–8th Grades

Interactive Teachers, peer leaders, external educators

1907

1 Year

Interactive External educators

2370 1635

Interactive External actors Passive

465 End of the program (unclear)

Interactive External educators

96

End of school year

Interactive External educators Passive Interactive External educators Passive

42

End of the program

34

End of the program

CTR: usual curricula

7th Grade Los Angeles and Orange (USA)

Adolescent Alcohol Prevention Trial (AAPT)

Drug Resistance Strategies Project

Hurry and McGurk (1997)

High school South Western (USA) 7–10 Years age Hackney (London)

Jones et al. (1990)

3rd Grade Virginia (USA)

Jones et al. (1995)

3rd Grade Virginia (USA)

EXP: skills CTR: usual curricula EXP: skills CTR: usual curricula EXP: affective, skills CTR: usual curricula EXP: skills CTR: knowledge

EXP: skills CTR: usual curricula Project CHARLIE EXP: skills (Chemical Abuse Resolution CTR: usual Lies in Education) curricula Rehearsal plus, traditional EXP: and attention control affective, skills CTR: knowledge Rehearsal plus and EXP: affective, general information knowledge CTR: usual curricula

Teachers

2 Years 1 Year, 2 Years

(continued on next page)

390

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

Table 2 (continued) Study

Population/ Setting

Intervention

Kim et al. (1989)

7th Grade Carolina (USA)

WHOA: A Great Way To Say No

Malvin et al. (1985) Moskowitz et al. (1984) Ringwalt et al. (1991) Sexter et al. (1984)

Sigelman et al. (2003) Snow et al. (1992), and Gersick et al. (1988) Sussman et al. (1998, 2002b) Sussman et al. (2002a, 2003) Werch et al. (1991)

Arm focus a

EXP: skills CTR: usual curricula 8th and 9th Grades Cross age tutoring EXP: skills and school store CTR: usual curricula 7th and 9th Grades NAPA project EXP: skills California (USA) CTR: usual curricula 5th and 6th Grades Drug Abuse Resistance EXP: skills North Carolina (USA) Education (DARE) CTR: usual curricula 5th through Humanistic education, peer EXP: skills, 9th Grades group, parent effectiveness, affective, know network and advocacy CTR: usual New York (USA) curricula 3rd through Basic, biologically enhanced EXP: skills 6th Grades and tobacco myths CTR: usual (USA) curricula 6th Grade Cognitive behavioural EXP: skills New England skills intervention CTR: usual (USA) curricula Continuation Project Towards No EXP: skills high schools Drug Abuse (TND) CTR: usual California (USA) curricula Continuation Project Towards No EXP: skills high schools Drug Abuse (TND) CTR: usual California (USA) curricula Elementary schools Keep A Clear EXP: skills Arkansas (USA) Mind Program CTR: usual (KACM) curricula

Technique Administrator

Analysis Follow-up period sample

Interactive External educators

367 End of the program (unclear)

Interactive External educators

43 21

End of the program

Interactive External educators

399 352

End of the program, 1 Year

Interactive Policemen

1270

End of the program

Interactive External educators Passive

1575 End of the program (unclear)

Interactive Self-administered, 337 external educators Passive Interactive External 1250 educators 1075

End of the program

Interactive External educators

1074

End of the program, 1 Year

Interactive External educators

575

2 Years

Interactive External educators, teachers

455

End of the program

End of the program, 2 Years

EXP: experimental group. CTR: control group. a Details on the classification of the components are given in the original publication (Faggiano et al., 2005).

continuous outcomes are used (Table 3). Among studies not included in the meta-analysis, Dent et al. (2001), Clayton et al. (1991, 1996), Lynam et al. (1999), and Cook et al. (1984) showed no effects, whereas Botvin et al. (1984, 1990b, 1997), Sexter et al. (1984) and Hecht et al. (1993) showed a positive effect of the skills-focused arm. Hard drug use was positively affected by the skills interventions (Fig. 3), when compared to usual curricula, in dichotomous variables (RR = 0.45; CI 95%: 0.24, 0.85) whereas it was not significant in continuous outcome (SMD = − 0.30; CI 95%: − 0.85, 0.25 – test for heterogeneity p b 0.0001 – data not shown) (Gersick et al., 1988; Snow et al., 1992; Sussman et al., 1998, 2002b) as well as the generic drug use (RR = 0.81; CI 95%: 0.64, 1.02) (Fig. 4) (Table 3); these results are confirmed by several studies not providing data for the meta-analysis (Botvin et al., 1997, Hecht et al., 1993; Dent et al., 2001), whereas generic drug use did not show differences at 1 year in the study by Cook et al. (1984), nor at 10 years in the study by Clayton et al. (1991, 1996) and Lynam et al. (1999). Skills-based intervention significantly reduced marijuana use in the meta-analysis of four studies (RR = 0.82; CI 95%: 0.73, 0.92) (Table 3, Fig. 2). This result was unchanged when the quality C study (Sussman et al., 2002a, 2003) was excluded (RR = 0.81; CI 95%: 0.72, 0.91). Among studies not included

in the meta-analysis, Botvin et al. (1990a, 1995a) showed a significant effect on marijuana use, whereas no significant differences were found by Ellickson and Bell (1990), Bell et al. (1993), and Ellickson et al. (1993a,b). Skills-based intervention has no effect on marijuana use when continuous outcomes are used (Table 3). Among studies not included in the metaanalysis, Dent et al. (2001), Clayton et al. (1991, 1996), Lynam et al. (1999), and Cook et al. (1984) showed no effects, whereas Botvin et al. (1984, 1990b, 1997), Sexter et al. (1984) and Hecht et al. (1993) showed a positive effect of the skillsfocused arm. Sexter et al. (1984) detected a positive effect on preventing the use of glues; however Botvin et al. (2001) found no effect on inhalant use on the continuous outcome and Furr-Holden et al. (2004) on the dichotomous outcome. Skills-based interventions significantly improved drug knowledge when compared to usual curricula (WMD = 2.60; CI 95%: 1.17, 4.03) (Table 3). This result is confirmed by some studies not included in the meta-analysis, Botvin et al. (1984, 1990b), Moskowitz et al. (1984) at 1-year follow-up and Botvin et al. (1990a, 1995a), while the studies by Malvin et al. (1985) did not show any significant differences for drug knowledge, and no differences in marijuana knowledge were shown by Werch et al. (1991) at post-test.

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

391

Table 3 Effect estimate comparing intervention and control groups, from stratified meta-analyses Comparison

Outcome

No. of References studies

Skills vs. usual curricula

Drug knowledge 1 Decision 2 making skills Drug use 2

No. of subjects Effect exp vs. ctrl measure

Effect estimate (CI 95%)

Test for heterogeneity ( p)

Hurry and McGurk (1997) Snow et al. (1992), Gersick et al. (1988), Hurry and McGurk (1997) Snow et al. (1992), Gersick et al. (1988) Ringwalt et al. (1991) Botvin et al. (1990a, 1995a), Sussman et al. (2002a, 2003), Ellickson et al. (2003), Furr-Holden et al. (2004) Sussman et al. (2002a, 2003), Furr-Holden et al. (2004)

48 vs. 37 646 vs. 583

WMD (fix) 2.60 (1.17, 4.03) SMD (rnd) 0.78 (0.46, 1.09)

– 0.09

1260 vs. 1111

RR (rnd)

0.81 (0.64, 1.02)

0.30

4072 vs. 3215

RR (rnd)

0.82 (0.73, 0.92)

0.37

392 vs. 354

RR (rnd)

0.45 (0.24, 0.85)

0.55

Drug knowledge 2 Decision 2 making skills

Corbin et al. (1993), Jones et al. (1995) Corbin et al. (1993), Jones et al. (1995)

36 vs. 27 36 vs. 27

SMD (rnd) SMD (rnd)

1.88 (1.27, 2.50) 1.35 (0.79, 1.91)

0.36 0.82

Knowledge vs. Drug knowledge 3 usual curricula Decision 2 making skills

Corbin et al. (1993), Jones et al. (1995), Sigelman et al. (2003) Corbin et al. (1993), Jones et al. (1995)

114 vs. 106

SMD (rnd)

0.91 (0.42, 1.39)

0.17

28 vs. 27

SMD (rnd)

− 0.06 (−0.60, 0.47) 0.34

Skills vs. affective

Drug knowledge 1 Self-efficacy 1

Jones et al. (1990) Jones et al. (1990)

15 vs. 15 16 vs. 15

WMD (fix) − 0.60 (−1.48, 0.28) – WMD (fix) 1.90 (0.25, 3.55) –

Skills vs. knowledge

Drug knowledge 2 Self-efficacy 2

Botvin et al. (1994, 1995b), Jones et al. (1990) 336 vs. 136 Botvin et al. (1994, 1995b), Jones et al. (1990) 336 vs. 136

SMD (rnd) SMD (rnd)

0.02 (− 0.18, 0.22) 0.13 (− 0.37, 0.63)

0.50 0.16

Affective vs. knowledge

Drug knowledge 3

51 vs. 40

SMD (rnd)

0.60 (0.18, 1.03)

0.94

36 vs. 28

SMD (rnd)

1.22 (0.33, 2.12)

0.11

336 vs. 336 vs. 199 vs. 200 vs.

SMD (rnd) SMD (rnd) RR (fix) RR (fix)

0.02 (− 0.18, 0.22) 0.13 (− 0.37, 0.63) 0.83 (0.59, 1.17) 0.46 (0.21, 0.99)

0.50 0.16 –

Affective vs. usual curricula

Interactive vs. passive techniques

Marijuana use

4

Hard drugs use

2

Decision making skills

2

Jones et al. (1990), Corbin et al. (1993), Jones et al. (1995) Corbin et al. (1993), Jones et al. (1995)

Drug knowledge Self-efficacy Marijuana use Hard drug use

2 2 1 1

Botvin et al. (1994, 1995b), Jones et al. (1990) Botvin et al. (1994, 1995b); Jones et al. (1990) Sussman et al. (2002a, 2003) Sussman et al. (2002a, 2003)

136 136 183 183

RR: relative risk; SMD: standardised mean difference; WMD: weighted mean difference. rnd: Random effect model; fix: fixed effect model.

Skills interventions were better in improving decisionmaking skills (SMD = 0.78; CI 95%: 0.46, 1.09) (Fig. 5), but not for Botvin et al. (1997, 1990a, 1995a), and peer pressure resistance (RR = 2.05; CI 95%: 1.24, 3.42), and slightly better in improving self-esteem (SMD = 0.22; CI 95%: 0.03, 0.40), when compared to usual curricula at post-test. The effect on peer pressure resistance is confirmed by Clayton et al. (1991, 1996) and Lynam et al. (1999), though self-esteem was lower in the intervention group at 10-year follow-up. In two studies (Cook et al., 1984; Botvin et al., 1990a, 1995a) self-esteem was significantly improved in the intervention arm.

(p = 0.004) and 2 years (p = 0.0003) of follow-up. In the study by Sexter et al. (1984) the use of psychedelic and stimulant drugs at post-test was significantly lower in the affective arm. Drug knowledge (SMD: 1.88, CI 95%: 1.27, 2.59) and decisionmaking skills (SMD: 1.35, CI 95%: 0.79, 1.91) were significantly improved in the affective arm compared with the usual curricula in two studies (Table 3). No significant differences were noted in knowledge and in self-reported behaviour in the study by Bernstein and Woodall (1987).

Affective vs. usual curricula

Drug knowledge was significantly improved at post-test (SMD = 0.91; CI 95%: 0.42, 1.39) when compared to the usual curriculum control group, whereas decision-making skills were not increased (Table 3). A knowledge arm was also evaluated in

In the study by Hansen et al. (1988) the affective group used significantly more marijuana than the controls at 1 year

Knowledge vs. usual curricula

392

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

Fig. 2. Cumulative meta-analysis of randomised trials comparing skills-focused intervention with usual curricular activities: marijuana use.

the Sexter et al. (1984) study, but no significant effect was showed. Skills vs. affective Jones et al. (1990) showed that skills-based interventions were better than affective in the improvement of self-efficacy (WMD = 1.90; CI 95%: 0.25, 3.55), but not in drug knowledge (Table 3). Skills vs. knowledge One study (Hansen et al., 1991; Palmer et al., 1998) evaluated the difference in marijuana use between two skillsfocused and a knowledge-focused programs: one skills-focused programs (normative education) reduced marijuana use at 1year follow-up but not at 2-year follow-up, whereas the other which focused on resistance training had no effect. Programs based on skills' enhancement were not different from knowledge-based ones on drug knowledge (SMD = 0.02; CI 95%: − 0.18, 0.22) (Table 3). Among studies not included in the meta-analysis, Botvin et al. (1994, 1995b) found a significant effect on marijuana knowledge in favour of the information-only control group, whereas no differences were evident in the improvement of decision-making skills (Botvin et al., 1994, 1995b), self-efficacy (Botvin et al., 1994, 1995b; Jones et al., 1990), and self-esteem (Botvin et al., 1994, 1995b).

Affective vs. knowledge By comparison with knowledge-based interventions, affective-focused interventions improve drug knowledge (SMD = 0.60; CI 95%: 0.18, 1.03) (Table 3) and decisionmaking skills (SMD = 1.22; CI 95%: 0.33, 2.12), whereas selfefficacy was unaffected in the study by Jones et al. (1990). Interactive vs. passive techniques We compared studies in which interactive as opposed to didactic techniques were used, irrespective of their type. Only three studies provided data suitable for meta-analysis: results were not statistically significant for drug knowledge, decisionmaking skills, self-esteem, self-efficacy, and marijuana use (Table 3). However, interactive techniques were more effective in reducing hard drug use in the study by Sussman et al. (2002a, 2003) (RR = 0.43; CI 95%: 0.19, 0.99). The role of peers When administered by peers as opposed to teachers, programs were significantly more effective with regard to marijuana use, knowledge and attitudes towards this drug at post-test (Botvin et al., 1984, 1990b). Marijuana attitudes at 1-year follow-up become lower in the teacher-led group (Botvin et al., 1984, 1990b). When compared with external educators, the effect of peers was evident for drug knowledge (WMD = − 3.42; CI 95%:

Fig. 3. Cumulative meta-analysis of randomised trials comparing skills-focused intervention with usual curricular activities: hard drug use.

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

393

Fig. 4. Cumulative meta-analysis of randomised trials comparing skills-focused intervention with usual curricular activities: drug use.

− 6.81, − 0.03) (Botvin et al., 1994, 1995b), but not significant for the other outcomes.

to those of the usual curricula, not different from resistance training programs, and less effective than normative education programs.

Discussion The findings have some limitations: In our systematic review the three groups of school-based prevention programs (skills, affective, and knowledge focused) revealed differences in efficacy. – Skills-focused programs appear to have a positive effect on both mediating variables (drug knowledge, decision making, self-esteem and peer pressure resistance) and final outcomes, compared to usual curricula. The pooled estimates showed a statistically significant 20% reduction of marijuana use in the intervention groups at the post-test, and 55% of hard drugs use. These results are consistent with those of studies not included in meta-analyses and persist in some studies even years after the intervention (for example, for marijuana 5 years in Furr-Holden et al., 2004 and 2 years in Sussman et al., 2002a, 2003), with most of the RCTs included having a satisfactory methodological quality (mainly quality score of B). – Affective-focused programs improve decision-making skills and drug knowledge compared to both usual curricula and knowledge-focused interventions, but no evidence of effectiveness is shown for prevention of drug use. – Knowledge-focused programs improve mediating variables (especially drug knowledge) compared with usual curricula, but are not more effective than skills-based programs, and are less effective than affective programs. When final outcomes are considered (drug use), their effects are comparable

– None of the RCTs satisfied all the quality criteria of the review. Nevertheless, we exclude that the findings can be fully explained by bias: for example, the main results on prevention of marijuana use come from three studies adjusting for the cluster effect out of four included in the meta-analysis, and three out of four studies with low level of attrition. In general, all but one of the studies included in the meta-analyses was assigned a quality score of B; – most results are outcomes at post-test and there are few data from long-term follow-ups. In the meta-analysis on marijuana use, two out of four studies included have two or more years of follow-up; – many studies did not present effect measures but only statistical indicators (f, p…) or other heterogeneous effect measures, so it was impossible to combine them in the metaanalysis; – the control for heterogeneity is not always satisfactory. Some sources of heterogeneity were covered by the design of the review (outcome, methods of the intervention, design and quality of the study), many other sources of “clinical” heterogeneity (grade of the target classes, intensity of the intervention, duration of follow-up) cannot be taken into account. There are not enough trials in the strata of each eligible variable to allow a meta-regression (Sterne et al., 2001).

Fig. 5. Cumulative meta-analysis of randomised trials comparing skills-focused intervention with usual curricular activities: decision-making skills.

394

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

Another limitation is that some complex structures, with a recognised role in determination of adolescent drug use, are not included in the studies. Peer, family and social context are strongly implicated in the causation of drug use in youth (Hawkins et al., 1992, 2002). Nevertheless, the aim of our review is to evaluate the role of intervention programs in a school setting, and the randomised model of evaluation allows their effects to be assessed apart from the other determinants of drug use. Despite these limits, the review suggested a consistent pattern of results: school-based programs based on life skills seem the most effective in reducing incidence of drug use. These programs are targeted to the individual-level risk and protective factors known to be associated with adolescent drug use; they are based on the concept that social and psychological factors are relevant in promoting the onset of drug use (Botvin, 2000a). From the results of the review the number needed to treat (NNT) (Altman, 1998) is 33 for marijuana use. This means that one out of every 33 students receiving the intervention will abstain from drug use because of it. Since the prevalence of marijuana use in the post-test of the control harm of the RCTs included in this comparison (see Fig. 2) was 16.5%, 5 out of 33 students (16.5% of 33) will use this drug. Of this, one would be prevented by the intervention, which corresponds to the 20% of the new initiators. Given that this estimation is based on four RCTs involving 7287 students, with the heterogeneity test negative, it can be considered reliable. The results of our work are consistent with those of Tobler et al. (2000) on drug use, and with the Cochrane reviews of alcohol prevention among young people (Foxcroft et al., 2004) and school-based smoking prevention (Thomas, 2004), though only in the short term. Two conclusions can be drawn from this consistency: first it is an indirect confirmation of the theory that merges the pathways of risk and risk factors for alcohol, tobacco and drug use among the young; second, it favours the delivery of a single school-level intervention to prevent the initial use of all harmful substances. Our results also provoke a more general comment on the question of effectiveness in this field: the vast amount of research undertaken especially since 1980 has not generated the evidence expected. We selected 50 RCTs, yet only a fraction of their data could be used for our review. Many studies were excluded because their quality was insufficient, thereby making the final pool 29 studies. It might have been supposed that this were enough to establish a sound and robust determination of the effectiveness of interventions for reducing drug use: this however was not the case. The wide variability in the indicators, scales and scores employed, and the limited reporting of data, made it difficult to summarise the evidence: the maximum number of RCTs comprised in a single meta-analysis was merely 4 out of 29. Examination of the main sources of bias for this very small sample shows that the validity and comparability of results are aspects that must be taken into consideration in future studies. The low validity of RCTs on primary prevention of drug addiction was due to (i) failure of the randomisation process where 11 out of 21 RCTs were excluded; (ii) attrition rates higher the 50% in two RCTs (five studies had attrition rates

higher than 30% and for three studies no information were given); and (iii) the uncontrolled cluster effect. This arises because children are the unit of analysis, but they are grouped into classes and schools making an entire school the typical unit randomised to an intervention arm. This procedure reduces the effective sample size and rises the random variability since there is a tendency for outcomes to show greater similarity between two children from the same cluster, as opposed to two children from different clusters (Campbell et al., 2001). Only six studies (Furr-Holden et al., 2004, Botvin et al., 2001, Dent et al., 2001, Sussman et al., 2002a, 2003, Hansen and Graham, 1991, Palmer et al., 1998; Ellickson et al., 2003) were designed to take account of the cluster effect. Lastly there is the question of generalisability. All but 1 of the 29 RCTs included were conducted in the USA. Since a nation's social context and drug policies have a significant influence on the effectiveness of its intervention programs, generalisation in this case would be hard to justify. Conclusions The results of this systematic review demonstrate that programs which develop individual social skills are the most effective form of school-level intervention for the prevention of the early drug use and should be selected, when planning community interventions against drug use. There are very little data on long-term effect of interventions. Our results need further corroboration in well-designed, long-term follow-up, randomised trials, particularly in countries other than the USA. Acknowledgments We thank Paola Petroni, Valentina Comba, Simonetta Lingua and Paride Angius for their assistance in the search strategies. Barbara Martin collected the full text papers and Federica Mathis made a valuable contribution to the extraction of data and assessment of the quality of the studies. John Iliffe acted as the language editor for the of the first version. We thank Sara Sanchez Del Mazo for the English revision of this version. Contributors: FF, PL and EV conceptualised the review; FVT, AZ and EV searched the literature and collected papers. FVT, FF, AB, AZ and EV reviewed the studies. FVT, FF and EV abstracted data from the papers for meta-analysis. FF wrote the Introduction, Results, Discussion and Conclusions sections. FVT wrote methods, description of studies and methodological quality. EV wrote the abstract and participated in the completion of the report. All authors commented on the final version. FF is entirely responsible for the paper. References Altman, D.G., 1998. Confidence intervals for the number needed to treat. BMJ 317, 1309–1312. Amato, L., Davoli, M., Ali, R., Faggiano, F., Farrell, M., Foxcroft, D., Ling, W., Vecchi, S., Zhao, C., Drugs and Alcohol Group, 2007. Cochrane Drugs and Alcohol Group. About The Cochrane Collaboration (Cochrane Review Groups (CRGs), Issue 2.

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396 Ambtman, R., Madak, P., Koss, D., Strople, M.J., 1990. Evaluation of a comprehensive elementary school curriculum-based drug education program. J. Drug Educ. 20 (3), 199–225. Bell, R.M., Ellickson, P.L., Harrison, E.R., Resnicow, K., Botvin, G.J., 1993. Do drug prevention effects persist into high school? How project ALERT did with ninth graders. Prev. Med. 22, 463–490. Bernstein, E., Woodall, W.G., 1987. Changing perceptions of riskiness in drinking, drugs, and driving: an emergency department-based alcohol and substance abuse prevention program. Ann. Emerg. Med. 16/12, 1350–1354. Botvin, G.J., 2000a. Preventing drug abuse in schools: social and competence enhancement approaches targeting individual-level etiological factors. Addict. Behav. 25, 887–897. Botvin, G.J., Baker, E., Renick, N., et al., 1984. A cognitive–behavioral approach to substance abuse prevention. Addict. Behav. 9 (2), 137–147. Botvin, G.J., Baker, E., Dusenbury, L., Tortu, S., et al., 1990a. Preventing adolescent drug abuse through a multimodal cognitive behavioral approach: results of a 3 year study. J. Consult. Clin. Psychol. 58, 437–446. Botvin, G.J., Baker, E., Filazzola, A.D., Botvin, E.M., 1990b. A cognitive– behavioral approach to substance abuse prevention: one-year follow-up. Addict. Behav. 15 (1), 47–63. Botvin, G.J., Schinke, S.P., Epstein, J.A., Diaz, T., 1994. Effectiveness of culturally focused and generic skills training approaches to alcohol and drug abuse prevention among minority youths. Psychol. Addict. Behav. 8 (2), 116–127. Botvin, G.J., Baker, E., Dusenbury, L., Botvin, E.M., Diaz, T., 1995a. Longterm follow-up results of a randomized drug abuse prevention trial in a white middle-class population. JAMA 273 (14), 1106–1112. Botvin, G.J., Schinke, S.P., Epstein, J.A., Diaz, T., Botvin, E.M., 1995b. Effectiveness of culturally focused and generic skills training approaches to alcohol and drug abuse prevention among minority adolescents: two-year follow-up results. Psychol. Addict. Behav. 9 (3), 183–194. Botvin, G.J., Epstein, J.A., Baker, E., Diaz, T., Ifill Williams, M., 1997. Schoolbased drug abuse prevention with inner-city minority youth. J. Child Adolesc. Subst. Abuse 6 (1), 5–19. Botvin, G.J., Griffin, K.W., Diaz, T., Scheier, L.M., Williams, C., Epstein, J.A., 2000b. Preventing illicit drug use in adolescents: long-term follow-up data from a randomized control trial of a school population. Addict. Behav. 25 (5), 769–774. Botvin, G.J., Griffin, K.W., Diaz, T., Ifill-Williams, M., 2001. Drug abuse prevention among minority adolescents: posttest and one-year follow-up of a school-based preventive intervention. Prev. Sci. 2 (1), 1–13. Bry, B.H., 1982. Reducing the incidence of adolescent problems through preventive intervention: one- and five-year follow-up. Am. J. Community Psychol. 10 (3), 265–276. Calafat, A., Amengual, M., Farres, C., Monserrat, M., 1984. Eficacia de un programa de prevencion sobre drogas, especialmente centrado en la “toma de decisiones”, segun sea desarrolado por especialistas o por profesores de los alumnos. Drogalcohol IX (3), 147–169. Campbell, M.K., Mollison, J., Grimshaw, J.M., 2001. Cluster trials in implementation research: estimation of intracluster correlation coefficients and sample size. Stat. Med. 20, 391–399. Clarke, M., Oxman, A.D. (Eds.), 2004. Cochrane Reviewers' Handbook 4.2.0. http://www.cochrane.dk/cochrane/handbookd/handbook.htm. Clayton, R.R., Cattarello, A., Walden, K.P., 1991. Sensation seeking as a potential mediating variable for school-based prevention intervention: a twoyear follow-up of DARE. Health Commun. 3 (4), 229–239. Clayton, R.R., Cattarello, A.M., Johnstone, B.M., 1996. The effectiveness of Drug Abuse Resistance Education (Project DARE): 5-year follow-up results. Prev. Med. 25 (3), 307–318. Cook, R., Lawrence, H., Morse, C., Roehl, J.A., 1984. An evaluation of the alternatives approach to drug abuse prevention. Int. J. Addict. 19 (7), 767–787. Corbin, S.K., Jones, R.T., Schulman, R.S., 1993. Drug refusal behavior: the relative efficacy of skills-based and information-based treatment. J. Pediatr. Psychol. 18 (6), 769–784. De La Rosa Lopez A. La prevencion de las drogodependencias en el ambito escolar. Una experiencia pratica evaluada. Barcelona: Thesis for Degree in Psychology, 1995.

395

Deeks, J.J., Altman, D.G., Bradburn, M.J., 2001. Statistical methods for examining heterogeneity and combining results from several studies in metaanalysis. In: Egger, M., Davey Smith, G., Altman, D.G. (Eds.), Systematic Reviews in Health Care. BMJ Publ. Group. Dent, C.W., Sussman, S., Hennesy, M., et al., 1998. Implementation and process evaluation of a school-based drug abuse prevention program: project towards no drug abuse. J. Drug Educ. 28 (4), 361–375. Dent, C.W., Sussman, S., Stacy, A.W., 2001. Project Towards No Drug Abuse: generalizability to a general high school sample. Prev. Med. 32 (6), 514–520. Donaldson, S.I., Graham, J.W., Hansen, W.B., 1994. Testing the generalizability of intervening mechanism theories: understanding the effects of adolescent drug use prevention interventions. J. Behav. Med. 17 (2), 195–216. Duncan, T.E., Duncan, S.C., Beauchamp, N., Wells, J., Ary, D.V., 2000. Development and evaluation of an interactive CD-ROM refusal skills program to prevent youth substance use: “refuse to use”. J. Behav. Med. 23 (1), 59–72. Eggert, L.L., Thompson, E.A., Herting, J.R., Nicholas, L.J., Dicker, B.G., 1994. Preventing adolescent drug abuse and high-school drop out through an intensive school-based social network development program. Am. J. Health Promot. 8 (3), 202–215. Ellickson, P.L., Bell, R.M., 1990. Drug prevention in junior high: a multi-site longitudinal test. Science 247 (4948), 1299–1305. Ellickson, P.L., Bell, R.M., Harrison, E.R., 1993a. Changing adolescent propensities to use drugs: results from Project ALERT. Health Educ. Q. 20 (2), 227–242. Ellickson, P.L., Bell, R.M., McGuigan, K., 1993b. Preventing adolescent drug use: long-term results of a junior high program. Am. J. Public Health 83 (6), 856–861. Ellickson, P.L., McCaffrey, D.F., Ghosh-Dastidar, B., Longshore, D.L., 2003. New inroads in preventing adolescent drug use: results from a large-scale trial of project ALERT in middle schools. Am. J. Public Health 93 (11), 1830–1836. Faggiano, F., Vigna-Taglianti, F.D., Versino, E., Zambon, A., Borraccino, A., Lemma, P., 2005. School-based prevention for illicit drugs' use. Cochrane Database Syst. Rev., Issue 2. Foxcroft, D.R., Ireland, D., Lister-Sharp, D.J., Lowe, G., Breen, R., 2004. Primary prevention for alcohol misuse in young people. Cochrane Database Syst. Rev., Issue 2. Furr-Holden, C.D.M., Ialongo, N.S., Anthony, J.C., Petras, H., Kellam, S.G., 2004. Developmentally inspired drug prevention: middle school outcomes in a school-based randomized prevention trial. Drug Alcohol Depend. 73 (2), 149–158. Gersick, K.E., Grady, K., Snow, D.L., 1988. Social–cognitive skill development with sixth graders and its initial impact on substance use. J. Drug Educ. 18 (1), 55–70. Graham, J.W., Johnson, C.A., Hansen, W.B., Flay, B.R., Gee, M., 1990. Drug use prevention programs, gender, and ethnicity: evaluation of three seventhgrade project smart cohorts. Prev. Med. 19 (3), 305–313. Griffin, K.W., Botvin, G.J., Nichols, T.R., Doyle, M.M., 2003. Effectiveness of a universal drug abuse prevention approach for youth at high risk for substance use initiation. Prev. Med. 36 (1), 1–7. Hansen, W.B., 1992. School-based substance abuse prevention: a review of the state of the art in curriculum, 1980–1990. Health Educ. Res. 7 (3), 403–430. Hansen, W.B., Graham, J.W., 1991. Preventing alcohol, marijuana, and cigarette use among adolescents: peer pressure resistance training versus establishing conservative norms. Prev. Med. 20 (3), 414–430. Hansen, W.B., Johnson, C.A., Flay, B.R., Graham, J.W., Sobel, J., 1988. Affective and social influences approaches to the prevention of multiple substance abuse among seventh grade students: results from project smart. Prev. Med. 17 (2), 135–154. Hawkins, J.D., Catalano, R.F., Miller, J.Y., 1992. Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: implications for substance abuse prevention. Psychol. Bull. 112, 64–105. Hawkins, J.D., Catalano, R.F., Arthur, M.W., 2002. Promoting science-based prevention in communities. Addict. Behav. 27, 951–976. Hecht, M.L., Corman, S.R., Miller-Rassulo, M., 1993. An evaluation of the Drug Resistance Project: a comparison of film versus live performance media. Health Commun. 5 (2), 75–88.

396

F. Faggiano et al. / Preventive Medicine 46 (2008) 385–396

Hurry, J., McGurk, H., 1997. An evaluation of a primary prevention programme for schools. Addict. Res. 5 (1), 23–38. Jones, R.T., McDonald, D.W., Fiore, M.F., Arrington, T., Randall, J., 1990. A primary preventive approach to children's drug refusal behavior: the impact of rehearsal-plus. J. Pediatr. Psychol. 15 (2), 211–223. Jones, R.T., Corbin, S.K.T., Sheehy, L., Bruce, S., 1995. Substance refusal: more than “Just Say No”. J. Child Adolesc. Subst. Abuse 4 (2), 1–26. Kim, S., McLeod, J.H., Shantzis, C., 1989. An outcome evaluation of refusal skills program as a drug abuse prevention strategy. J. Drug Educ. 19 (4), 363–371. Kim, S., McLeod, J.H., Shantzis, C., 1993. An outcome evaluation of Here's Looking At You 2000. J. Drug Educ. 23 (1), 67–81. Kroger, C.B., 1994. A review of the effectiveness of health education and health promotion. IUPHE—International Union for Health Promotion and Education, Utrecht. Leshner, A.I., 1997. Drug abuse and addiction treatment research: the next generation. Arch. Gen. Psychiatry 54, 691–694. Losciuto, L., Ausetts, M., 1988. Evaluation of a drug abuse prevention program: a field experiment. Addict. Behav. 13 (4), 337–351. Lynam, D.R., Milich, R., Zimmerman, R., et al., 1999. Project dare: no effects at 10-year follow-up. J. Consult. Clin. Psychol. 67. Malvin, J.H., Moskowitz, J.M., Schaps, E., Schaeffer, G., 1985. Evaluation of two school-based alternative programs. J. Alcohol Drug Educ. 30 (3), 98–108. Moskowitz, J.M., Malvin, J.H., Schaeffer, G.A., Schaps, E., 1984. An experimental evaluation of a drug education course. J. Drug Educ. 14 (1), 9–22. O'Donnell, J., Hawkins, J.D., Catalano, R.F., Abbott, R.D., Day, L.E., 1995. Preventing school failure, drug use, and delinquency among low-income children: long-term intervention in elementary schools. Am. J. Orthopsychiatr. 65 (1), 87–100. Olton, AL., 1985. The effect of locus of control and perceptions of school environment on outcome in three school drug abuse prevention programs. J. Drug Educ. 15 (2), 157–169. Palmer, R.F., Graham, J.W., White, E.L., Hansen, W.B., 1998. Applying multilevel analytic strategies in adolescent substance use prevention research. Prev. Med. 27 (3), 328–336. Ringwalt, C., Ennett, S.T., Holt, K.D., 1991. An outcome evaluation of Project DARE (Drug Abuse Resistance Education). Health Educ. Res. 6 (3), 327–337. Rollin, S.A., Rubin, R., Hardy Blake, B., Allen, P., et al., 1994. Project K.I.C.K., a school-based drug education research project: peers, parents and kids. J. Alcohol Drug Educ. 39 (3), 75–86. Rollin, S., Rubin, R., Marcil, R., Ferullo, U., Buncher, R., 1995. Project kick: a school-based drug education health promotion research project. Couns. Psychol. Q. 8 (4), 345–359. Schinke, S.P., Tepavac, L., Cole, K.C., 2000. Preventing substance use among Native American youth: three-year results. Addict. Behav. 25 (3), 387–3897. Sexter, J., Sullivan, A.P., Wepner, S.F., Denmark, R., 1984. Substance abuse: assessment of the outcomes of activities and activity clusters in school-based prevention. Int. J. Addict. 19 (1), 79–92. Shope, J., Copeland, L., Marcoux, B., Kamp, M., 1996. Effectiveness of a schoolbased substance abuse prevention program. J. Drug Educ. 26 (4), 323–337. Shope, J.T., Copeland, L.A., Kamp, M.E., Lang, S.W., 1998. Twelfth grade follow-up of the effectiveness of a middle school-based substance abuse prevention program. J. Drug Educ. 28 (3), 185–197.

Short, J.L., 1998. Evaluation of a substance abuse prevention and mental health promotion program for children of divorce. J. Divorce Remarriage 28 (3–4), 139–155. Sigelman, C.K., Bridges, L.J., Leach, D.B., et al., 2003. The efficacy of an education program to teach children a scientific theory of how drugs affect behavior. Appl. Dev. Psychol. 24, 573–593. Snow, D.L., Tebes, J.K., Arthur, M.W., Tapasak, R.C., 1992. Two-year follow-up of a social–cognitive intervention to prevent substance use. J. Drug Educ. 22 (2), 101–114. Snow, D.L., Tebes, J.K., Ayers, T.S., 1997. Impact of two social–cognitive interventions to prevent adolescent substance use: test of an amenability to treatment model. J. Drug Educ. 27 (1), 1–17. Sterne, J.A.C., Egger, M., Davey Smith, G., 2001. Investigating and dealing with publication and other biases. In: Egger, M., DSG, AD (Eds.), Systematic Reviews in Health Care. BMJ Publ Group. Sussman, S., Dent, C.W., Stacy, A.W., Craig, S., 1998. One-year outcomes of Project Towards No Drug Abuse. Prev. Med. 27 (4), 632–642. Sussman, S., Dent, C.W., Craig, S., Ritt-Olsen, A., McCuller, W.J., 2002a. Development and immediate impact of a self-instruction curriculum for an adolescent indicated drug abuse prevention trial. J. Drug Educ. 32 (2), 121–137. Sussman, S., Dent, C.W., Stacy, A.W., 2002b. Project towards no drug abuse: a review of the findings and future directions. Am. J. Health Behav. 26 (5), 354–365. Sussman, S., Sun, P., McCuller, W.J., Dent, C.W., 2003. Project Towards No Drug Abuse: two-year outcomes of a trial that compares health educator delivery to self-instruction. Prev. Med. 37 (2), 155–162. Thomas, R., 2004. School-based programmes for preventing smoking (Cochrane Review). The Cochrane Library, Update Software. Oxford, Issue 2. Thompson, E.A., Horn, M., Herting, J.R., Eggert, LL., 1997. Enhancing outcomes in an indicated drug prevention program for high-risk youth. J. Drug Educ. 27 (1), 19–41. Tobler, N.S., 1986. Meta-analysis of 143 adolescent drug prevention programs: quantitative outcome results of program participants compared to a control or comparison group. J. Drug Issues 16 (4), 537–567. Tobler, N.S., Stratton, H.H., 1997. Effectiveness of school-based drug prevention programs: a meta-analysis of the research. J. Prim. Prev. 18 (1), 71–128. Tobler, N.S., Roona, M.R., Ochshorn, P.M., Diana, G., Streke, A.V., Stackpole, K.M., 2000. School-based adolescent drug prevention programs: 1998 metaanalysis. J. Prim. Prev. 20 (4), 275–336. United Nations Office for Drug Control and Crime Prevention, 2003. SchoolBased Drug Education: A Guide for Practitioners and the Wider Community. Villalbi, J.R., Auba, J., Garcia Gonzalez, A., 1993. Results of a school addictive substances abuse prevention program: the Barcelona PASE pilot project. Gac. Sanit. 7 (35), 70–77. Werch, C.E., Young, M., Clark, M., Garrett, C., Hooks, S., Kersten, C., 1991. Effects of a take-home drug prevention program on drug-related communication and beliefs of parents and children. J. Sch. Health 61 (8), 346–350. White, D., Pitts, M., 1997. Health Promotion With Young People for the Prevention of Substance Misuse. Health Education Authority, London. White, D., Pitts, M., 1998. Educating young people about drugs: a systematic review. Addiction 93 (10), 1475–1487.